Technical Guide To SRAN Network Design (GO Applicable To SRAN10.0 & GBSS17.0 & BSC6910) - 20140902-A-1.0

Technical Guide to SRAN Network Design (GO Applicable to the SRAN10.
0 &
GBSS17.0 & BSC6910)
Product Name
Confidentiality Level
GSM BSC6910
Internal Public
Product Version
Total 258 pages
V900R017C00
Technical Guide to SRAN Network Design

(GO Applicable to SRAN10.0 & GBSS17.0 &
BSC6910)
(For internal use only)
Prepared By
fuqiang (employee ID:00283077)
Date
2014-07-18
Reviewed By
lishuanghua (employee ID: 00101863)
Date
2014-07-28
Date
2014-09-03
Li Yongqing (employee ID: 00141602)

chenyin (employee ID: 00179448)
Hu Chunhua (employee ID: 00257638)
Approved By
hepeng (employee ID: 00110002)
Huawei Technologies Co., Ltd.

All rights reserved
CONFIDENTIAL
CONFIDENTIAL
Technical Guide to SRAN Network Design (GO Applicable to the SRAN10.0 &
GBSS17.0 & BSC6910)
Change History
Version
Prepared/Revie
wed By
Date
Description
Approved By
V0.1
Li Bo
2012-11-30
Initial draft
Mei Weifeng,
Huang Yanzhong
V0.2
Li Bo
2012-12-07
The document is
modified according
to comments of the
delivery department.
Mei Weifeng
V0.3
Li Bo
2012-12-15
The document is
modified according
to comments of the
network information
service (NIS)
department.
Mei Weifeng
V0.4
Li Bo
2012-12-22
The document is
modified according
to review
comments.
Mei Weifeng
V0.5
Li Bo
2013-2-4
Section
18.2.2"Design
Examples" is
modified.
Songruining
V0.6
Li Bo
2013-5-7
Add a note about the

relationship between
traffic model and
BSC specification in
the contract:
Songruining
Specifications and
capacity
configuration of the
BSC must be based
on a certain traffic
model, all contracts
must be established
on a given traffic
model to ensure the
accuracy of the
contract. If you are
2015-11-13
Huawei Confidential
Page 2 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Version
Prepared/Revie
wed By
Date
Description
Approved By
unable to obtain
accurate traffic
model, we
recommend using
the default Huawei
traffic model for the
contract traffic
model.
V0.7
Tang Xiaoli
2013-06-19
Deleted eGSM and

used the eGBTS to
replace independent
NE.
Songruining
V0.8
Tang Xiaoli
2013-07-30
Added section
19.3"A Interface
Design (TDM)" and
section 19.6"Abis
Interface Design (IP
over E1)."
Songruining
V0.9
Tang Xiaoli
2013-11-22
Updated the
document to adapt
to the GBSS16.0
version.
Songruining
V1.0
Tang Xiaoli
2014-03-03
Revised the
document based on
TR5 review
comments.
Songruining
V1.1
Liuqi
2014-05-06
Add 22.5.2
Constraints
Songruining
V1.2
2014.07.18
Add VAMOS FR
Songruining
Add 16.4 source IP

route
2015-11-13
Huawei Confidential
Page 3 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Contents
Foreword....................................................................................... 19
1.1 Objectives.....................................................................................................................................................................19
1.2 Scope............................................................................................................................................................................20
1.3 Constraints....................................................................................................................................................................20
1.4 Dependency..................................................................................................................................................................20
2 Overview of Network Design........................................................21

3 Overall Guidance Principles.........................................................22
4 Overview of Key NEs...................................................................23
5 Overview of the Network Design Tool...........................................24
6 Important Reference Document...................................................25
7 Product Specifications.................................................................26
7.1 BSC Specifications.......................................................................................................................................................26
7.1.1 Hardware Capacity....................................................................................................................................................26
7.1.2 Estimation of BSC Configuration Capacity..............................................................................................................27
7.2 Board Specifications.....................................................................................................................................................28
7.2.1 BSC6910 Board Specifications.................................................................................................................................28
7.2.2 Service Processing Modules......................................................................................................................................28
7.2.3 Interface Modules......................................................................................................................................................31
8 BOQ Review Guide......................................................................34

8.1 Design Overview..........................................................................................................................................................34
8.1.1 Purpose of the Design................................................................................................................................................34
8.1.2 Input of the Design....................................................................................................................................................34
8.1.3 Contents of the Design..............................................................................................................................................34
8.1.4 Design Reference.......................................................................................................................................................34
8.2 Overview of Pre-sales Network Design.......................................................................................................................34
8.3 BOQ Review Principles...............................................................................................................................................35
8.4 CS Traffic Models........................................................................................................................................................36
8.5 PS Traffic Models.........................................................................................................................................................38
8.6 Relationship Between Traffic Model and Traffic Statistics..........................................................................................40
2015-11-13
Huawei Confidential
Page 4 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
9 Parameters for Capacity Calculation.............................................42

10 Capacity Calculation..................................................................42
11 Design of Resource Allocation....................................................44
11.1 Design Overview........................................................................................................................................................44
11.1.1 Purpose of the Design..............................................................................................................................................44
11.1.2 Input of the Design..................................................................................................................................................45
11.2 BSC Load Allocation..................................................................................................................................................45
11.2.1 Signaling Storm.......................................................................................................................................................46
11.3 BSC Board Layout Design.........................................................................................................................................52
11.3.1 Design Guide...........................................................................................................................................................52
12 Naming Rules Design.................................................................58

12.1 Design Overview........................................................................................................................................................58
12.2 NE Naming Rules.......................................................................................................................................................58
12.2.1 Naming Rules of Areas............................................................................................................................................58
12.2.2 Naming Rules of Offices.........................................................................................................................................59
12.2.3 Naming Rules of Manufacturers.............................................................................................................................59
12.2.4 Naming Rules of NEs..............................................................................................................................................60
12.2.5 Naming Rules of Signaling Points..........................................................................................................................61
12.3 NE Numbering Rules.................................................................................................................................................61
12.3.1 Numbering Rules of Entity IDs...............................................................................................................................61
12.3.2 Numbering Rules of BTS IDs.................................................................................................................................62
12.3.3 Numbering Rules of Cell IDs..................................................................................................................................62
12.3.4 Numbering Rules of LACs......................................................................................................................................62
12.3.5 Numbering Rules of MCCs and MNCs...................................................................................................................62
12.3.6 Numbering Rules of SPXs and DPXs.....................................................................................................................62
13 BSC6910 Networking Principles..................................................63

13.1 Technical Principles....................................................................................................................................................63
13.1.1 Overview.................................................................................................................................................................63
13.1.2 Technical Specifications..........................................................................................................................................67
13.1.3 Technical Description..............................................................................................................................................68
14 Optical Interface MSP................................................................70

14.1 MSP Design Guide.....................................................................................................................................................70
14.1.1 STM-1 Tributary Mode Selection...........................................................................................................................70
14.1.2 MSP Mode Selection...............................................................................................................................................70
14.1.3 Parameter Configuration.........................................................................................................................................71
14.1.4 S1 Configuration.....................................................................................................................................................72
14.1.5 C2 Configuration.....................................................................................................................................................73
14.1.6 MSP Support Capabilities of Boards.......................................................................................................................73
2015-11-13
Huawei Confidential
Page 5 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
14.2 MSP Technical Description........................................................................................................................................74
15 Detection Mechanism................................................................80
15.1 Restrictions of the Design..........................................................................................................................................80
15.2 BFD Detection............................................................................................................................................................82
15.3 ARP Detection............................................................................................................................................................84
15.4 IP PM Detection.........................................................................................................................................................84
16 IP Interworking Design..............................................................86
16.1 IP Planning on the BSC Side......................................................................................................................................86
16.2 IP Planning on the BTS Side......................................................................................................................................87
16.3 Routing Design on the BSC Side...............................................................................................................................88
16.4 Routing Design on the BTS Side................................................................................................................................88
16.5 VLAN Design.............................................................................................................................................................88
16.6 QoS Design.................................................................................................................................................................89
17 Network Topology Design...........................................................91

17.1 Design Overview........................................................................................................................................................91
17.2 Network Structure Design..........................................................................................................................................91
17.2.1 Design Guide...........................................................................................................................................................91
17.2.2 Typical Networking.................................................................................................................................................92
18 Reliability Design....................................................................102
18.1 Design Overview......................................................................................................................................................102
18.1.1 Purpose of the Design............................................................................................................................................102
18.1.2 Input of the Design................................................................................................................................................102
18.2 Network Reliability Design......................................................................................................................................102
18.2.1 Design Guide.........................................................................................................................................................102
18.2.2 Design Examples...................................................................................................................................................103
19 Transmission Interface Design..................................................118

19.1 Design Overview......................................................................................................................................................118
19.2 A Interface Design....................................................................................................................................................118
19.2.1 Interface Description.............................................................................................................................................118
19.2.2 Networking Design................................................................................................................................................119
19.2.3 SCTP Multi-Homing Design.................................................................................................................................129
19.2.4 Signaling Bandwidth Calculation..........................................................................................................................135
19.2.5 Signaling Configuration Principles.......................................................................................................................135
19.2.6 Traffic Bandwidth Calculation..............................................................................................................................135
19.2.7 IP Address Planning (A over IP)............................................................................................................................136
19.2.8 Routing Planning (A over IP)................................................................................................................................138
2015-11-13
Huawei Confidential
Page 6 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
19.2.9 QoS Design (A over IP).........................................................................................................................................139

19.2.10 Interface Interworking.........................................................................................................................................142
19.3 A Interface Design (TDM)........................................................................................................................................144
19.3.3 Transmission Bandwidth Design...........................................................................................................................146
19.3.4 Signaling Configuration Principles.......................................................................................................................146
19.4 Gb Interface Design..................................................................................................................................................148
19.4.3 Bandwidth Calculation..........................................................................................................................................157
19.4.4 IP Address Planning...............................................................................................................................................157
19.4.5 Routing Planning (Gb over IP)..............................................................................................................................159
19.4.6 QoS Design (Gb over IP)......................................................................................................................................160
19.4.7 Configuration Principles........................................................................................................................................161
19.4.8 Interface Interworking...........................................................................................................................................161
19.4.9 Interworking Instances..........................................................................................................................................165
19.5 Abis Interface Design...............................................................................................................................................168
19.5.4 IP Address Planning...............................................................................................................................................183
19.5.5 Routing Planning...................................................................................................................................................187
19.5.6 QoS Design............................................................................................................................................................188
19.5.7 Abis Port Allocation Design..................................................................................................................................191
19.6 Abis Interface Design (IP over E1)...........................................................................................................................192
19.6.3 Transmission Bandwidth Design...........................................................................................................................194
19.6.4 Configuration Principles........................................................................................................................................194
19.6.5 IP Planning............................................................................................................................................................194
19.6.6 Route Planning......................................................................................................................................................194
19.6.7 QoS Planning.........................................................................................................................................................194
19.6.8 Clock Synchronization..........................................................................................................................................195
19.7 Lb Interface Design..................................................................................................................................................195
19.7.2 Function Interaction...............................................................................................................................................196
19.7.3 Constraints and Limitations...................................................................................................................................196
19.7.5 Positioning Modes.................................................................................................................................................197
19.7.7 Parameter Design...................................................................................................................................................198
2015-11-13
Huawei Confidential
Page 7 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
19.8 BTS Homing Allocation...........................................................................................................................................199
20 Clock Synchronization Design...................................................201

20.1 Design Overview......................................................................................................................................................201
20.2 Clock Description.....................................................................................................................................................201
20.2.1 Definition of Synchronization...............................................................................................................................201
20.2.2 SyncE.....................................................................................................................................................................201
20.2.3 IEEE 1588 V2........................................................................................................................................................202
20.2.4 Advantages and Disadvantages of Clock Protocols..............................................................................................202
20.2.5 QoS Requirements of Clock Protocols..................................................................................................................205
20.3 Clock Source Selection.............................................................................................................................................205
20.4 Clock Design in Abis over TDM Mode...................................................................................................................206
20.5 Clock Design in Abis over IP Mode.........................................................................................................................207
20.6 Design of the IP Clock Server..................................................................................................................................209
21 Time Synchronization Design...................................................214

21.1 Design Overview......................................................................................................................................................214
21.2 Description of Time Synchronization.......................................................................................................................214
21.3 NTP...........................................................................................................................................................................214
21.4 Selection of a Time Synchronization Source............................................................................................................215
21.5 Transmission Mode..................................................................................................................................................215
21.6 Typical Networking..................................................................................................................................................215
21.7 Typical Application...................................................................................................................................................216
22 Function Design......................................................................217
22.1 Design of Broadcast Solutions for Cells..................................................................................................................217
22.1.1 Standard Broadcast................................................................................................................................................217
22.1.2 Simple Cell Broadcast...........................................................................................................................................222
22.2 Design of Radio Measurement Data Interface for Navigation (TOM-TOM)..........................................................223
22.2.1 Overview...............................................................................................................................................................223
22.2.2 Reference Document.............................................................................................................................................224
22.2.3 Limitations on Specifications................................................................................................................................224
22.2.4 Software and Hardware Configuration..................................................................................................................224
22.2.6 Bandwidth Design.................................................................................................................................................227
22.2.7 Time Synchronization............................................................................................................................................227
22.3 MOCN II Design......................................................................................................................................................227
22.3.1 Overview...............................................................................................................................................................227
22.3.3 Capacity Planning..................................................................................................................................................228
2015-11-13
Huawei Confidential
Page 8 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
22.3.4 Interface Design.....................................................................................................................................................228

22.4 Design of BSC Node Redundancy...........................................................................................................................229
22.4.1 Overview...............................................................................................................................................................229
22.4.2 Constraints.............................................................................................................................................................230
22.4.4 Capacity Planning..................................................................................................................................................233
22.4.5 Interface Design.....................................................................................................................................................234
22.5 LCS Function Design...............................................................................................................................................239
23 BTS Design.............................................................................. 243

23.1 BTS Cable Design....................................................................................................................................................243
23.2 Design Tool of the BTS Cable Diagram...................................................................................................................243
23.3 BTS Transmission Design........................................................................................................................................243
23.3.2 BTS Transmission.................................................................................................................................................243
23.3.3 eGBTS Networking...............................................................................................................................................246
24 OM Networking Design............................................................248
24.1 Design Overview......................................................................................................................................................248
24.1.2 Design Content......................................................................................................................................................248
24.1.3 Reference...............................................................................................................................................................248
24.2 Introduction to OMU................................................................................................................................................248
24.2.1 Standalone OMU...................................................................................................................................................248
24.2.2 Dual OMU.............................................................................................................................................................249
24.3 OM Networking Design...........................................................................................................................................250
24.3.1 Networking for Part of E1/T1 Timeslots...............................................................................................................250
24.3.2 Entire E1/T1 Networking......................................................................................................................................252
24.3.3 IP Networking........................................................................................................................................................253
24.3.4 Networking Instances............................................................................................................................................254
24.4 OM IP Address Planning..........................................................................................................................................255
24.5 Route Planning.........................................................................................................................................................256
24.6 Impact of eGBTS on the O&M................................................................................................................................256
2015-11-13
Huawei Confidential
Page 9 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figures
Figure 2-1 Position of the BSS network design in the entire network construction process................................21
Figure 11-1 Average service duration...................................................................................................................51
Figure 13-1 Port switchover..................................................................................................................................71
Figure 13-2 Board switchover...............................................................................................................................72
Figure 15-1 Diagram of the promoted commercial solution.................................................................................86
Figure 17-1 Networking of the BSC connected to a single MGW.......................................................................96
Figure 17-2 BSC/MGW multi-homing networking..............................................................................................97
Figure 17-3 MSC Pool networking mode 1..........................................................................................................98
Figure 17-4 MSC Pool networking mode 2..........................................................................................................99
Figure 17-5 Typical networking of the SGSN pool............................................................................................100
Figure 17-6 All-IP networking............................................................................................................................101
Figure 17-7 Typical IP-based networking...........................................................................................................101
Figure 17-8 Hybrid networking..........................................................................................................................102
Figure 17-9 Logical networking of the transmission resource pool...................................................................103
Figure 17-10 Physical networking of the transmission pool with active/standby boards...................................103
Figure 17-11 Physical networking of the transmission pool with independent boards......................................104
Figure 18-1 Improving reliability by active/standby links on ports....................................................................106
Figure 18-2 Reliability design of the Gb interface.............................................................................................107
Figure 18-3 Reliability design of IP transmission routes....................................................................................108
Figure 18-4 Reliability design of IP transmission routes....................................................................................108
Figure 18-5 BSC/MGW multi-homing networking............................................................................................109
Figure 18-6 MSC Pool networking mode 1........................................................................................................110
Figure 18-7 MSC Pool networking mode 2........................................................................................................111
Figure 18-8 Typical networking diagram of the SGSN pool..............................................................................112
Figure 18-9 IP networking topology of A interface boards based on the dynamic loading balancing...............112
Figure 18-10 Standalone EOMU........................................................................................................................113
2015-11-13
Huawei Confidential
Page 10 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 18-11 Dual EOMUs.................................................................................................................................114

Figure 18-12 Clock subsystem of the BSC6910.................................................................................................115
Figure 18-13 BSC belonging to two layer-2 transmission devices in the dual-homing mode............................117
Figure 18-14 BSC belonging to one layer-2 transmission device in the single-homing mode...........................117
Figure 18-15 Inter-board link aggregation in the inter-board pool networking scenario....................................118
Figure 18-16 Manual active/standby LAGs on the BSC side+router adopting the VRRP networking mode....119
Figure 18-17 LAG of the active/standby board+router adopting the VRRP networking mode.........................120
Figure 19-1 Reference protocol model on the control plane of the A interface..................................................122
Figure 19-2 Reference protocol model on the user plane of the A interface......................................................122
Figure 19-3 BSC/MGW multi-homing networking............................................................................................123
Figure 19-4 Typical A over IP networking mode (pool of standalone boards)...................................................124
Figure 19-5 Typical A over IP networking mode (pool of standalone boards)...................................................126
Figure 19-6 Typical A over IP networking mode (pool of active/standby interface boards+dual-active ports). 128
Figure 19-7 Typical A over IP networking mode (pool of active/standby boards+manual active/standby LAGs)
.............................................................................................................................................................................129
Figure 19-8 SCTP four-homing between the BSC and the MSC server.............................................................133
Figure 19-9 Two M3UA links and SCTP four-homing between the BSC and the MSC server.........................134
Figure 19-10 SCTP dual-homing on the MSC server side and SCTP single-homing on the BSC side (1)........135
Figure 19-11 SCTP dual-homing on the MSC server side and SCTP single-homing on the BSC side (2)........136
Figure 19-12 SCTP single-homing on the MSC server side and SCTP dual-homing on the BSC side.............137
Figure 19-13 IP network topology of the BSC...................................................................................................141
Figure 19-14 Promoted detection mode in active/standby mode........................................................................143
Figure 19-15 Reference protocol model on the control plane of the A interface................................................148
Figure 19-16 Gb over IP protocol stack..............................................................................................................155
Figure 19-17 Embedded PCU networking..........................................................................................................156
Figure 19-18 Direction connection (Gb over IP)................................................................................................156
Figure 19-19 IP transmission network connection (Gb over IP)........................................................................157
Figure 19-20 Typical Gb over IP networking mode (active/standby boards+manual active/standby LAGs)....158
Figure 19-21 Typical Gb over IP networking mode (active/standby boards+dual-active ports)........................161
Figure 19-22 Logical connection between the NS layer and the SSGP layer.....................................................169
Figure 19-23 Abis over HDLC interface protocol..............................................................................................175
Figure 19-24 TDM networking when the Abis interface adopts STM-1 transmission.......................................177
Figure 19-25 IP networking when the Abis adopts MSTP transmission............................................................177
Figure 19-26 IP networking when the Abis adopts data network transmission..................................................177
2015-11-13
Huawei Confidential
Page 11 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 19-27 BTS networking diagram..............................................................................................................178

Figure 19-28 Two E1s connected to different interface boards..........................................................................181
Figure 19-29 Two E1s connected to different ports on the same interface board...............................................181
Figure 19-30 Typical A over IP networking mode (pool of active/standby boards+manual active/standby
LAGs+single IP address).....................................................................................................................................182
Figure 19-31 Typical A over IP networking mode (pool of active/standby boards+dual-active ports+single IP
address)................................................................................................................................................................184
Figure 19-32 SMLC-based network topology for the Lb interface....................................................................203
Figure 19-33 Direct connection between the BSC and the SMLC.....................................................................204
Figure 19-34 Connection through STP...............................................................................................................204
Figure 20-1 Clock networking instance 1...........................................................................................................213
Figure 20-2 Clock networking instance 2...........................................................................................................213
Figure 20-3 MSTP-based GSM IP solution........................................................................................................216
Figure 20-4 IP Clock synchronization networking.............................................................................................217
Figure 21-1 Typical networking for time synchronization.................................................................................222
Figure 22-1 Network topology of the cell broadcast..........................................................................................225
Figure 22-2 Cable connection diagram between the interface board and the CBC............................................226
Figure 22-3 Topology of the simple cell broadcast system................................................................................230
Figure 22-4 Logical networking for the TOM-TOM..........................................................................................232
Figure 22-5 Physical networking on the VNP interface.....................................................................................233
Figure 22-6 Networking of the active/standby OMUs with a single port and directly connected routers.........233
Figure 22-7 Networking for time synchronization.............................................................................................234
Figure 22-8 Logical structure of the LCS system on the GSM network............................................................235
Figure 22-9 Logical structure of the NSS-based SMLC.....................................................................................237
Figure 22-10 Logical structure of the BSS-based SMLC...................................................................................237
Figure 22-11 LCS flow initiated by an external LCS client...............................................................................238
Figure 23-1 Networking topology change of the eGBTS...................................................................................243
Figure 23-2 Change of northbound and southbound interfaces of the eGBTS...................................................243
Figure 24-1 Standalone OMU.............................................................................................................................245
Figure 24-2 Dual OMUs.....................................................................................................................................246
Figure 24-3 25-pin D model interface.................................................................................................................247
Figure 24-4 Networking for part of E1/T1 timeslots..........................................................................................248
Figure 24-5 Entire E1/T1 Networking................................................................................................................249
Figure 24-6 OM network topology.....................................................................................................................249
2015-11-13
Huawei Confidential
Page 12 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 24-7 IP networking in dual OMU mode..................................................................................................250

Figure 24-8 OM E1 networking instance 1.........................................................................................................250
Figure 24-9 OM E1 networking instance 2.........................................................................................................251
Figure 24-10 Change of the OM structure..........................................................................................................252
2015-11-13
Huawei Confidential
Page 13 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Tables
Table 7-1 Typical maximum configuration of HW69 R13 boards in BSC6900 GSM where the BM and TC are
integrated...............................................................................................................................................................21
Table 7-2 Typical maximum configuration of HW69 R13 boards in BSC6900 GSM where the BM and TC are
separated................................................................................................................................................................22
Table 7-3 Typical maximum configuration of HW69 R13 boards in BSC6900 GSM where the Abis over TDM
or A over IP is adopted...........................................................................................................................................22
Table 7-4 Typical maximum configuration of HW69 R13 boards in BSC6900 GSM where the Abis over IP or A
over IP is adopted...................................................................................................................................................22
Table 7-5 Board specifications..............................................................................................................................24
Table 8-1 Basic PS traffic model (new in the R13)...............................................................................................31
Table 8-2 PS user model........................................................................................................................................31
Table 8-3 PS coding ratio and average rate...........................................................................................................32
Table 8-4 Performance counters corresponding to basic procedures....................................................................32
Table 9-1 BSC capacity planning table.................................................................................................................35
Table 10-1 Manufacturer short names...................................................................................................................43
Table 10-2 NE short names...................................................................................................................................44
Table 12-1 MSP advantages and disadvantages....................................................................................................53
Table 12-2 MSP support capabilities of the boards of the controller....................................................................56
Table 12-3 Framing mode comparison..................................................................................................................69
Table 12-4 Optical interface interworking parameters..........................................................................................70
Table 13-1 Restrictions of the fault detection mechanism of the controller.........................................................73
Table 13-2 Restrictions of the fault detection mechanism of the base station......................................................74
Table 17-1 Calculation result of A interface bandwidth in TDM transmission mode.........................................121
Table 17-2 Calculation result of A interface bandwidth in IP transmission mode..............................................121
Table 17-3 A interface interworking parameters.................................................................................................129
Table 17-4 Design principles of A interface networking.....................................................................................133
Table 17-5 Configuration of O&M links for the Ater interface..........................................................................136
Table 17-6 Configuration of signaling links for the Ater interface.....................................................................136
2015-11-13
Huawei Confidential
Page 14 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Table 17-7 Gb over IP interworking parameters.................................................................................................151

Table 17-8 Gb over IP interworking parameters.................................................................................................153
Table 17-9 Performance test results of parameters a, b, c, and d........................................................................170
Table 17-10 Estimates of data related to parameters a, b, c, and d.....................................................................170
Table 18-1 Support for 2G-based 1588v2 clocks................................................................................................189
2015-11-13
Huawei Confidential
Page 15 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Technical Guide to SRAN Network Design (GO

Applicable to SRAN10.0&GBSS17.0&BSC6910)
Keywords: network design

Abstract: The BSC6910 is introduced in the GBSS15.0. The BSC6910 R15 has less
networking scenarios than the BSC6900. In GBSS17.0, the networking is enhanced.
Specifically, the BSC6910 (configured with the POUc) supports A over TDM networking.
The following describes the networking scenarios of the BSC6910:
The BSC6910 does not support an external PCU, without any Pb interface.
The BSC6910 does not support TC, remote TC subracks, or local independent TC
subracks, without any Ater interface.
The BSC6910 does not support Abis over HDLC.
The A interface does not support IP over E1/T1.
Calculation of the BSC6910 capacity does not require calculation of Ater or HDLC
transmission. Contents considering the deleted networking scenarios are removed from this
document, for example, TC Pool and local switching. For details, see this document.
The following table lists the differences in network design between the GBSS17.0 BSC6900
and BSC6910:
2015-11-13
Item
BSC6900
BSC6910
Resource allocation
Not supports the EXOUa in 10

GE or EGPUa. Supports TDM
exchange, TNU boards, EIUa,
EIUb, OIUa, OIUb, FG2a,
FG2c, FG2d, PEUa, PEUc,
GOUa, GOUc, GOUd, GOUe,
XPUa, XPUb, and
DPUa/c/d/e/f/g.
Supports the EXOUa in 10 GE

and EGPUa, FG2c, GOUc,
POUc, GOUd, GOUe, and
FG2d boards. Not supports
TDM exchange, TNU boards,
and TC subracks.
Capacity
Supports the calculation of the CS

traffic volume, the number of
BSC subscribers, the BSC CS
BHCA, the CIC of A interface and
Ater interface, the number of
PDCH, IWF resources, the
TDM&IP and IP&IP using the
IWF, and the Gb interface
throughput.
Supports the calculation of the

CS traffic volume, the number
of BSC subscribers, the BSC CS
BHCA, the CIC of A interface,
the number of PDCH, IWF
resources, the TDM&IP and
IP&IP using the IWF, and the
Gb interface throughput. Not
supports the CIC calculation of
Ater interface.
Overall
capacity
calculati
on of the
BSC
Huawei Confidential
Page 16 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
2015-11-13
Bandwid
th
calculati
on of
Abis
interface
Supports the bandwidth

calculation of Abis over TDM
over E1, Abis over TDM over
STM1, Abis over IP over E1, Abis
over IP over STM1, Abis over IP
over
FE/GE(electrical)/GE(optical),
and Abis over HDLC over E1.
Not supports the bandwidth
calculation of 10GE(optical)
interface.
Supports bandwidth calculation

of Abis over TDM over STM1
and Abis over IP over
FE/GE(electrical)/GE(optical)/1
0GE(optical).
Bandwid
th
calculati
on of A
interface

calculation of A over TDM over
E1, A over TDM over STM1, A
over IP over E1, A over IP over
STM1, and A over IP over
FE/GE(electrical)/GE(optical).
Supports the calculation of the
number of M3UA links over IP.
calculation of A over IP over 10
GE(optical).
Supports bandwidth calculation

of A over IP over
0GE(optical), and the
calculation of the number of
M3UA links over IP.
Ater
interface

calculation of Ater over TDM
over E1, Ater over TDM over
STM1, and Ater over IP over
STM1.
Not supports this interface.
Pb
interface

calculation of Pb interface circuits
and the calculation of the
bandwidth occupied by Pb
interface links.
Not supports this interface.
Gb
interface

calculation of GB over FR over
TDM E1, GB over FR over TDM
STM1, and Gb over IP over
FE/GE(electrical)/GE(optical).
calculation of Gb over IP over 10
GE(optical).
Not supports Gb over FR over

TDM E1 or Gb over FR over
TDM STM1. Supports Gb over
IP over
0GE(optical).
Naming rules
The value of both BTS ID and

CELL ID ranges from 0 to 2047.
The value of DPX (integer)
ranges from 0 to 186.
The value of both BTS ID and

CELL ID ranges from 0 to 7999.
The value of DPX (integer)
IP networking
Supports the IP design on the

BSC and BTS sides, the route
design on the BSC and BTS sides,
the VLAN design, and the QoS
design.
Supports the IP design on the

BSC and BTS sides, the route
design on the BSC and BTS
sides, the VLAN design, and the
QoS design. Not supports the IP
path design.
Huawei Confidential
Page 17 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Network topology
Supports the TDM networking,

the BSC and MGW singlehoming networking, the BSC and
MGW multi-homing networking,
the MSC Pool networking, the
SGSN pool networking, all-IP
networking, hybrid networking,
and the transmission resource
pool networking.
Not supports the TDM

networking or TC subracks.
Therefore, networking in the
BM and TC separated mode
does not exist. Supports the
BSC and MGW single-homing
networking, the BSC and MGW
multi-homing networking, the
MSC Pool networking, the
SGSN pool networking, all-IP
networking, hybrid networking,
and the transmission resource
pool networking.
Reliability
Supports the reliability design of

active/standby port links, loadbalancing, data configuration,
multiple transmission channels,
the VRRP in IP networking, the
SCTP multi-homing, the BSC
multi-homing MGWs, the MSC
pool, the SGSN pool, the
transmission resource pool over A
interface, the OM, clock, and
Ethernet link aggregation.
Supports the reliability design of

active/standby port links, loadbalancing, data configuration,
multiple transmission channels,
the VRRP in IP networking, the
SCTP multi-homing, the BSC
multi-homing MGWs, the MSC
pool, the SGSN pool, the
transmission resource pool over
A interface, the OM, clock, and
Ethernet link aggregation. Not
supports the reliability design of
the TC pool.
A
interface
Supports A over TDM, A over IP

over
FE/GE(electrical)/GE(optical),
and A over IP over E1.
Not supports A over IP over E1.

Supports A over IP over
0GE(optical).
Abis
Interface
Supports Abis over TDM over

STM1, Abis over TDM over E1,
Abis over IP, and Abis over
HDLC.
Supports Abis over TDM over

STM1 and Abis over IP over
EF/GE(electrical)/10GE(Optical
).
Gb
interface
Supports Gb over FR and Gb over

IP.
Supports Gb over IP. Not

supports Gb over FR.
Transmis
sion
interface
2015-11-13
Huawei Confidential
Page 18 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Ater
interface
Supported
Not supported
Pb
interface
Supported
Not supported
Supports line clock, GPS, BITS

clock, and external clock.
Not supports line clock.

Supports GPS, BITS clock, and
external clock.
Clock
synchronization
Foreword
1.1 Objectives
This document guides global system for mobile communications (GSM) base station
subsystem (BSS) network design engineers through the network design and delivery of GSM
BSS establishment, migration, expansion, and optimization. With the help of this document, a
GSM BSS network design engineer can use high-level design (HLD) and low-level design
(LLD) templates for GSM BSS network design to work out a final GSM BSS network design
report for a customer.
A network design report consists of the HLD and LLD. The HLD provides the customer with
the design of the network topology, networking, transmission, interfaces, resource capacity,
function services, operation and maintenance (O&M), clock, and time synchronization. This
document covers all the guidance principles. The LLD is intended for engineering guidance,
and provides the design of the device board layout, cable connections, and key data
configuration. You can use the network equipment planning (NEP) tool to generate the LLD.
2015-11-13
Huawei Confidential
Page 19 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
1.2 Scope
This document describes the design principles, design methods, and output formats of the
BSS networking, transmission, interfaces, services, and O&M. The core network elements
(NEs) involved in BSS network design are the base station controller (BSC), packet control
unit (PCU), and BTS, and the involved interface NEs are the mobile switching center (MSC)
server, media gateway (MGW), M2000, serving GPRS support node (SGSN), and local
maintenance terminal (LMT).
1.3 Constraints
This document is developed based on GBSS17.0 BSC6910 and is applicable to the GSM
Only mode of the BSC6910. Network design of the BSC6900 is described in Technical Guide
to Single RAN Network Design V100R003 (GO applicable to
SRAN10.0&GBSS17.0&BSC6910). The GU mode is described in the Single RAN network
design guide.
1.4 Dependency
Before the network design, you must collect the required data based on the information
collection template for network design. During the network design, you need to
effectively communicate with the operator and core network engineers to ensure that the
required information is accurate and the change causes and change results are recorded.
The network design personnel must be global technical service (GTS) engineers who are
familiar with the BSC6910 and are engaged in engineering or maintenance for more than
one year.
The network design guide is updated based on changes in the BSC and application
scenarios and is available at http://support.huawei.com. You can obtain the latest version
of the guide from the following path:
Documentation > Wireless > Wireless Public > Wireless Professional Services
Product > Technical Guides
2015-11-13
Huawei Confidential
Page 20 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Overview of Network Design
The BSS network design service is provided in the engineering preparation and delivery
stage. The network planning (NP) provided by the network design department of Sales &
Services, the network development planning provided by the operator, and the radio network
plan provided by the network planner are the input of the HLD and LLD. The BSS network
design guides the follow-up network deployment design and engineering.
Figure 1.1 shows the position of the BSS network design in the entire network construction
process:
Figure 1.1 Position of the BSS network design in the entire network construction process
The GSM BSS network design service involves the overall designs of the networking,
transmission, interfaces, resource capacity, functional services, O&M, and clock of the
network. Focusing on the security, balance, and extensibility of the network, the GSM BSS
network design provides guidance for engineering and construction and guarantees highquality network operation for operators.
2015-11-13
Huawei Confidential
Page 21 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Overall Guidance Principles
Generally, the scale of GSM BSS network construction is large, numerous NEs are involved,
and the interface interworking is complicated. The BSS network design principles are as
follows:
Principle of area-based design

The BSS network design is implemented based on the network construction plan of the
operator, areas, and stages in engineering. Generally, design is implemented in a "cakecutting" manner with an MSC and all the BSCs mounted to the MSC as a cluster. In this
way, the design work can be simplified, and the design process is lengthened so that the
design workload can be distributed properly based on the engineering schedule.
Principle of interworking
The BSS network design and core network design are closely related. Therefore, during
the BSS network design, designers must effectively communicate with core network
designers on issues, such as NE homing, interface interworking, and device capacity.
Principle of security in network design

The purpose of network design is to provide the customer with an available and reliable
network that can handle burst traffic and can recover quickly in the event of network
faults.
Principle of proper utilization of resources

The design principle of resources on a network varies with the development stage of the
network. For example, if the number of users on a network rapidly increases, the
resource usage cannot be designed too high. Otherwise, after the network is constructed,
new BSCs may be required, and then the new BTSs result in hybrid networking and
require re-homing, or new TRXs cannot be added for capacity expansion due to capacity
limitation of the BSC after resources are used up.
Principle of interface independence

The A, Gb, and Abis interfaces are physically independent. That is, a physical board can
be configured with only one type of logical interfaces. Do not configure the A, Abis, and
Gb interfaces on the same interface board because of the inconvenience for follow-up
maintenance and expansion and the greater impact from board faults.
2015-11-13
Huawei Confidential
Page 22 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Overview of Key NEs
BSC
The BSC connects to the MSC and BTS through the A interface and Abis interface
respectively. A PCU is embedded to implement radio resource management, BTS
management, power control, handover control, radio network configuration, and radio
network performance measurement.
BTS
The BTS/eGBTS connects to the BSC through the Abis interface and communicates with
mobile stations (MSs) through the radio interface. The BTS/eGBTS provides radio
functions in the BSS. For example, the BTS transmits and receives radio signals,
measures the quality of the radio network, controls power, and implements channel
coding, interleaving, and encrypting for radio channels.
PCU
The built-in PCU connects to the SGSN through the Gb interface. The PCU is introduced
in the BSS so that the BSS supports the general packet radio service (GPRS) packet
service. The PCU manages packet radio resources, controls packet calls, and transmits
data packets on the radio interface and Gb interface
MSC server
The MSC server provides switching functions and implements call switching between
the public land mobile network (PLMN) and the public switched telephony network
(PSTN). The MSC server provides telecom services, bearer services, and supplementary
services for mobile subscribers.
SGSN
The SGSN is a core network device in the GSM packet switched (PS) domain. It
implements functions, such as mobility management, session management, data packet
routing and forwarding, charging, SMS, customized applications for mobile network
enhanced logic (CAMEL), and quality of service (QoS) management.
2015-11-13
Huawei Confidential
Page 23 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Overview of the Network Design

Tool
The NEP tool is developed to improve the efficiency of network design delivery. This tool can
complete most network designs automatically. If you use the NEP tool in network design, the
efficiency can be greatly improved. For detailed information, contact Li Yongqing (employee
ID: 00141602), network design delivery representative of Network Integration Service (NIS).
2015-11-13
Huawei Confidential
Page 24 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Important Reference Document

The Transmission Configuration Specifications describes the specifications of IP planning,
QoS parameters, and VLAN planning used for IP transmission. Related links are available at
http://support.huawei.com:
A&GB Interface Configuration Specification_IP(GBSS17.0)

Wireless > Wireless Public > Wireless Professional Services Product > Technical
Guides
http://support.huawei.com/support/pages/navigation/gotoKBNavi.do?
actionFlag=getAllJsonData&colID=ROOTWEB|CO0000000064&level=4&itemId=20300051453&itemId0=29-7&itemId1=3-154&itemId2=1-632&itemId3=20200051452&itemId4=20300051453&itemId5=&itemId6=&itemId7=&itemId8=&itemId9=&materialType=1232&isHedexDocType=&pageSize=20
2015-11-13
Huawei Confidential
Page 25 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Product Specifications
The specifications vary with the product version. For details about the capacity specifications of the
BSC of a certain version, see the officially released documents of that version.
Specifications and capacity configuration of the BSC must be based on a certain traffic model, all
contracts must be established on a given traffic model to ensure the accuracy of the contract. If you are
unable to obtain accurate traffic.
7.1 BSC Specifications

For details about BSC specifications, see BSC6910 GU Product Description in the Hedex
BSC6910 GU product documentation.
Specifications of a BSC adopting an all-IP network change as follows: The number of TRXs
increases from 8192 in the BSC6900 to 24000 in the BSC6910.
7.1.1 Hardware Capacity

Table 1.1 lists the typical maximum configuration of R16 boards in BSC6910 GSM. The
GBSS17.0 BSC6910 has a maximum configuration of one cabinet and three subracks in the
GO mode and supports A over TDM, but it cannot be configured in BM/TC separated mode.
The GBSS15.0 does not support A over TDM.
Table 1.1 Typical maximum configuration of R16 boards in BSC6910 GSM (Abis over TDM and
A over IP are adopted.)
2015-11-13
Specification and Subrack Name
1 MPS+2 EPS
Number of cabinets
Maximum BHCA (M)
15
Traffic volume (Erlang)
43750
Number of TRXs
7000
Number of PDCHs that can be activated (MCS-9)
28000
Huawei Confidential
Page 26 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Gb throughput (G)
2.688
Table 1.2 lists the typical maximum configuration of traffic volume of R16 boards in
BSC6900 GSM where the BM and TC are separated.
Table 1.2 Typical maximum configuration traffic volume of R16 boards in BSC6900 GSM (all-IP
mode used)
Specification and Subrack Name
1 MPS+2 EPS (Number of

Subracks Can Be Changed)
Number of cabinets
Maximum BHCA (M)
52
Traffic volume (Erlang)
150000
Number of TRXs
24000
Number of PDCHs that can be activated (MCS-9)
96000
Gb throughput (G)
7.1.2 Estimation of BSC Configuration Capacity

BSC configuration capacity is estimated based on two key BSC counters: Busy Hour Call
Attempts (BHCA) and traffic volume. The actual configuration capacity is related to the
number of interface boards and the number of service processing boards and is the minimum
capacity calculated based on each board. The estimation of the configuration capacity
conducted currently, however, is based on the number of EGPUa(GCUP) boards. This section
describes a simple method for estimating BSC configuration capacity based on the number of
EGPUa(GCUP) boards.
The following describes how to calculate the maximum number of BHCA and maximum
traffic volume:
2015-11-13
The method for calculating the maximum number of BHCA allowed by the current
configuration is as follows:
If all interfaces adopt the IP transmission mode, the maximum number of BHCA
allowed by the current configuration is calculated as follows: Maximum number of
BHCA = MIN ((Number of EGPUa(GCUP) pairs on the current BSC x Number of
BHCA supported by a pair of EGPUa(GCUP)s x 80%, 52,000,000)
If all Abis interfaces adopt the TDM transmission mode, the maximum number of
BHCA allowed by the current configuration is calculated as follows: Maximum
number of BHCA = MIN ((Number of EGPUa(GCUP) pairs on the current BSC x
Number of BHCA supported by a pair of EGPUa(GCUP)s x 80%, 21,000,000)
If the Abis interfaces adopt the TDM/IP hybrid transmission mode, the maximum
number of BHCA allowed by the current configuration is calculated as follows:
Maximum number of BHCA = MIN ((Number of EGPUa(GCUP) pairs on the
current BSC x Number of BHCA supported by a pair of EGPUa(GCUP)s x 80%,
52,000,000)
Huawei Confidential
Page 27 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The simplified method for calculating traffic (expressed in Erlang) is as follows:
If all interfaces adopt the IP transmission mode, the maximum traffic volume allowed
by the current configuration is calculated as follows: Maximum traffic volume = MIN
((Number of EGPUa(GCUP) pairs on the current BSC x Traffic volume supported by
a pair of EGPUa(GCUP)s, 150,000)
If all Abis interfaces adopt the TDM transmission mode, the maximum traffic volume
allowed by the current configuration is calculated as follows: Maximum traffic
volume = MIN ((Number of EGPUa(GCUP) pairs on the current BSC x Traffic
volume supported by a pair of EGPUa(GCUP)s, 62,500)
If the Abis interfaces adopt the TDM/IP hybrid transmission mode, the maximum
traffic volume allowed by the current configuration is calculated as follows:
Maximum traffic volume = MIN ((Number of EGPUa(GCUP) pairs on the current
BSC x Traffic volume supported by a pair of EGPUa(GCUP)s, 150,000)
7.2 Board Specifications

7.2.1 BSC6910 Board Specifications
For details about board specifications, see Boards in BSC6910 GSM Hardware Description in
the BSC6910 documentation package. This document is officially issued to customers.
Hardw
are
Versio
n
Involved Board
HW6910
R15
SCUbGCGaGCUaGCUbGCGbFG2cGOUcEGPUaEXOUaEOMUa
ESAUaENIUaEXPUa
HW6910
R16
SCUbGCUbGCGbFG2cGOUcEGPUaEXOUaEOMUaESAUaENI
UaGOUeEXPUa
HW6910
R17
SCUbFG2cEGPUaEXOUaEOMUaESAUaENIUa,
SPUc GCGb GCUbGOUeEXPUa
In the BSC6910, only the POUc boards support Abis over TDM and A over TDM. The POUc
supports 1024 TRXs (without extra license control). In the BSC6900, the POUc supports 512
TRXs and can be used in the BSC6910. In the BSC6910, POUc boards support TDM and IP
over E1 transmission.
In A over TDM transmission mode, DPUf boards must be configured to process user-plane
CS data. The number of configured DPUf boards is determined according to the number of
CICs. The DPUf supports N+1 backup mode.
Number of Configured DPUf = RoundUp(MaxACICPerBSCTDM/ TCNoPerDPUf,0)
where
MaxACICPerBSCTDM indicates the maximum number of required A CICs on a BSC and is
calculated based on the traffic model.
2015-11-13
Huawei Confidential
Page 28 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
7.2.2 Service Processing Modules

Table 1.1 lists the specifications of service processing boards.
Table 1.1 Specifications of service processing boards
Boar
d
Logical
Functi
on
Full Name
of Logical
Function
Descriptio
n
Specificati
ons
Condition
EGPUa
RMP
Resource
management
Resource
management
processing
This board is
for resource
management of
the system.
A BSC is
configured
with a pair of
EGPUa
boards.
This board
(GCUP)
processes
services of
control plane
and user plane
integration. In
addition, it
supports CS
and PS
services of the
standard TRX.
This board
processes
services of
control plane
and user plane
integration. The
specification is:
1000 TRXs
The BHCA is
based on
Huawei default
traffic model.
processing
GCUP
2015-11-13
GSM BSC
control plane
and user plane
processing
600 BTSs
600 CELLs
3000 PDCHs.
GMCP
GSM BSC
mathematics
calculation
processing
If the board is
used for GSM
BSC
mathematics
calculation
processing, it
can calculate
using the
Interference
Based
Channel
Allocation
(IBCA)
algorithm.
None
The GMCP
needs to be
configured if
the IBCA
feature is
enabled.
NASP
Network
assisted
service
process
Network
assisted
service
processing
unit
None
The NASP
needs to be
configured if
Intelligent WiFi Detection
and Selection
is enabled.
Huawei Confidential
Page 29 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
EXPUa
RMP
Resource
management
processing
Resource
management
processing
board.
This board is
for resource
management of
the system.
A BSC is
configured
with a pair of
EXPUa
boards.
GCUP
GSM BSC
control plane
and user plane
processing
This board
(GCUP)
processes
services of
control plane
and user plane
integration. In
addition, it
supports CS
and PS
services of the
standard TRX.
This board
processes
services of
control plane
and user plane
integration. The
specification is:
The BHCA is
based on
Huawei default
traffic model.
If the board is
used for GSM
BSC
mathematics
calculation
processing, it
can calculate
using the
IBCA
algorithm
None
The GMCP
needs to be
configured if
the IBCA
feature is
enabled.
GMCP
GSM BSC
mathematics
calculation
processing
1000 TRXs
600 BTSs
600 CELLs
3000 PDCHs.
ENIUa
NIU
Evolved
network
intelligence
unit
Evolved
network
intelligence
unit
An ENIUa
board has a
capacity of
8000 M PS
throughput in
the RAN15.0.
The ENIUa
needs to be
configured if
the Evolved
Deep Packet
Inspection
function is
enabled.
ESAUa
SAU
Evolved
service aware
unit
Evolved
service aware
unit
The SAU board

is for
collecting,
filtering, and
gathering data
of the service
board, and
periodically
sending it to
Nastar.
The SAU
needs to be
configured on
the BSC, if a
user purchases
Nastar. A BSC
is configured
with only one
ESAUa board.
2015-11-13
Huawei Confidential
Page 30 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
DPUf
DPU
CS Data
Processing
Unit (1920
CICs)
This board
provides the
TC function to
process CS
data and
works in N+1
backup mode.
This board
provides the TC
function
(supporting
1920 CICs) in
A over TDM
mode.
If common
AMR is used,
the DPUf
supports 1920
CICs. If WB
AMR is used,
the number of
supported
CICs is halved.
That is, the
board
capability
required by
WB AMR calls
is two times
greater than
that required
by common
calls.
7.2.3 Interface Modules

Table 1.1 lists the interfaces applicable to the boards.
Table 1.1 Interfaces applicable to the boards
Board Name
Description
Applicable Interface
FG2c
IP Interface Unit (12 FE/4

GE, Electric)
IP: A/Abis/Lb/Gb/Iur-g
GOUc
IP Interface Unit(4 GE,

Optical)
EXOUa
Evolved 10GE Optical

interface Unit
POUc
TDM Interface Unit(4 STM1, channelized)
TDM: Abis
Table 1.2 lists the specifications of interface boards over different interfaces.
Table 1.2 Specifications of interface boards over different interfaces
2015-11-13
Huawei Confidential
Page 31 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Model
Trans
missi
on
Mode
Port Type
Port
No.
TRX
A CIC
(64
kbit/s)
Ater CIC
(16
kbit/s)
Gb
Through
put(Mbit
/s)
WP1D000FG201
(FG2c)
IP
FE/GE
electrical
port
12/4
2048
23040
N/A
2000
WP1D000GOU01
(GOUc)
IP
GE optical
port
2048
23040
N/A
2000
QM1D00EXOU00
(EXOUa)
IP
10 GE
optical port
8000
75000
N/A
8000
WP1D000GOU03
(GOUe)
IP
GE optical
port
2048
23040
N/A
2000
WP1D000POU01
(POUc)
TDM
CSTM-1 port
1024
7680
N/A
488
IP
IP CSTM-1
2048
N/A
N/A
N/A
The total number of required interface boards is the sum of interface boards over all
interfaces. Interface boards work in 1+1 backup mode. The BSC does not support BM/TC
separated mode and is not configured with the Ater interface. The A, Gb, and Abis interfaces
must be configured on the BM subrack side. On a GSM network, it is not recommended that
the A, Abis, and Gb share an interface board. Interface boards are configured over different
interfaces.
Calculation of the number of Abis interface boards
Select appropriate transmission ports based on the network plan. Calculate the number of
required Abis interface boards based on the service capability (TRX support capability) and
port requirements, and then select the maximum value.
Number of Abis interface boards = 2 x RoundUp(MAX(Number of TRXs in the transmission
mode/Number of TRXs supported by the interface board, number of ports in the transmission
mode/number of ports supported by the interface board),0)
When configuring Abis interface boards, concern the following aspects:
2.
In Abis over TDM transmission mode, the BSC6910 only supports the POUc and does
not support the TDM over E1/T1 interface board. If the Abis uses TDM over E1/T1
transmission on the BSC side, optical or electrical switching devices, such as Huawei
OSN device, are required to perform switching between E1/T1 and STM-1.
3.
The BSC6910 cannot be configured with a 10GE EXOUa interface board. Instead, it can
only be configured with the FG2 or GOUc working as the GE interface board when both
of the following conditions are met:
The BTS uses IP over E1 transmission.
The BSC uses IP transmission.
4.
Only the POUc board of the BSC6910 supports IP over E1 transmission.
Calculation of the number of A interface boards

required A interface boards based on the service capability (CIC support capability).
2015-11-13
Huawei Confidential
Page 32 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Number of A interface boards = 2 x RoundUp(A CIC Number/Support capability of the

A interface board,0)
When configuring A interface boards, concern the following aspects:
In A over TDM transmission mode, the BSC6910 only supports the POUc interface board
(TDM over STM-1) and does not support the TDM over E1/T1 interface board. If the Abis
uses incoming TDM over E1/T1 transmission, optical or electrical switching devices, such as
Huawei OSN device, are required to perform switching between E1/T1 and STM-1.
Calculation of the number of Gb interface boards

required Gb interface boards based on the service capability (bandwidth support
capability).
Number of Gb interface boards = 2 x RoundUp(Gb throughput/Support capability of the
Gb interface board,0)
Calculation of the total number of required interface boards

The total number of required interface boards is calculated as follows:
Total number of interface boards = Number of Abis interface boards + Number of A
interface boards + Number of Gb interface boards
2015-11-13
Huawei Confidential
Page 33 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
BOQ Review Guide
8.1 Design Overview

8.1.1 Purpose of the Design
Review the pre-sales bill of quantities (BOQ) configuration based on accurate network
planning information to ensure that the BOQ meets network construction requirements.
8.1.2 Input of the Design
Device BOQ
Network planning information (obtain the information, including the BSC coverage,
traffic, location area code (LAC) partitioning, and BTS homing from the on-site network
planning department.)
Information about the equipment room, power supply, or transmission of the customer,
and special requirements of the customer
8.1.3 Contents of the Design
BOQ review result
Defects in the BOQ and solution (suggestions)
8.1.4 Design Reference

GSM BSC Configuration Manual
8.2 Overview of Pre-sales Network Design

This section describes the pre-sales network planning and design as well as BOQ principles
and process to guide network design personnel through BOQ review.
In the network deployment scenario, the pre-sales network design procedures are as follows:
1.
2015-11-13
The pre-sales network planner plans the number of TRXs and the number of BTSs in the
areas based on the capacity, coverage, and information, such as coverage, predicted
number of subscribers, and traffic per subscriber, provided by the customer.
Huawei Confidential
Page 34 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
2.
The pre-sales network designer confirms the BSC locations and network topology based
on the transmission conditions, equipment room resources, and core network of the
customer.
3.
The pre-sales network designer uses the design tool to calculate the subrack and board
BOQ configuration of each BSC based on the number of TRXs of each BSC, number of
BTSs, half-rate ratio, transmission type, and number of BSCs.
In the network expansion scenario, the pre-sales network design procedures are as follows:
1.
The pre-sales network planner plans the number of sites and carriers to be added based
on the congestion rate, coverage, and frequency planning of the live network.
2.
The pre-sales network designer calculates the number of pieces of BSC hardware
required based on the number of sites, number of TRXs, half-rate ratio, and transmission
type, deducts the number of pieces of existing hardware from the number of pieces of
BSC hardware required to obtain the number of pieces of hardware to be added, and then
generates the BSC device BOQ.
The pre-sales planning of the core network is different from that of the BSS. The BOQ and interface
bandwidth data of the core network are obtained directly based on the number of subscribers, traffic per
subscriber, and certain redundancy. Therefore, interface bandwidth inconsistency may occur. Generally,
the calculation result of the core network is smaller, and this causes the interface bandwidth
inconsistency. The BSS planning does not involve bandwidth bottleneck and facilitates follow-up
network development.
8.3 BOQ Review Principles

Use the GSM NEP tool for BOQ review.
In BOQ review, the number of pre-sales configured boards, especially the number of Abis
interface boards, is reviewed. If spare BOQ hardware is configured, the review is successful.
If the BOQ hardware is insufficient, check with marketing personnel whether to change the
delivery.
In the new network construction and migration scenarios, use the GSM NEP tool to calculate
the required BSC hardware based on the number of BTSs, number of TRXs, and traffic
model, and check the requirements against the pre-sales BOQ.
In the expansion scenario, use the GSM NEP tool to calculate the required BSC hardware
based on the number of BTSs, number of TRXs, and traffic model after expansion, deduct the
existing BSC hardware to obtain the number of pieces of hardware to be added, and check the
number against the pre-sales expansion BOQ.
To meet the special requirements of some operators, the actual number of pieces of hardware
in BOQ delivery is far greater than the actual required number of pieces of hardware. In terms
of BTS distribution, the following principles are recommended (you communicate with the
operator to learn the follow-up expansion plan):
BTSs are evenly distributed to subracks and Abis interface boards based on a certain
redundancy ratio. This facilitates follow-up TRX expansion or BTS addition.
Traffic model parameters
The following content is quoted from the GBSS15.0 BSC6910 system capacity calculation
manual. Ensure that the following content is for internal use only.
2015-11-13
Huawei Confidential
Page 35 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
In this document, parameters are described in tables. Different colors in tables convey
different meanings as follows:
Parameter
Title
TRX
Number
Parameter name
0.02
It is an input parameter, which is entered based on the network planning

and design result. You can use the default value (if available) of an input
parameter if the entered value cannot be obtained. The calculation result
based on the default value is different from the actual situation. Generally,
the result calculated based on the default value is larger. That is, more
device resources are required.
300
It is an advanced parameter. You can enter a value or directly use the

default value.
98%
Automatically calculated result. Do not change this value unless you are
absolutely confident of the new value. If you can provide the dimension
result, you can use it, but you must ensure that the modification is correct.
8.4 CS Traffic Models

The circuit switched (CS) traffic model affects the BSC system capacity in the following
aspects:
CS traffic on the control plane in the system. It is measured by the BHCA. If the traffic
on the air interface in the system is specified, the traffic model affects the BHCA traffic
on the control plane in the system.
CS traffic on the user plane in the system. It is measured in Erlang. If the number of
subscribers on the network is specified, the traffic model affects the traffic on the user
plane in the system.
Table 1.1 CS traffic model parameters

Parameter
2015-11-13
Name
Defau
lt
Value
Description
Average voice traffic

CSErlPerSub
per
subscriber@BH(Erlang)
0.02
Average busy-hour CS
traffic per subscriber
Average Call
Duration(Second)
CSCallDuration
60
Average busy-hour
conversation duration per
subscriber
Percent of Mobile
originated calls
CSMOCRatio
50%
Average busy-hour MOC

ratio
Percent of Mobile
terminated calls
CSMTCRatio
50%
Average busy-hour MTC

ratio
Huawei Confidential
Page 36 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Parameter
Name
Defau
lt
Value
Description
Average LUs/sub/BH
CSLUPerSubinBH
1.2
Number of busy-hour
location updates per
subscriber
Average IMSI
Attach/sub/BH
CSAttachPerSubinBH
0.15
Average busy-hour IMSI

attachments per subscriber
Average IMSI
Detach/sub/BH
CSDetachPerSubinBH
0.15
Average busy-hour IMSI

detachments per subscriber
Average MO-SMSs
/sub/BH
CSMOSMSPerSubinB
H
0.6
Average busy-hour sent

SMSs per subscriber
Average MT-SMSs
/sub/BH
CSMTSMSPerSubinB
H
Average busy-hour received

SMSs per subscriber
Average intra-BSC HOs CSIntraHOPerSubinB

/sub/BH
H
1.1
Average busy-hour intraBSC handovers per

subscriber
Average inter-BSC HOs

/sub/BH
CSInterHOPerSubinB
H
0.1
Average busy-hour interBSC handovers per

subscriber
Paging Retransfer Ratio
PagingRetransferRatio
35%
Ratio of paging retries on

the A interface in busy
hours
Table 1.2 CS signaling load parameters

Parameter
Name
Defau
lt
Value
Description
64k SS7 signaling links

load
64kSS7SigLinkLoad
0.2
Busy-hour 64K signaling

load
2M SS7 signaling links

load
2MSS7SigLinkLoad
0.2
Busy-hour 2M signaling
load
Table 1.3 GoS-related parameters
2015-11-13
Parameter
Name
Defau
lt
Value
Description
Grade of Service (GoS)

on Um interface
UmBlockRatio
0.02
Um interface block ratio
Huawei Confidential
Page 37 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Grade of Service (GoS)

on A interface
ABlockRatio
0.001
Device block ratio
Table 1.4 Other related parameters

Parameter
Name
Defau
lt
Value
Description
Average MOCs/sub/BH
CSMOCPerSubinBH
0.6
Number of busy-hour
calling times per subscriber
= CSErlPerSub x
3600/CSCallDuration x
CSMOCRatio
Average MTCs/sub/BH
CSMTCPerSubinBH
0.6
Number of busy-hour called

times per subscriber =
CSErlPerSub x
3600/CSCallDuration x
CSMTCRatio
MR report/sub/BH
CSMRPerSubinBH
144
Average number of MRs

reported by each subscriber
in busy hours. Its weight in
BHCA is zero. It is used
only for reference.
Paging retransfer
/sub/BH
CSRetransferPagingPe
rSubinBH
0.56
Average number of paging

retransmission times per
subscriber in busy hours on
the A interface.
Parameter relationship in the CS traffic model

2.
Relationship between CSMTCRatio and CSMOCRatio:

CSMTCRatio = 1 CSMOCRatio
3.
Relationship between CSErlPerSub, CSCallDuration, CSMOCPerSubinBH,

CSMOCPerSubinBH, and CSMTCPerSubinBH:
CSMOCPerSubinBH = (CSErlPerSub x 3600/CSCallDuration) x CSMOCRatio
CSMTCPerSubinBH = (CSErlPerSub x 3600/CSCallDuration) x CSMTCRatio
4.
Calculation of CSMRPerSubinBH:
CSMRPerSubinBH = (CSMTCPerSubinBH + CSMOCPerSubinBH) x CSCallDuration
x2
In the preceding formula, the MRs that are not reported in the call stage. For example, the MRs
reported in the short message service (SMS) and signaling connection stages, are not included.
5.
Relationship between CSRetransferPagingPerSubinBH and PagingRetransferRatio:

CSRetransferPagingPerSubinBH = (CSMTCPerSubinBH + CSMTSMSPerSubinBH) x
PagingRetransferRatio
2015-11-13
Huawei Confidential
Page 38 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
8.5 PS Traffic Models

The packet switched (PS) traffic model affects the BSC system capacity in the following
aspects:
If the number of subscribers in the system is specified, the PS traffic model determines the
total traffic of PS services in the BSC.
The PS traffic model consists of the following:
Basic PS traffic model (This model is new in the R13. In R12 and earlier versions, only
the PS user model is available.)
PS user model (For details about this model, see Table 1.2.)
Table 1.1 Basic PS traffic model

Parameter
Name
Value
Description
Uplink TBF Est & Rel

/ Second/TRX
TBFUpPerSec
PerTRX
1.75
It indicates the average number of

uplink TBFs per second for each
TRX in peak hours. Its default
value is 1.75 for common networks
and is 3.5 for PS networks with
heavy traffic.
Downlink TBD Est &

Rel / Second/TRX
TBFDownPerS
ecPerTRX
0.9
It indicates the average number of

downlink TBFs per second for each
TRX in peak hours. Its default
value is 0.9 for common networks
and is 1.8 for PS networks with
heavy traffic.
PS Paging / Sub/BH
PSPagingPerS
ub
1.25
It indicates the number of received

peak-hour pagings for each PS
subscriber. Its default value is 1.25
for common networks and is 2.5 for
PS networks with heavy traffic.
Table 1.2 PS user model
2015-11-13
Parameter
Name
Valu
e
Description
GPRS Active Sub
PSSubAct
10000
Number of online
GPRS/EGPRS subscribers
average traffic per sub

in busy hour (bit/s)
PSTrafficPerSubinBH
300
Average GPRS/EGPRS traffic

per online subscriber in busy
hours (application layer)
PS Traffic Peak Ratio
PSPeakRatio
25%
Ratio of the difference

between the PS peak traffic
and the average traffic to the
average traffic. Do not use this
parameter if it is not required.
Huawei Confidential
Page 39 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
average IP packet data

length in Gb (Bytes)
PayloadLenGb
300
Average packet length on the

Gb interface. Do not use this
parameter if it is not required.
Table 1.3 PS coding ratio and average rate

code scheme
Ratio
CS1
0%
CS2
0%
CS3
0%
CS4
0%
MCS1
0%
MCS2
0%
MCS3
0%
MCS4
0%
MCS5
0%
MCS6
100%
MCS7
0%
MCS8
0%
MCS9
0%
The sum of the preceding coding rate ratios must be 100%.

Generally, the customer cannot provide the ratios of coding rates during calculation. You can
use the customer-expected average rate to replace the inputs. For example, the customerexpected average rate is about 30 kbit/s. According to the preceding table, this average rate is
within the rate range of the MCS6 coding mode. In this case, you can simply enter 100% as
the ratio of the MCS6 coding mode.
8.6 Relationship Between Traffic Model and

Traffic Statistics
Traffic model indicates the average number of typical subscriber behaviors for a subscriber.
The total number of these subscriber behaviors can be obtained from the traffic statistics. The
traffic model for a subscriber equals the total number divided by the number of subscribers.
The number of subscribers (SubPerBSC) served by a BSC must be available and accurate in
the calculation of the traffic model.
Table 1.1 lists the performance counters corresponding to basic procedures.
2015-11-13
Huawei Confidential
Page 40 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Table 1.1 Performance counters corresponding to basic procedures
2015-11-13
Basic Procedure
(Subscriber
Operation)
Performance Counters (Sum of Cell

Performance)
CS LUs (Location
Update)
A300F: Channel Requests (Location Updating)
Average IMSI Attachs

(IMSI Attachs)
From MSC
Average IMSI Detachs

(IMSI Detachs)
From MSC
CS calls
A300A: Channel Requests (MOC) + A300C: Channel

Requests (MTC) CA334A: Total Uplink Point-to-Point
Short Messages CA334B: Total Downlink Point-to-Point
Short Messages
MR Reports
S329: Number of Power Control Messages per Cell
CS SMS (sending and

receiving)
CA334A: Total Uplink Point-to-Point Short Messages +

CA334B: Total Downlink Point-to-Point Short Messages
Intra-Hos (intra BSC)
CH310: Number of Outgoing Internal Inter-Cell Handover

Requests
Inter-HOs (Inter BSC)
CH330: Outgoing External Inter-Cell Handover Requests +

CH340: Incoming External Inter-Cell Handover Requests
CS Paging
A330: Delivered Paging Messages for CS Service
Uplink TBF Est
A9201: Number of Uplink EGPRS TBF Establishment

Attempts + A9001: Number of Uplink GPRS TBF
Establishment Attempts
Downlink TBF Est
A9301: Number of Downlink EGPRS TBF Establishment

Attempts + A9101: Number of Downlink GPRS TBF
Establishment Attempts
PS Paging
A331: Delivered Paging Messages for PS Service
Huawei Confidential
Page 41 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Parameters for Capacity

Calculation
Please refer to the latest BSC6900 Capacity Calculation Manual which you can download
from http://3ms.huawei.com.
http://3ms.huawei.com/mm/docMaintain/mmMaintain.do?
method=showMMDetail&f_id=GSM14040308540024
10
Capacity Calculation
Please refer to the latest BSC6900 Capacity Calculation Manual which you can download
from http://3ms.huawei.com.
http://3ms.huawei.com/mm/docMaintain/mmMaintain.do?
method=showMMDetail&f_id=GSM14040308540024
Reference: Impact on Interface Transmission Bandwidth After VLAN Is Deployed

VLAN is a data exchange technology derived from traditional LAN.
VLAN allows LAN devices to be logically grouped into multiple network segments (that is,
smaller LANs) to implement virtual workgroups. The hosts in the same VLAN communicate
2015-11-13
Huawei Confidential
Page 42 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
with each other through VLAN switches. The hosts in different VLANs are separated from
each other and they only communicate with each other through routers. A VLAN is a
broadcast domain, that is, a host in a VLAN can receive broadcast packets from the other
hosts in the same VLAN but cannot receive any broadcast packets from other VLANs.
The advantages of VLAN are as follows:
Suppresses broadcast storm
Improves transmission security
Provides differentiated services
VLAN Frame Format

The VLAN frame format is defined in IEEE 802.1Q. Compared with a standard Ethernet
frame, the VLAN frame is added with a four-byte VLAN tag in its header, as shown below.
The fields of the VLAN tag are described as follows:
TPID: specifies the VLAN tag protocol identifier defined by IEEE. If a VLAN frame
complies with IEEE 802.1Q, TPID is permanently set to 0x8100.
VLAN priority: specifies the priority of a VLAN frame. The priority ranges from 0 to 7.
Ethernet provides differentiated services based on the VLAN priority.
Canonical Format Indicator (CFI): specifies the format of a frame that is exchanged
between the bus Ethernet and a Fiber Distributed Data Interface (FDDI) or between the
bus Ethernet and the token ring network.
VLAN ID: specifies the VLAN to which a frame is to be sent. Each VLAN is identified
by a VLAN ID.
Application scenario: Only Ethernet IP networks.

The related BSC6910 parameters are as follows:
VLANID: This parameter specifies the identifier of a VLAN. The VLAN ID mapping
should be preconfigured in the BSC6910. According to the VLAN ID mapping, the
BSC6910 determines the VLAN ID to send a VLAN frame. The BSC6910 supports two
VLAN configuration modes:
Configuring VLAN by next hop:

The VLAN ID is determined according to the preconfigured mapping between the
next-hop IP address and the VLAN ID.
The related parameters are IPADDR and VLANID.
2015-11-13
Huawei Confidential
Page 43 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Configuring VLAN by data flow

The VLAN ID is determined according to the preconfigured mapping between the
SCTP link, IP path, and VLAN ID.
The parameters related to SCTP link are SCTPLNKN, PATHID, and VLANID.
VLANPRI: This parameter specifies the priority of a VLAN frame.
VLAN configuration modes supported by different interfaces on the BSC6910 on the GSM
networks are as follows:
The A and Abis interfaces support configuring VLAN by next hop or data flow.
The Gb interface supports configuring VLAN by next hop.
The Ater interface supports configuring VLAN by next hop or data flow when IP over
E1 is not in use.
Impact assessment: With the increasing deployment of IP networking, in particular, with the
increasing deployment of VLAN networking on IP networks, VLAN tags have certain impact
on IP transmission bandwidth over the Abis interface. The actual impact varies according to
different compresses and transmission rate, and the average impact is about 3.5%.
Detailed calculation method:
If IP multiplexing (MUX) is not in use, a four-byte (32-bit) VLAN tag is added to a 20-ms
voice (data) frame.
If MUX is in use, a four-byte (32-bit) VLAN tag is added to voice (data) frames that are
transmitted at an interval of 20 ms. Therefore, VLAN tag resources are saved if MUX is in
use.
2015-11-13
Huawei Confidential
Page 44 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
11
Design of Resource
Allocation
11.1 Design Overview

Review the traffic and BHCA load of each device based on accurate network planning
information. If a load risk exists or the traffic exceeds the specifications, adjust the BTS
homing. If the BTS homing cannot be adjusted, negotiate with the customer and
marketing personnel to purchase more devices (under the guidance of marketing
personnel).
Configure BSC boards in proper slots based on the BSC traffic and BHCA to balance the
BSC load, improve the device resource usage, and improve the anti-attack capability.
Review the specifications information about the MSC, MGW, and SGSN to check
whether the capacities are enough and assess the risk.
Device BOQ
Network planning information (Obtain the information, including the BSC coverage,
traffic, LAC partitioning, and BTS homing from the on-site network planning
department.)
Information about the equipment room, power supply, or transmission of the customer,
and special requirements of the customer
11.2 BSC Load Allocation

This section assesses the BSC load risks, including the current traffic model and target traffic
model based on the current device processing capability, and lists the percentages of the
BHCA and traffic load of each BSC in the design specifications.
11.2.1.1 Design Principles
2015-11-13
The predicted BSC traffic load does not exceed 70% of the design specifications.
The predicted BHCA load does not exceed 70% of the design specifications.
TRXs are allocated to subracks evenly to balance the load and reduce signaling transfer
between subracks.
The number of TRXs configured in each subrack

needs to be less than 70% (it is for flexible follow-up adjustment and expansion but is
Huawei Confidential
Page 45 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
not mandatory) of the specifications to facilitate follow-up site adjustment and

expansion.
If the preceding principles conflict with the marketing BOQ, follow the marketing BOQ, and improve
the overall system capacity by optimizing capacity resource allocation. If the traffic load or BHCA load
exceeds 60%, experts in the Huawei headquarters assess the risk.
Confirm the BSC traffic capacity and BHCA specifications in the current configuration based
on the configuration in the marketing BOQ.
Assess whether the BSC resource load meets the requirements based on the traffic model,
traffic capacity, and BTS homing.
The BSC traffic calculated by the GSM NEP is the traffic capacity of the BSC. It is obtained
based on the number of TRXs, number of BTSs, congestion rate, and erlang_B table. The
actual traffic can be obtained from the customer or network planner. The calculation formula
is as follows:
Actual BSC traffic = Predicted number of subscribers x Busy-hour traffic per subscriber
Actual BSC BHCA = Predicted number of subscribers x Busy-hour BHCA per

subscriber
Actual BSC traffic load = Actual BSC traffic/BSC traffic specifications
Actual BSC BHCA load = Actual BSC BHCA/BSC BHCA specifications
Huawei's recommended expansion standards are as follows:
The number of TRXs configured for the BSC reaches 70% of the capacity specifications.
The busy-hour traffic exceeds 70% of the specifications.
The busy-hour BHCA exceeds 70% of the specifications.
The busy-hour central processing unit (CPU) usage exceeds 70%.
The SS7 link load exceeds 40%.
The CIC traffic per line exceeds 0.7 ERL.
Output of the design

Table 1.1 BSC capacity planning table
BSC
Name
BTS
Numbe
r
Traffic
BHCA
Foreca
st
Foreca
st
TRX
Numbe
r
TRX
TRX
Traffic
BHCA
Capacit
y
Perce
nt
Perce
nt
Percen
t
For BHCA calculation note, see section Error: Reference source not found.
9.1.1 Signaling Storm

11.2.1.2 Concept of Signaling Storm
Signaling storms first rose on the 3G network. Under the impact of a large number of
signaling messages, major signaling processing channels become the bottleneck of the
2015-11-13
Huawei Confidential
Page 46 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
network, and CPU usage on the control plane of both the radio network controller (RNC) and
NodeB increases sharply. Moreover, a smart terminal attempts to access the network at an
increasingly short interval, once some signaling messages are discarded. Consequently, the
already heavy signaling traffic becomes even heavier, which causes a signaling storm. The
signaling storm threatens equipment security of both the RNC and NodeB, and seriously
decreases the processing capacity of the system. A typical phenomenon is that the serving
capacity of the system decreases. That is, high RRC and radio access bearer (RAB) rejection
rates occur when the data traffic is low.
11.2.1.3 Current State of Signaling Storm

As the penetration rate of smart terminals increases, the feature of high signaling traffic on
smart terminals stands out. Statistics shows that smart terminals have seen an increase in
signaling traffic, which is 15 times that in traditional terminals. Part of the UMTS network has
been affected by signaling storms. For example, some subscribers of China Unicom in Beijing
failed to access the network after the congestion rate increases in the core network (CN) of the
X office in Beijing in 2010. Some subscribers of StarHub in Singapore had the same problem
in 2011. Moreover, some subscribers of TELUS in Canada failed to access the network after
many signaling messages were discarded by the CN of the X office. Some subscriber of NTT
DoCoMo in Japan failed to access the network, and performance and reliability of the
network were affected for the same reason.
11.2.1.4 Causes of Signaling Storm

A signaling storm rises from too many signaling messages, which are caused by the following
factors:
Smart phone penetration rate increases year on year. After smart phones were introduced,
services, including low-traffic services, have been diversified. Smart terminals and mobile
network services are increasingly popular. To provide better user experience, smart terminals
periodically send heartbeat packages to the network server to synchronize the information at
the request of such applications as QQ and MSN Messenger. Heartbeat packages are small
data packages of hundreds or thousands of bytes sent every dozen seconds or tens of seconds.
The heartbeats of different applications and the system result in frequent PS calling.
According to the 3rd Generation Partnership Program (3GPP), a terminal in the connected
state can send a signaling connection release indication (SCRI) to the RNC in some scenarios.
An SCRI carries different cause values in different scenarios. For example, a smart terminal
sends an SCRI with the cause value being "UE Requested PS Data session end." to indicate
the end of a PS data session.
The greatest bottleneck of an MS lies in the battery. To save power, a smart terminal
automatically sends an SCRI to the RNC at the end of a data session to release the RRC
signaling connection and returns to the idle state. However, some applications on the terminal
need to periodically send heartbeat packages to the application server. As a result, a
connection to the RRC is re-established, and the UE returns to the connected state. After a
small-size heartbeat package is sent, the RRC connection is released again, and the cycle goes
on and on. As many as 30 pieces of signaling over the Uu interface and Iub interface are
required in every PS data transmission, which makes the traffic model change significantly
and the packet calling attempts exceed the voice calling attempts.
11.2.1.5 Impact of Smart Phones on the 2G Networks

Signaling storms in the UMTS network rise from too many signaling messages caused by a
large number of smart phones and low-traffic services. Whether signaling storms are likely to
2015-11-13
Huawei Confidential
Page 47 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
raise in the 2G networks as the penetration rate of smart phones rapidly increases? An analysis
is made in multiple dimensions.
Impact of the heartbeat service model of smart phones over the average service model in
the GSM
Use the live network in Hangzhou city of China Mobile Group Zhejiang Company Ltd as
an example. As shown in Figure 1.1, the heartbeat service model of smart phones is the
same as the average service model in the GSM.
Figure 1.1 Average service duration
Impact of access to the GSM over the core network
When uplink data exists on the MS, the MS is switched to the Ready state and
directly sends the data. No authorization or encryption is required.
The state changes from Ready to Standby, if the time when no signaling message
exists over the Gb interface exceeds a time prescribed by a timer on the SGSN side.
When downlink data exists on the network side, the Gb interface sends a paging
message. In response, the MS returns a correct logical link control (LLC) frame.
Frequent service triggering does not increase signaling messages other than paging
messages over the Gb interface.
To sum up, signaling interworking does not exist between the GSM service access and the
core network.
2015-11-13
Huawei Confidential
Page 48 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Signaling in each access to the GSM

When the GSM network is accessed, the number of signaling messages is less than three
on the wireless network side, much less than the 25 signaling messages in the UMTS
network (not considering interworking with the core network).
2015-11-13
Signaling load on the common control channel (CCCH)
Based on the loading capacity of the base station subsystem (BSS), the loading
efficiency of the packet data channel (PDCH) is 9 kbit/s when a cell uses a maximum
of 64 PDCHs. Two CCCHs can meet the requirements of the PS signaling load when
the cell enables the multi-CCCH function.
The CCCH resource usage efficiency can be further improved by multi-layered

paging.
Channel management by layer improves signaling resource usage efficiency, gives

priority to access of voice services, and prevents the heavy PS load from affecting the
CS services.
The CCCH resources can bear the signaling load and do not form a bottleneck.
Huawei Confidential
Page 49 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Signaling load on the LAPD resources

The load on the LAPD link mainly comes from the load on the B RSL link, which affects
the CS paging messages, CS immediate assignment messages, PS paging messages, and
PS immediate assignment messages. The LAPD resources can bear the signaling load
and do not form a bottleneck.
The number of LAPD links required by the PS service is calculated according to the
following specifications: On a 16 kbit/s timeslot, the maximum signaling load has
2000 Bytes/s.
Number of bytes in paging messages: 21 Bytes
Number of bytes in PS immediate assignment messages: 27 Bytes
The eXtensible processing unit (XPU) resources

The following figure describes the subsequent networking planning requirements of
China Mobile Group. Huawei's XPU design specifications can meet the BHCA
requirements and do not form a bottleneck. In all-IP mode, the latest product BSC6910
has a BHCA specification of 52,000 K, which is much higher than the specification of
the BSC6900.
2015-11-13
Huawei Confidential
Page 50 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Flow control on the XPU of the BSC ensures that the PS paging times in a period and
the channel requests by the PS services in a period are controllable.
The increase in signaling services caused by the increase in the traffic of the current
data services does not have an impact over the XPU of the BSC. In event of sudden
increase in signaling messages, the flow control on the XPU guarantees the loading
security on the XPU and deals with the impact of the PS services over the CS
services.
The BSC BHCA specification of the BSC6910 is 52000 K (all-IP networking mode).
The data processing unit (DPU) resources

As the version is updated, the DPU supports an increasing large number of PDCHs. The
DPU resources can bear the signaling load and do not form a bottleneck.
The BSC6910 does not use the XPU and DPU boards separately. Functions of both XPU
and DPU boards are integrated in EGPUa boards. An EGPUa (GCUP) board supports
1000 TRXs and 3000 PDCHs, almost twice the number of TRXs (512) and PDCHs
(1024) supported by the original XPU and DPU boards. Based on the above analysis, the
BSC6910 does not have a bottleneck in processing capability of the XPU and DPU
boards.
2015-11-13
Impact of smart phones over the 2G network (based on the data of China Mobile Group
Zhejiang Company Ltd)
The 2G network subscriber base remains unchanged. However, traffic of smart phone
users is 2.9 times that of non-smart phone users.
Market penetration rate of small phones reached 19% in 2011. The total traffic
increases to 2.35 times that of the current traffic, if non-smart phones are substituted
by smart phones.
The average loading efficiency of the PDCH is 4 kbit/s on the live network. When the
loading efficiency of the PDCH increases to 9 kbit/s and the specification of
equipment is improved to support more channels, the traffic in the 2G network can
increase to 2.35 times the current traffic.
Huawei Confidential
Page 51 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Heartbeat duration of application services is adjusted to reduce the impact over the
network
Tencent increases the heartbeat duration of its program QQ (30 to 180s) to reduce the
impact over the network.
Apple launched APNS, a server to send notification messages to such terminals as

iPhone in a secure and timely manner, to manage heartbeats and increase heartbeat
duration.
Development trend
View of the operator (VF):
The GSM network will evolve to be a low-cost and low-traffic network for the following
reasons:
The spectrum resources of the GSM will decrease because part of the resources is
given to the UMTS and LTE in spectrum refarming. As a result, configuration for the
BTS degrades in the GSM.
The legacy UE evolves towards smart phones and the traffic becomes increasingly
low.
11.2.1.6 Conclusion of Impact of Smart Phones over the 2G

Network
Based on the analysis above, signaling storms do not occur on the 2G network. The
conclusion may be updated depending on the future development.
11.3 BSC Board Layout Design

11.3.1 Design Guide
Design board layout between subracks for the BSC based on the BOQ for load balancing and
work out the board configuration figure.
Purpose of board layout design:
2015-11-13
Decrease the number of messages forwarded between subracks to improve the BSC
performance.
Balance the load between subracks to improve the anti-attack capability.
Huawei Confidential
Page 52 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Reserve certain port redundancy to facilitate site adjustment and expansion.
Deploy logical boards of the same type in a centralized manner to reduce interleaving
with boards of different types.
Deploy electrical interface boards on one side and optical interface boards on the other
side to facilitate cable connection.
Use different boards to provide 2G and 3G services to reduce the impact of software
upgrade and board adjustment on services.
Allocate slots properly to maximize the board processing capability (the switching
capability of the slots on the backplane differs).
In an office where traffic is heavy, board layout is important. Proper board resource allocation
can maximize the processing capability of the device, balance the load, and improve the antiattack capability of the device.
11.3.1.1 BSC6910 Design Principles

Design Principles (Use the NEP Tool to automatically generate the board layout figure, and
the following principles are only for your reference)
Reduce inter-subrack signaling transfer. Ensure that the processing capabilities of the
Abis interface board, A interface board, and embedded packet control unit (PCU) in the
same subrack match each other.
Balance the load between the subracks of the BSC. The GMPS needs to process data,
such as operation and maintenance (O&M), traffic measurement, and alarms. The XPU
load is relatively high. The number of TRXs configured in the GMPS subrack is
relatively small. Therefore, in the case of Abis interface board imbalance between BM
subracks, the number of Abis interface boards configured for the GMPS is small.
Install interface boards in rear slots and service processing boards in non-fixed slots.
Therefore, preferentially install service processing boards in front slots. Deploy the A
interface board, Abis interface board, and Gb interface board separately, and deploy
logical interface boards of the same type (A interface board, Abis interface board, and Gb
interface board) together.
Deploy boards of the same type (physical boards or logical boards) from the middle to
sides in the subrack to facilitate follow-up board expansion.
Deploy optical interface boards and electrical interface boards on different sides in the
subrack. Do not deploy them on the same side.
The ENIU board (data service identification board with a specification of 1000 Mbit/s
over the Gb interface) can be inserted in slots that do not hold the OMU and GGCU of
the BM subrack. The recommended slots are slots 2 and 3, and the priorities of slots are
2 to 7. ENIU boards are preferably configured with the same subracks of the Gb
interface board (to reduce traffic between subracks). The system can be configured with
a maximum of 15 ENIUa boards. The ENIUa board can only be configured in 10 G slot.
When a customer purchases and uses Huawei's Nastar, the ESAUa boards need to be
inserted in the BSC6910. The ESAUa board may be inserted in other idle slots other than
the fixed slots. An ESAUa board occupies two slots. Configure the ESAUa board in the
active subrack.
Deployment of 10G Slots in the BSC6910
2015-11-13
Huawei Confidential
Page 53 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The following describes a method for configuring a main subrack:

Every BSC6910 must be configured with only one PCS main subrack.
Assign EOMUa switch boards to slot 10 to 13. SCUb boards are assigned to slot 20 and
21, and EGPUa boards for resource management are assigned to slot 8 and 9.
Configure the BSC6910 with two PCS GCUa boards, when a GPS clock is required.
Configure the BSC6910 with two PCS GCGa boards, when a GPS clock is not required.
Assign GCUa/GCGa boards to slot 14 and 15.
When a customer purchases Huawei's Nastar, ESAUa boards are required in the
BSC6910.
EGPUa/ESAUa boards can be inserted in other idle slots other than the fixed slots. The
following assignment is recommended:
2015-11-13
Assign ESAUa boards to slot 0 and 1.
Preferred slots for EGPUa boards are slot 2 to 7.
The following assignment is recommended for GOUc/FG2c/EXOUa/POUc boards:
EXOUa boards can only be assigned to slot 16 to 19 and slot 22 to 25.
Preferred slots for GOUc/FG2c/POUc boards are slot 16 to 19 and slot 22 to 25.
When these slots are inadequate, they are assigned to slot 26 to 27.
Huawei Confidential
Page 54 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The following describes a method for configuring an extended processing sub rack:
SCUb switch boards are assigned to slot 20 and 21.
If a customer purchases Huawei's Nastar, ESAUa boards are required. Configure ESAUa
boards in the main subrack.
EGPUa boards can be assigned to other idle slots other than slot 20 and 21. The
recommended slots are slot 0 to 13.
GOUc/FG2c/EXOUa/POUc boards are interface boards.
EXOUa/POUc boards can only be assigned to slot 16 to 19 and slot 22 to 25.
Preferred slots for GOUc/FG2c boards are slot 16 to 19 and slot 22 to 25. When these
slots are inadequate, assign GOUc/FG2c boards to slot 26 and 27.
Principles of EGPUa/EXPUa configuration
2015-11-13
Huawei Confidential
Page 55 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Service processing boards used by the BSC6910 include EGPUa and EXPUa boards. EXPUa
boards are used in the GSM other than the Universal Mobile Telecommunications System
(UMTS). Logical types of service processing boards are RMP, GCUP, GMCP, or NASP.
EXPUa and EGPUa boards can be configured in both GO and GU mode. By default,
EXPUa boards are configured in GO mode and EXPUa in GU mode.
In the UO mode, only EGPUa boards can be configured, instead of EXPUa boards.
Principles of EGPUa/EXPUa configuration for the RMP: In the GO mode, the RMP can
use EXPUa or EGPUa boards. By default, the RMP uses the same board as the GCUP. In
the GU/UO mode, the RMP can only use EGPUa boards.
Principles of EGPUa/EXPUa configuration for GMCP: In the GO/GU mode, the GMCP
can use XPUa or EGPUa boards. By default, the GMCP uses the same board as the
GCUP.
Principles of EGPUa/EXPUa configuration for the NASP: The NASP can only use
EGPUa boards, instead of EXPUa boars.
Principles of RMP configuration

The system is configured with only one RMP pair in the MSP subrack, one active and one
standby board.
Principles of GCUP configuration

Service processing boards are configured according to the BSC capacity planning. Different
calculation methods are applied for the BSC6910 and BSC6900.
In the BSC6900, numbers of XPUa/XPUb boards on the control plane, DPUd/DPUg boards
on the PS user plane, and DPUc/DPUf boards on the CS user plane are calculated differently.
For boards on the control plane, the number is the larger value calculated based on the
planned TRX number and the comprehensive BHCA. For boards on the PS user plane, the
number is calculated based on the number of PDCHs. For boards on the CS user plane, the
number is calculated based on the predicated traffic volume.
In the BSC6910, the EGPUa boards are used. Each GCUP board has the following
specifications: BTS number, cell number, TRX number, comprehensive BHCA, number of
PDCHs, and traffic volume. Divide the site-planned overall specification by the basic
specifications above respectively, to obtain several board numbers. The greatest of these
numbers is the number of boards to be configured.
Table 1.1 describes the specifications of the EGPUa boards.
2015-11-13
Huawei Confidential
Page 56 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Table 1.1 Specifications of the EGPUa board

TRX
1000
Cell
600
BTS
600
Traffic
volume
6250
6.25 Erlang per TRX on average
PDCH
3000
3 PDCHs per TRX on average
PS
throug
hput
300 Mbit/s
3000 100 kbit/s, EGRPS2A
Compre
hensive
BHCA
2200 K
The value is based on the actual benchmark

weights and considers the PS BHCA. The
PS BHCA is based on the comprehensive
BHCA of Huawei's default traffic model.
GCUP boards do not support the active/standby mode. The number of redundant boards can
be manually specified in the redundancy configuration. By default, if the number of GCUP
boards required is X in capacity calculation, another GCUP board is configured. Each BSC is
configured with at least two redundant boards.
Principles of GMCP configuration

GMCP boards are configured according to the IBCA deployment requirements. If the IBCA
function is enabled, one GMCP board supports 2048 TRXs. GMCP boards do not support the
active/standby mode.
Principles of NASP configuration

NASP boards are configured according to the deployment requirement of network assisted
WLAN identification. If this feature is enabled, a BSC is configured with only once NASP
board.
Principles of ENIUa configuration

If the feature of intelligent service identification is enabled, ENIUa boards are required. A
BSC is configured with only one ENIUa board.
Principles of ESAUa configuration

If a customer purchases the Nastar, the ESAUa board needs to be configured in the BSC. A
BSC is configured with only one ESAUa board.
2015-11-13
Huawei Confidential
Page 57 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
12
Naming Rules Design

This section designs the numbering and naming rules for all the NEs on the network to
make network topology clear and facilitate network management.
Standard numbering and naming facilitate maintenance. Directly locate faults by using
alarm information to improve maintenance efficiency.
Information, such as geographical distribution and the number of NEs, area names, office
names, and NE types
NE naming specifications and requirements of the customer. In the high-level design

(HLD), naming rules of NEs are determined based on naming conventions and planning
requirements of the customer, and Huawei's naming rules.
This document describes the naming and numbering rules recommended by Huawei. However, most
customers use their own NE naming rules. In actual applications, communicate with the customer and
then determine the naming and numbering rules based on the customer requirements and the rules
recommended in this document.
12.2 NE Naming Rules

12.2.1 Naming Rules of Areas
Network design is performed based on areas. Therefore, it is necessary to name the areas.
Use the short name of the geographical name of an area to name the area. For a geographical
name in China, use the capital letters in the full pinyin name.
For example, the publicly known short name of Lagos, Nigeria is LOS. Use LOS as the name
of Lagos.
2015-11-13
Huawei Confidential
Page 58 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
For example, the capital letters of the full pinyin name of Guangzhou is GZ. Use GZ as the
name of Guangzhou.
12.2.2 Naming Rules of Offices

Devices are installed in offices. Therefore, office naming is part of NE naming, and it is
important for locating NEs quickly.
Use the first three letters in the full English geographical name of an office. For a
geographical name in China, use the capital letters in the full pinyin name.
For example, a BSC office is located in Adekula in Nigeria. Use ADE as the name of the
Adekula office.
For example, a BSC office is located in Dian Xin Guang Chang in Xi'an. Use DXGC as the
name of the Dian Xin Guang Chang office.
If the short names of two are the same, lengthen the short name of one office to distinguish
them.
For example, two BSC offices are located in Okuno and Okuani in Nigeria. Use OKUN and
OKUA as the names of the offices in Okuno and Okuani respectively.
12.2.3 Naming Rules of Manufacturers

The network of an operator may use the devices of multiple manufacturers. Naming the
manufacturers and using the manufacturer names in NE naming can help quickly distinguish
the manufacturer of an NE.
Use the commonly used manufacturer short names in the telecom field.
Table 1.1 Manufacturer short names
Manufacturer
Short Name
Huawei
HW
Ericsson
ERI
ZTE
ZTE
Nortel
NOR
Motorola
MOT
Samsung
SAM
Alcatel-Lucent
AL
UTSTARCOMM UT
UT
Nokia-Siemens
NSN
Cisco
CIS
There are numerous manufacturers, and this document lists only the commonly known ones.
2015-11-13
Huawei Confidential
Page 59 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
12.2.4 Naming Rules of NEs

This document describes the naming rules only of the NEs on the radio side and the NEs
closely related to the radio side. The short names of the NEs are as follows:
Table 1.1 NE short names
Network Element Type
Description
BTS
Base transceiver station
eGBTS
Evolved GSM Base Transceiver Station
BSC
Base station BSC
AN
Access network
OMU
Operation and maintenance unit
MSC
Mobile switching center
MSCe
Mobile switching center emulation
MGW
Media gateway
HLR
Home location register
STP
Signaling transfer point
SGSN
Serving GPRS support node
GGSN
Gateway GPRS support node
DNS
Domain name server
BG
Border gateway NE08, NE40
RT
Router
LSW
LAN switch
FW
Firewall
M2K
M2000 server
Huawei recommends the following naming rules for NEs other than the BTS:
<A>_<B>_<C><D><E>
A stands for the short name of the area where the NE is located.
B stands for the short name of the office where the NE is located.
C stands for the short name of the manufacturer of the NE.
D stands for the short name of the NE.
E stands for the sequence number in the area where the NE is located.
For example, the second BSC (manufactured by Huawei) in Xi'an is located in Dian Xin
Guang Chang, and the BSC is named as follows:
2015-11-13
Huawei Confidential
Page 60 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
XA_DXGC_HWBSC2
For example, the first PDSN (manufactured by ZTE) in Lagos in Nigeria is located in
Adekula, and the PDSN is named as follows:
LOS_ADE_ZTEPDSN1
This document recommends this naming rule. Engineers can determine the specific naming rule based
on actual conditions. For example, if the devices on the operator's network are all provided by Huawei,
omit the manufacturer name to simplify the NE names.
The OMU and PCU are embedded in the BSC. However, they are independent NEs in terms of GSM
network structure. Therefore, name them independently. The OMU and the PCU are named based on the
BSC.
Generally, the number of BTSs is large. Therefore, simplify the BTS names. Name BTSs as
follows:
<A><B>
A stands for the name of area where the BTS is located, or the name of the property company
that manages the area where the BTS is located.
B stands for the BTS ID. For details about BTS IDs, see 12.3.2"Numbering Rules of BTS
IDs."
For example, BTS2 in Parkview in Nigeria is named Parkview2.
12.2.5 Naming Rules of Signaling Points

A signaling point refers to the originating signaling point code (OPC) or destination signaling
point code (DPC) of the BSC. The DPC naming rule is as follows:
The DPC naming rule is as follows:
<A>_OPC<B>
The DPC naming rule is as follows:
<A>_DPC<B>
A stands for the name of the BSC to which the DPC belongs. For details about BSC names,
see 12.2.4"Naming Rules of NEs."
B stands for the sequence number of the DPC in the BSC to which the DPC belongs.
For example, the second DPC of XA_DXGC_HWBSC2 is named as follows:
XA_DXGC_HWBSC2_DPC2
An OPC or DPC consists of 1 to 49 characters.
12.3 NE Numbering Rules

12.3.1 Numbering Rules of Entity IDs
An entity ID is a BSC ID. It is used to identify BSCs. The combination of MSC ID+BSC ID
can uniquely identify a BSC in the system. The value range is 0 to 65535. Each BSC ID in the
same MSC must be unique.
2015-11-13
Huawei Confidential
Page 61 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Number the BSCs with the entity IDs starting from 1 based on the order in which the BSCs
are launched in commercial use.
12.3.2 Numbering Rules of BTS IDs

A BTS ID is used to uniquely number a BTS in the BSC. The value range is 0 to 7999. Each
BTS ID in the BSC must be unique. Provide this numbering rule after you communicate with
the RF network planner. Name BTSs based on geographical areas. That is, allocate a
consecutive number range to a geographical area. Then, the maintenance personnel can learn
the geographical area where the BTS is located from the BTS ID (BTS name).
12.3.3 Numbering Rules of Cell IDs

A cell ID is used to uniquely number a cell in the BSC. The value range is 0 to 7999. Each
cell ID in the BSC must be unique. Provide this numbering rule after you communicate with
the RF network planner.
12.3.4 Numbering Rules of LACs

In actual applications, the RF network planner determines the LAC numbers.
12.3.5 Numbering Rules of MCCs and MNCs

Each country has a mobile country code (MCC). The operator provides the MCC. Each
country has multiple mobile network codes (MNCs). The operator provides the MNC.
MCC+MNC+LAC+CI comprise the cell global identifier (CGI) of a cell. A CGI can uniquely
identify a cell.
12.3.6 Numbering Rules of SPXs and DPXs

An SPX uniquely identifies an originating signaling point. The value range is 0 to 4.
Configure up to five originating signaling points. Use four originating signaling points for
GSM and reserve one for universal mobile telecommunications system (UMTS) for follow-up
GU convergence expansion.
A DPX uniquely identifies a destination signaling point. The BSC allocates indexes to all
destination signaling points. The indexes range from 0 to 427, that is, there can be up to 428
destination signaling points.
2015-11-13
Huawei Confidential
Page 62 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
13
BSC6910 Networking
Principles
This chapter describes the principles, advantages, and disadvantages of the mainstream
networking of the IP interface of the BSC6900.
13.1 Technical Principles

13.1.1 Overview
Table 1.1 lists the advantages and disadvantages when the GBSS15.0 BSC6910 adopts the
typical networking mode of IP transmission resource pool.
Table 1.1 Advantages and disadvantages when the GBSS15.0 BSC6910 adopts the typical
networking mode of IP transmission resource pool
Networking
Scheme
Description
Advantages
Promoted scheme: Pool

of independent interface
boards
Each port of the router

works in load-sharing
mode. The independent
BSC interface boards
form the resource pool.
Basic reliability: The

BTS is homed to
BSC's multiple
interface boards
that provide
resource pool,
thereby generating
a 1:N protection.
Low hardware cost:

Multiple interface
boards work
independently to
dynamically
balance the load,
thereby providing
the highest usage.
Low maintenance
2015-11-13
Huawei Confidential
Disadvantages
The reliability is low. The
SCTP is configured to multihoming, resulting call drop for
the ongoing calls instead of
the newly connected
subscribers.
Page 63 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
cost: The
maintenance is IPpath-free
configuration. IP
address can be
added on the core
network device
without changing
the configuration on
the BSC. The
configuration of the
network device is
simple and ports
can be easily added
to a board.
Optional scheme 1: Pool

of active/standby
interface boards+dualactive ports
Optional scheme 2: Pool

of active/standby
interface boards+manual
2015-11-13
Each port of the router

works in load-sharing
mode. The BSC interface
boards form the resource
pool, in which the
interface boards with
dual-active ports work in
active/standby mode. The
ports work independently.
The gateways also work
in active/standby mode.
The ports of the router

work in pairs using the
Virtual Router
High adaptability:
This scheme has no
special
requirements on
network device
configuration.
High reliability:
Fault of port or
board does not
affect services. In
addition, the 1:N
protection of
resource pool is
configured between
multiple pairs of
interface boards.
Low hardware cost:

Multiple pairs of
interface boards
dynamically
balance the load,
thereby providing
the highest usage.
Low maintenance
cost: The
configuration. IP
address can be
added on the core
network device
without changing
the configuration on
the BSC.
High reliability:
Fault of port or
board does not
Huawei Confidential
After a switchover in case of

a fault, one port takes over
the services from another,
affecting those excess
services.
Compared with the solution

with independent boards,
the data processing and
connection capabilities
decreases by half.
There are special

requirements on the
configuration of network
devices, for example,
active/standby route
policies are required.
High hardware cost: The

number of port and
interface boards doubles
Page 64 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
active/standby link
aggregation groups
(LAGs)
Redundancy Protocol
(VRRP). The BSC
interface boards working
in active/standby mode
form the resource pool.
The ports work in
active/standby mode. The
ports work independently.
affect services. In
addition, the 1:N
protection of
resource pool is
configured between
multiple pairs of
interface boards.
Compared with nonpool solutions, the

hardware cost is
lower. Multiple
pairs of interface
boards dynamically
balance the load,
thereby providing
the highest usage.
Low maintenance
cost: The
configuration.
Transmission
capacity expansion
and adjustments do
not require
migration.
Reconstruction from
the solution of
active/standby
interface boards
with dual-active
ports requires few
changes. Only
small changes are
required on the
BSC.
compared with other two

resource pool solutions.
Complicated configuration:
This scheme has special
requirements on the
configuration of network
devices, for example, the
router requires the
configuration of VRRP,
VLANIF, and layer-2
interface. The BSC
requires the configuration
of dual BFD+ARP
detection.
The trunk connection

between routers may not
be reliable and affects the
bandwidth. If the trunk is
disconnected, services may
be affected. If the
bandwidth is insufficient,
services are congested.
Table 1.2 lists the advantages and disadvantages when the BSC6910 does not adopt the
typical networking mode of IP transmission resource pool.
Table 1.2 Advantages and disadvantages when the BSC6910 does not adopt the typical
networking mode of IP transmission resource pool
Networ Description
king
Scheme
Advantages
Disadvantages
Promoted
scheme:
Pool of
active/stan
dby
The implementation
is simple, and the
application
technology is proven.
2015-11-13
The ports of the router

work in pairs using
the VRRP. The BSC
interface boards work
in active/standby
Complicated
configuration: The router
requires the configuration
of VRRP, VLANIF, and
layer-2 interface. The
Huawei Confidential
Selection
Principle (SubScenario)
The promoted
solution is the
default solution. Try
your best to
recommend this
Page 65 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
interface
boards+ma
nual
active/stan
dby LAGs
(pool of
independe
nt IP
addresses)
mode with manual

active/standby LAG.
Either the active or
standby port bears the
data streams.
Optional
scheme 1:
Pool of
active/stan
dby
interface
boards+du
al-active
ports (pool
of
independe
nt IP
addresses)
Ports of routers work

in load-sharing mode.
Each port provided by
the active and standby
boards of the BSC is
dual-homed to one
router.
Not
recommen
ded:
independe
nt interface
board
2015-11-13
Divide the signaling

and IP addresses of
the local and peer
ends into two groups,
in which one group of
IP is routed to the left
path and another
group of IP is routed
to the right path
through the highpriority route
configuration, thereby
achieving load
sharing. Furthermore,
the two routes work in
backup mode,
improving reliability.
Ports of routers work

in load-sharing mode.
Two independent
boards are connected
to two routers,
respectively. The ports
are set to LAG mode.
In addition, interboard dual-homing
BSC requires the

configuration of dual
BFD+ARP detection.
Load sharing is
implemented. The
router and
network path
usage is
increased.
Effective load
sharing: The load
both in the
sending and
receiving
directions are
shared. In
addition, an endto-end
deployment can
be implemented,
that is, dual-path
protection is
configured for
intermediate
networks.
solution to the
customer.
The trunk connection

between routers may not
be reliable and affects the
bandwidth. If the trunk is
disconnected, services
may be affected. If the
bandwidth is insufficient,
services are congested.
Complicated planning and

configuration: IP addresses
need to be divided to two
groups. Different route
priorities are configured for
each group.
The data flow path

is clear and
consistent in
sending and
receiving, which
provides good
maintainability
and helps fault
location and QoS
monitoring.
Basic reliability:
The reliability is low.
The BTS is
Services are affected if the
homed to BSC's board is faulty.
multiple interface
boards that
provide resource
pool, thereby
generating a 1:N
protection. The
Huawei Confidential
The customer
requires load-sharing
networking which
satisfies the
following
requirements:
Services address in
pairs must be
configured for
core network
devices.
Active/standby
route policies are
required.
Not recommended
Page 66 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
needs to be configured
on the bearer plane.
LAG protection is
configured for
intra-board ports.
Low maintenance
cost: The
configuration is
simple.
Expansion can be
easily
implemented. The
data path is clear.
High adaptability:
This scheme has
no special
requirements on
network device
configuration.
13.1.2 Technical Specifications

Table 1.1 Technical specifications
Fault Detection
Port
Failure
Detection
on the
Physical
Layer
BFD
ARP
(100 ms x
3)
(300 ms x
3)
Best: 300 ms
Best: 200 ms
Worst: 600 ms
Worst: 300 ms
Best: 6
seconds
Switchover
ARP Message Total

Sending
Best: 100 ms
Immediate
transmission
Generally, it is
shorter than 1
second.
Immediate
Worst: 2 seconds transmission
Generally, it is
shorter than 3
seconds.
Worst: 1 second
Worst: 9
seconds
Board
Failure
The fault
NA
detection
duration varies
with the
component
where the fault
occurs.
Best: 100 ms
Remote Same as local fault detection

Failure
Note: The switchover duration is related to the number of routes, number of IP paths/SCTP links, and CPU
usage of the interface board/SCU.
2015-11-13
Huawei Confidential
Page 67 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
These specifications are internally used. Do not promise these to the customer unless it is necessary.
There are multiple detection modes. The actual switchover duration is determined by the
quickest detection mode.
13.1.3 Technical Description

1.
2.
Networking schemes in GBSS15.0:
Pool of independent interface boards: An IP pool is added to a network that has

independent dual-active interface boards.
Pool of dual-active ports of active/standby boards: An IP pool is added to a network

that has load-sharing ports and active/standby boards.
Pool of active/standby boards+manual active/standby LAGs: It is an alternative for

the solution of active/standby boards+active /standby ports. Networking changes are
invisible to customers, but only internal configuration varies with flexible change.
Furthermore, the reliability of the IP pool is improved and use this scheme on the A
interface.
Principle description of active/standby boards+manual active/standby LAGs

Manual active/standby LAGs are adopted on active/standby boards. Link aggregation
means to aggregate multiple physical links of the active and standby boards to a logical
link, thereby generating an LAG. The LAG can increase bandwidth, improve
transmission reliability, and transmit data streams of users over multiple links at the same
time. Therefore, if an Ethernet port in an LAG is faulty, services are not interrupted. The
BSC supports 1:1 active/standby link mode. If the active link is available, it sends and
receives data. If the active link is faulty, the standby link takes over the data sending and
receiving from the active link. This is known as a port active/standby mode based on link
aggregation. In port active/standby mode, the active port transmits and receives data, and
the standby port is generally not used or is used only for link detection. In addition, the
active/standby port can be inconsistent with the active/standby board. That is, the active
port can be on the active board or standby board.
Figure 2.1 describes the port switchover.

Figure 2.1 Port switchover
As shown in Figure 2.1, after switchover, active and standby boards remain unchanged. Port 0
of the active and standby boards are switched over. IP1 moves to port 0 in the standby board.
Figure 2.2 describes the board switchover.
2015-11-13
Huawei Confidential
Page 68 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 2.2 Board switchover
Use networking based on active/standby boards, manual active/standby LAG, Layer-3

networking, router VRRP, and device IP for communication.
Typical configuration:
Device IP address: The active and standby boards share the same device IP address, and the
default IP subnet mask is 32-bit subnet mask, which cannot be configured.
ETHIP: Each port is configured with an IP address, and uses the 29-bit subnet mask. Different
ports belong to different network segments.
If ARP detection is not enabled on the standby port, generally no IP address is configured.
Recommended detection mode: Dual BFD detection on the active port. In this mode, the
destination addresses are the port IP addresses corresponding to the two routers.
In the ARP detection on the standby port, the destination addresses are the VRRP IP addresses
of the two routers.
2015-11-13
Huawei Confidential
Page 69 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
14
Optical Interface MSP
This chapter describes the optical interface (STM-1 interworking design).
14.1 MSP Design Guide

Optical interface interworking mainly describes the optical interface boards on the BSC side
and the directly connected optical interface NEs (including the design of interworking
between transmission devices).
14.1.1 STM-1 Tributary Mode Selection

For channelized optical interfaces, the ITU standard does not clearly specify the mode in
which timeslots are sequenced in the VC-4. Therefore, the VC-12 numbering rules in the VC4 of different manufacturers are different. Three modes are available based on the numbering
rules:
Huawei mode: In this mode, numbering is implemented based on timeslot numbers, and this
mode is known as the sequence-based mode. The VC-12 is numbered based on TUG-3, then
based on TUG-2, and finally based on TU-12.
Lucent mode: In this mode, numbering is implemented based on line numbers, and this mode
is known as the insertion mode. The VC-12 is numbered based on TU-12, then based on
TUG-2, and finally based on TUG-3.
Alcatel Mode: This mode is seldom used. In this mode, the VC-12 is numbered based on
TUG-2, then based on TU-12, and finally based on TUG-3.
The optical interface of the BSC supports the preceding modes. In the case of interworking
with network devices, find out the mode used at the peer end, and then configure the mode at
the local end accordingly.
In addition, the VC-12 numbers of network devices start from 1 generally, and the E1T1
numbers of BSCs start from 0. Therefore, the E1T1 number on the BSC is not the same as the
VC-12 number of the network device even if the numbering modes are the same. Instead,
E1T1 number = VC-12 number 1.
14.1.2 MSP Mode Selection

By default, use the MSP 1+1 single-end non-recovery mode.
2015-11-13
Huawei Confidential
Page 70 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Table 1.1 MSP advantages and disadvantages

Solution Advantage
Disadvantage
MSP 1+1
single-end
nonrecovery
mode
A number of
It is the default mode and
boards/backplanes of is promoted.
Huawei's BSC do
not support the
single-end mode.
Negotiation between the

two ends is not required,
and the compatibility is
high.
The switchover speed is
high.
Selection Principle
(Sub-Scenario)
The paths in the two

The self-healing rate is high directions are
inconsistent, and the
when multi-point faults
implementation is
occur.
complicated.
MSP 1+1
double-end
nonrecovery
mode
Negotiation between the

two ends is required. The
paths in the two directions
must be the same for easy
understanding and fault
location.
Negotiation between
the two ends is
required, and
protocol
compatibility is
required.
MSP 1:1
double-end
recovery
mode
The boards/backplanes
that do not support the
1+1 single-end mode
use this mode.
The remote device or the

customer requires this
mode.
The remote device does

not support MSP 1+1
protection and only
supports 1:1 protection.
14.1.3 Parameter Configuration

Table 1.1 J0, J1, and J2 configuration
2015-11-13
Length
Mode
(Byte)
Sending
Receiving
Selection Principle
16
The user can enter 15

bytes (the most
significant bit cannot be
1), and the first byte is
automatically generated
(15-byte CRC7 check
value, the most
significant bit is 1).
Then, the bytes are sent.
The receiving
end adopts the
same
algorithm for
comparison.
It is the standard mode and

promoted by default. The
detailed contents are
negotiated with the remote
end. The default SBS 155 is
converted into hexadecimal
0X534253203135352020202
020202020.
The single bytes set by

the user are sent
consecutively.
The bytes are

checked one
by one.
The remote end does not

support the 16-byte mode.
The detailed contents are
end.
64
The user can enter 62
The receiving
The remote end does not
Huawei Confidential
Page 71 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Length
Mode
(Byte)
NULL
Sending
Receiving
Selection Principle
bytes (0D0A cannot be

present), and the last two
bytes 0D0A are
automatically generated.
Then, the bytes are sent.
end adopts the

same
algorithm for
comparison.
support the 16-byte mode.

The detailed contents are
end.
All zeros without a

format are sent
consecutively.
A check is not
performed.
The remote configuration

cannot be known, or the
formats at the two ends do
not match.
The J2 of the AOUa and POUa does not support the 1-byte mode.
If all zeros are received, no alarm is generated regardless of the format set at the local end.
If the number of bytes/characters that the user enters is smaller than the required number, zeros or spaces
are padded automatically (this can be configured by running Set OPT: JAUTOADD=ZERO/SPACE;).
By default, spaces are padded. In interworking, ensure that the padding modes (ZERO/SPACE) at the
two ends are the same.
In NULL mode, the J byte does not provide the continuity check function and do not use it as the default
configuration.
Two configuration modes, that is, the character mode and the hexadecimal number mode, are available.
Use the character mode because it is clear and not prone to errors.
14.1.4 S1 Configuration
The S1 byte is in the first column and the ninth row in MSOH of the SDH frame structure.
The least significant four bits (bits 5 to 8) transmit the synchronization status information
(SSM). Generally, the S1 byte refers to the least significant four bits. The most significant
four bits are reserved. Huawei network devices expand the most significant four bits so that
they can be used to transmit the clock ID.
Table 1.1 describes the definition of the S1 in G.707. The larger value indicates the lower
clock quality.
Table 1.1 Definition of the S1 in G.707
2015-11-13
Quality
Level
San1, San2, San3, and

San4
Synchronization Quality
Level
0000
The quality is unknown (on the

existing synchronization network).
0010
G.811 recommended clock
0100
SSU-A (G.812 transit exchange)
1000
SSU-B (G.812 local office)
11
1011
Synchronous equipment timing

source (SETS)
Huawei Confidential
Page 72 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Quality
Level
San1, San2, San3, and

San4
Synchronization Quality
Level
15
1111
Synchronization is unavailable.
The requirement for the clock of the BSC is stratum 3 and category A. Generally, the clock is
obtained from the optical interface on the core network side. The BSC selects the clock based
on the configured priority and does not process the S1 byte sent by the remote end of the
optical interface.
The synchronous digital hierarchy (SDH) network of the Iub/Abis interface possesses its own
clock and seldom obtains the clock from the BSC. In special cases, for example, the SDH
network is the small-scale network dedicated for BTS backhaul, the SDH may trace the clock
of the BSC. That is, the optical interface of the BSC seldom works as the clock source of the
SDH devices.
Conclusion:
Set S1 to the default value 11 on the BSC, that is, the SETS.
The BSC does not have a requirement for the S1 byte of the remote device.
14.1.5 C2 Configuration
On the BSC, the C2 byte does not need to be configured by the user but is determined by the
board application type. If the interface is a channelized optical interface, the C2 byte is 0X02
(TUG structure); if the interface is a non-channelized optical interface, the C2 byte can be
0X13 (ATM mapping) or 0X16 (PPP Mapping). Table 1.1 lists the details of the C2
configuration.
Table 1.1 Details of the C2 configuration
Board
Function
C2 Value
C2 Interpretation
POUc
TDM+FR
0X02
TUG structure
It is required that the remote configuration be the same as the local configuration.
14.1.6 MSP Support Capabilities of Boards

Table 1.1 describes the MSP support capabilities of boards.
Table 1.1 MSP support capabilities of the boards of the BSC
2015-11-13
Board
Function
MSP 1:1
MSP 1+1
single-end
MSP 1+1
double-end
Remarks
POUc
TDM+FR
Y(*)
It supports the singleend mode only after it is

inserted in the new
backplane.
Huawei Confidential
Page 73 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The new backplane refers to the backplane whose version is VER.C or later. The backplanes in the
subracks delivered after about April 2008 are new backplanes.
14.2 MSP Technical Description

Technical Description of the BSC Optical Interfaces
Framing mode
In optical interface interworking, three framing modes, that is, Huawei mode, Lucent mode,
and Alcatel mode, are available. The names of the three modes are different from the names
that partners use. This makes the communication between the technical personnel difficult for
interworking negotiation. Therefore, the following describes the three modes and provides a
reference for the communication between technical personnel.
Three SDH optical interface framing modes are available. They are Huawei mode (X
increments first, then Y, and finally Z), Lucent mode (Z increments first, then Y, and finally
X), and Alcatel mode (Y increments first, then Z, and finally X). The arrangement of the 63
PCMs on the optical interface varies with the framing mode. However, the first E1s in the
three modes are the same.
For details, see Table 1.1.
2015-11-13
Huawei Confidential
Page 74 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Table 1.1 Framing mode comparison
2015-11-13
Huawei Confidential
Page 75 of 258
GBSS17.0 & BSC6910)
2015-11-13
Huawei Confidential
CONFIDENTIAL
Page 76 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Names and description of the optical interface interworking parameters:

Table 1.2 Optical interface interworking parameters
Parameter
Name
Recommended
Value
Description
Tributary
Numbering
Negotiated with the

peer end
It indicates the tributary arrangement sequence

in an SDH frame. Optional modes: Lucent
mode, Huawei mode, and Alcatel mode.
J0 trace
mismatch alarm
switch
Negotiated with the

peer end
It indicates whether RTIM is reported if TX J0

Byte (Hexadecimal, Character String) or
Expect RX J0 Byte (Hexadecimal, Character
String) on both ends are inconsistent in optical
interface interworking. Optional values: YES
and NO.
J0 Mode
Negotiated with the

peer end
It indicates the maximum length of TX J0 Byte

(Hexadecimal, Character String) and Expect
RX J0 Byte (Hexadecimal, Character
String).
The J0 byte lengths of the optical interface
devices provided by different manufacturers are
different. Therefore, the maximum length of
TX J0 Byte (Hexadecimal, Character String)
and Expect RX J0 Byte (Hexadecimal,
Character String) can be adjusted so that
Huawei's device can interwork with the optical
interface devices of other manufacturers and
can be compatible with the earlier versions.
The setting of this parameter must be consistent
with the corresponding parameter on the peer
optical interface device.
TX J0 Byte
(Hexadecimal,
Character String)
Negotiated with the

peer end
This byte is used to repeatedly send the

segment access point identifier so that the
receiving end can determine based on this byte
that the receiving end and the specified sending
end are in the continuous connection state. The
value is a character string of 0 to 15 characters.
Expect RX J0
Byte
(Hexadecimal,
Character String)
Negotiated with the

peer end
This byte is used to repeatedly send the

segment access point identifier so that the
receiving end can determine based on this byte
that the receiving end and the specified sending
end are in the continuous connection state.
The value is a character string of 0 to 15
characters.
J1 trace
mismatch alarm
switch
2015-11-13
Negotiated with the

peer end
It indicates whether HPTIM is reported if TX

J1 Byte (Hexadecimal, Character String) or
Expect RX J1 Byte (Hexadecimal, Character
String) on both ends are inconsistent in optical
Huawei Confidential
Page 77 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Parameter
Name
Recommended
Value
Description
interface interworking.
J1Mode
Negotiated with the

peer end
It indicates the maximum length of TX J1 Byte

(Hexadecimal, Character String) and Expect
RX J1 Byte (Hexadecimal, Character
String).
The J1 byte lengths of the optical interface
devices provided by different manufacturers are
different. Therefore, the maximum length of
TX J1 Byte (Hexadecimal, Character String)
and Expect RX J1 Byte (Hexadecimal,
Character String) can be adjusted so that
Huawei's device can interwork with the optical
interface devices of other manufacturers and
can be compatible with the earlier versions.
The setting of this parameter must be consistent
with the corresponding parameter on the peer
optical interface device.
The optional lengths are 1 byte, 16 bytes, and
64 bytes.
TX J1 Byte
(Hexadecimal,
Character String)
Negotiated with the

peer end
The J1 byte is used to repeatedly send the

higher order path access point identifier so that
the receiving end of the path can determine
based on this byte that the receiving end and the
specified sending end are in the continuous
connection (the path is continuously connected)
state.
characters.
Expect RX J1
Byte
(Hexadecimal,
Character String)
Negotiated with the

peer end

state.
characters.
2015-11-13
Expect RX J1
Byte
(Hexadecimal,
Character String)
Negotiated with the

peer end

state. The value is a character string of 0 to 15
characters.
Set TX S1 Byte
Negotiated with the

peer end
It indicates whether the S1 byte is enabled.

Optional values: YES and NO.
Huawei Confidential
Page 78 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Parameter
Name
Recommended
Value
Description
TX S1 Byte
Negotiated with the

peer end
It indicates the synchronization state byte. It is

a character string of 0 to 15 characters.
TX Frame
Format
Negotiated with the

peer end
In E1 working mode, only DOUBLE_FRAME

and CRC4_MULTIFRAME are supported.
In T1 working mode, only SUPER_FRAME
and EXTENDED_SUPER_FRAME are
supported.
RX Frame
Format
Negotiated with the

peer end
In E1 working mode, only DOUBLE_FRAME

and CRC4_MULTIFRAME are supported.
In T1 working mode, only SUPER_FRAME
and EXTENDED_SUPER_FRAME are
supported.
Configuration principles of the optical interface interworking data

If optical interface interworking is adopted, pay attention to the preceding parameters and
configure the tributary numbering mode, J0, J1, J2, S1, frame format, and C2. The
configuration principles are as follows:
Tributary Numbering on the BSC side must be consistent with Tributary Numbering
on the MSC side.
TX J0 Byte (Hexadecimal, Character String) on the BSC side must be consistent with
Expect RX J0 Byte (Hexadecimal, Character String) on the MSC side. Expect RX J0
Byte (Hexadecimal, Character String) on the BSC side must be consistent with TX J0
Byte (Hexadecimal, Character String) on the MSC side. Otherwise, ALM-20225 is
generated.
generated.
generated.
The S1 synchronization state flag must be consistent with the flag on the MSC side, and
the K1 and K2 parameters must be consistent with the parameters on the MSC side.
The sending and receiving framing formats on the two sides must be consistent.
The C2 parameter cannot be configured on the BSC6000. This parameter is set internally in the
software, and the default value is 0x02. Therefore, check whether this parameter is set to 0x02 on the
MSC side. If it cannot be confirmed, right-click the corresponding OIUa and select Query Interface
Board Port State to check the C2 value sent by the remote end. The values on the two sides must be
consistent. Otherwise, the interworking fails.
2015-11-13
Huawei Confidential
Page 79 of 258
GBSS17.0 & BSC6910)
15
CONFIDENTIAL
Detection Mechanism
15.1 Restrictions of the Design

Table 1.1 Restrictions of the fault detection mechanism of the BSC
2015-11-13
Huawei Confidential
Page 80 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Physical
detection
Fixed: 300 ms x 2 times
ETH EFM
It is associated with the port status. It can trigger an active/standby port
Detection switchover but cannot trigger a trunk switchover.
The GB IP interface does not support ETH OAM.
The EFM of the RNC requires that the PDU packet sending period of the
remote end be the same as the period of the local end. Otherwise,
negotiation fails. Set the period to one second.
ETH CFM
The GB IP interface does not support ETH OAM.
Detection
PPP
Default: 10 seconds x 5 times
Status
Detection
Default: 300 ms x 3 times
ARP
Detection 16/board. The number of detections on a port is not limited. One port
cannot perform BFD and ARP detections at the same time.
The ARP automatically associates the next-hop route and sets the port as
the key detection association port.
The standby port supports only one ARP detection.
The detected IP address of the standby port and the IP address of the active
port can be in the same network segment.
ARP does not support flow-based VLAN tagging and supports only nexthop-based tagging.
Only the FE and GE ports support BFD, and the POS interface does not
BFD
Detection support BFD.
16/board (SBFD+MBFD); 512/board (only the AIU). The number of
detections on a port is not limited.
Default: 30 ms x 3 times
One IP address can perform either SBFD or ARP detection, but the two
detections cannot be performed at the same time.
Configure the MBFD associated IPRT and IP path. A 10-second delay is
introduced in the case of path faults.
The SBFD automatically associates the next-hop route and sets the port as
the key detection association port.
The active port can be associated with multiple detections (multiple ARP
and BFD detections). The port becomes Down only if all the detections fail.
The standby port does not support the BFD detection and supports only the
ARP detection.
The route and path can be associated with multiple detections (ARP/SBFD,
MBFD, and ping). The route or path is considered faulty if any detection
fails.
The SBFD and MBFD distinguish sessions based on the source and
destination IP addresses. They do not support multiple BFD detections on
the same address pair or VLAN-based BFD.
The MBFD uses port 4784 by default (stipulated in the new draft). After
you enable the negotiation switch (by running SET BFDPROTOSW), the
2015-11-13
Huawei Confidential
Page 81 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
local port number can be automatically adjusted to the same port number as
the remote end. It can be used for the interworking with Huawei's router of
an earlier version (the port number is 3784).
SCTP
Default: MinRTO = 1000 ms; MaxRTO = 3000 ms; HB interval = 1000 ms;
Detection Association max retrains = 4; Path max retrans = 2
Table 1.2 Restrictions of the fault detection mechanism of the base station
NodeB Restrictions (RAN12.0)
GBTS
Restrictions
(GBSS9.0)
Physical
Detection
No smoothness on the upper layer
NA
ETH EFM
Detection
NA
NA
ETH CFM
Detection
NA
NA
PPP Status
Detection
NA
NA
ARP
Detection
Not supported
Not supported
BFD
Detection
The SBFD automatically associates the route whose

next hop is the BFD destination address and does not
associate with ports or other routes.
2 BFDs/board
The MBFD automatically associates the single-point

route of the BFD destination address and does not
associate with ports or other routes. The MBFD does
not detect the standby route.
SCTP
Detection
2015-11-13
Does the SBFD

automatically
associate the route
whose next hop is
The BFD destination
address?
In the active/standby route scenario, the applications

with and without BFD are different. After the active
route fails and the data is switched to the standby
route, if the active route is recovered, then: 1) If the
BFD is not configured, the priority of the active route
is higher than that of the standby router; therefore, the
data is switched back to the active route. 2) If the BFD
is configured, the active route is available but the
system still uses the standby route.
Enable the MBFD to

associate the IP
route.
Same as the RNC. Default: MinRTO = 1000 ms;

MaxRTO = 3000 ms; HB interval = 1000 ms;
Association max retrains = 4; Path max retrans = 2
N/A
Huawei Confidential
Page 82 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
15.2 BFD Detection

BFD definition:
The BFD can detect channel faults between forwarding engines with light load and quick
response. Such faults include interface faults, data link faults, and forwarding engine faults.
The BFD provides an independent mechanism to detect any media or protocol layer in real
time. The detection time and payload ranges are wide, ensuring that only the BFD is used and
avoiding conflicts with other detections.
BFD is short for bidirectional forwarding detection.
The BFD is implemented on the service layer of the system and the detection is specific to the
connectivity for service packet forwarding.
The BFD can be used on the top layer of any data protocol to detect multiple network layers.
The BFD can detect faults on any type of channel between systems.
The BFD completes detection in milliseconds. It can implement rapid detection and
switchover in a short time and can help to locate link faults.
One port can enable either the ARP or BFD detection at a time.
Recommended commercial solution:

In the most recommended solution, the dual BFD detection is implemented on the active port,
the ARP detection is implemented on the standby port, and associated objects are ports.
Enable two BFD sessions on the active port. One BFD session detects the physical IP address
of the active router, and the other detects the physical IP address of the standby router. Set
both BFDs to key detections, and to enable automatic switchover of the active/standby ports
when two BFDs are interrupted simultaneously.
Enable one ARP probe session on the standby port to detect virtual IP addresses of the two
VRRP routers.
If both BFD sessions on the active port detect a fault but the ARP probe on the standby port
indicates no abnormality, the switchover between the active port and the standby port is
triggered automatically.
2015-11-13
Huawei Confidential
Page 83 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Diagram of the promoted commercial solution
Application analysis:
The BFD application in the current VRRP networking is defective: If the BFD is enabled on
the active and standby ports, the remote address in the BFD after the active and standby ports
are switched over remains unchanged; therefore, the remote address must be a fixed virtual
address. In VRRP networking, this address can only be a VRRP virtual address, but the VRRP
address of Huawei's router does not support the BFD detection, and the routers of other
suppliers may not support the BFD detection. Therefore, the VRRP virtual address mode
cannot be used.
Hence, Huawei proposes the dual BFD detection solution. On the BSC side, two BFD
detections are initiated on the active port, and the remote ends are the interface addresses of
the two routers. Currently, the VLAN interface addresses are used for implementing solutions,
such as VLAN separation. On the router side, the BFD is associated with static routes, and
then static routes are introduced into the dynamic routing protocol to ensure rapid
convergence of downlink routes.
The active port on the BSC side determines the status of the two BFD detections. The active
port considers the link down only when the status of both BFD detections is Down. In this
way, rapid switchover can be implemented. When the active port is normal, a fault on the
standby port does not affect services; therefore, rapid detection and switchover are not
required, and use the ARP detection.
Application scenario of the multi-hop BFD (do not use):
Generally, the intermediate network provides the dynamic routing protocol and rapid
switchover mechanisms for protection. A service NE requires only the protection between the
NE and the access router instead of the protection across the intermediate network. Therefore,
the E2E multi-hop BFD is not required.
If the scale of the bearer network is large, the multi-hop BFD cannot be completed quickly,
and increases the network load.
Therefore, the multi-hop BFD is used only in the following scenarios:
2015-11-13
The redundant path is independent on the network. The network is not protected. The
end-to-end detection is required.
The number of routers is small, and no dynamic routing protocol is used between routers.
Huawei Confidential
Page 84 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The access router does not run the dynamic routing protocol.
15.3 ARP Detection

Compared with the BFD detection, the ARP detection is slow and completed in seconds.
Any gateway can respond to ARP requests. Therefore, the ARP detection does not require
support from the peer end. In the ARP detection, broadcast packets are used and only the
FE/GE addresses in the same network segment can be detected. Generally, the BFD detection
is preferred, and the ARP detection is seldom used.
The main scenarios of the ARP detection in the recommended networking solution are as
follows:
When the dual BFD detection is configured on the active port, the standby port adopts
the ARP detection.
The ARP detection is used independently.
The purpose of the ARP detection is to detect the connectivity of gateways. If the current
networking mode is VRRP, the addresses detected in the ARP detection are VRRP virtual
addresses.
15.4 IP PM Detection
Functions of IP PM
IP performance monitor (PM) is an IP transmission QoS detection solution. The IP
transmission QoS detection is the basis for the RAN system to perform flow control and
admission control.
IP PM provides the following functions:
Detects the delay, jitter, and packet loss in IP transmission.
Obtains the packet loss and delay of all the IP paths of a logical port. Then, based on
a certain algorithm module, dynamically adjusts the bandwidth of the logical port. If
the bandwidth is reduced, packet loss is reduced and the efficiency is improved. If the
bandwidth is increased, the bandwidth usage is improved.
Detects the IP path connectivity and uses alarms to report the detection result to users.
Basic principles of IP PM
IP PM is similar to ATM OAM PM. Forward monitoring (FM) and backward reporting
(BR) are used to detect the packet loss condition along the path. One end sends FM
messages periodically to indicate the number of packets sent by this end. The period can
be specified to an interval or the number of sent packets. After receiving the FM
messages, the remote end replies with BR responses to report the number of received
packets. Then, the transmitting end measures the packet loss condition based on the BR
responses.
The basic process is as follows:
2015-11-13
Huawei Confidential
Page 85 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
When the PM initiator needs to start the PM function, the PM initiator sends an ACT
frame to the PM receiver. After receiving the ACT frame, the PM receiver enables the
BR function from the PM receiver to the PM initiator and replies the PM initiator with an
ACT ACK frame. After receiving the ACT ACK frame, the PM initiator enables the FM
function from the PM initiator to the PM receiver. Then, the PM initiator sends FM
frames periodically to the PM receiver, and the PM receiver replies the PM initiator with
BR frames after receiving the FM frames.
To enable the bidirectional link QoS detection, set up PM sessions in the directions from
A to B and from B to A.
Recommended IP PM solution on the BSC side

If IP paths require the IP PM detection and statistics to monitor the link quality, enable IP
PM.
For the multiple flows whose source and destination addresses are the same (for example, when the
boards work in active/standby mode and the ports work in load-sharing mode), the detection data and
subscriber data may travel along different paths, and therefore the IP PM result is not reliable. In this
case, disable IP PM.
Currently, only the Abis IP path supports IP PM. This function must be used with the BTS. Therefore,
ensure that the versions of both ends support this function.
GBSS14.0 supports only the BSC providing FE and BTS providing E1.
Requirements of IP PM for the transmission network

The DSCP value cannot be modified on the transmission network.
Scenarios that require IP PM

In the scenario where the quality of an IP transmission link fluctuates a lot (for example,
ADSL), use dynamic bandwidth adjustment for the logical interface. In this case, IP PM
must be enabled so that the currently available bandwidth can be adjusted based on the
detection result.
Basic principle: The closed loop control system is used to eliminate packet loss in the
BSC, and the output rate of the logical interface of a BTS cannot exceed the measured
bottleneck bandwidth of the logical interface.
The closed loop control system detects the jitter or packet loss ratio of a detected flow on
the transmission link from the BSC to the BTS. When the jitter or packet loss ratio of the
detected link is increased, the closed loop control system adjusts the output rate of the
logical interface of the link to reduce the link load and relieve the link congestion.
The adjustment of the rate of the logical interface inevitably affects the queue of the
logical interface. If the queue threshold is reached, the system informs the radio service
processing board of the queue congestion in back pressure mode. The service processing
board reduces the pressure. IP PM closed loop adjustment is used externally and closed
loop adjustment of congestion back pressure is used internally. In this way, end-to-end
congestion avoidance is implemented externally and internally in a joint manner.
2015-11-13
Huawei Confidential
Page 86 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
16
IP Interworking Design
This section describes the IP interworking design, including IP technical solutions, such as
layer-2 and layer-3 networking design in IP networking, IP address planning, VLAN planning,
and address planning for the maintenance channel, in IP RAN networking.
16.1 IP Planning on the BSC Side

1.
Description of the IP addresses of interface boards

Interface board FG2c/GOUc/EXOUa that supports Ethernet ports:
Each interface board supports up to 16 device IP addresses (DEVIP, logical IP address in
the board).
Each port can be configured with one active Ethernet port address (ETHIP) and up to
five ETHIPs (configured only for the active port in port backup). Each port being used
must be configured with at least the active Ethernet address.
2.
2015-11-13
Restrictions of IP address planning
IP addresses are determined by the customer based on the actual network conditions
and belong to the A/B/C category. An IP address consists of a network segment and a
host segment. The host segment cannot be all 0s or all 1s (when CIDR is used, the
selected IP address cannot be an invalid address in the A/B/C category). The first byte
of an IP address cannot be 0 or 127.
The network segment of the planned IP address and the network segment of the BAM
internal network address cannot partially or totally overlap.
Device IP address: The network segment of a device IP address and the network
segment of the BAM internal network address cannot partially or totally overlap. A
device IP address cannot be the same as or in the same network segment as any
configured IP address in the BSC (it can be an IP address of the Ethernet port or
remote address of the SCTP link). The device IP addresses configured on the same
BSC cannot be in the same subnet, and the device IP addresses configured for
different interface boards cannot be the same. When the CBS or MDSP service uses a
device IP address, other services (such as the SCTP link) cannot use this device IP
address. When a service other than CBS and MDSP uses a device IP address, CBS or
MDSP cannot use this device IP address. However, the services other than CBS and
MDSP can share the same device IP address.
ETHIP: The network segments of different Ethernet port IP addresses cannot partially
or totally overlap. The network segments of the active and standby IP addresses of the
Huawei Confidential
Page 87 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
same port cannot partially or totally overlap. An Ethernet port IP address cannot be
the same as or in the same network segment as any configured IP address in the BSC
(it can be a device IP address).
3.
Principles of IP address planning
On the BSC side, use the device IP communication mode. The default 32-bit and 29bit subnet masks are used for the device IP address and port IP address respectively.
In network segment planning, IP addresses must be sufficient, and consider

reservation for extensibility. Use the simplest possible configuration for the remote
router.
In IP layer-2 (FE) networking, the ETHIPs of the ports on the BSC interface board
are in the same network segment as the FE port IP addresses of the BTS. Use port IP
communication for the BTS. One BTS uses one port IP address. The signaling,
services, and O&M share the same IP address. Whether an independent NE adopts
separate O&M IP address and service IP address depends on whether the O&M and
service are separated on the bearer network. If the O&M and service are separated on
the bearer network, the NE adopts separate logical O&M IP address and service IP
address.
In IP layer-3 networking, the length of the subnet mask of the address on the BSC
side is irrelevant to the number of BTSs but is determined by the number of required
addresses. In this case, further network segment separation is required. For example,
an ETHIP must be in the same network segment as its gateway. If the gateway uses
only one IP address, two valid addresses are required, and 30-bit subnet masks can be
used. If the gateway adopts VRRP, the gateway requires one virtual IP addresses and
two real IP addresses, and 29-bit subnet masks can be used.
The device IP addresses of interface boards are in different network segments. The
30-bit or 32-bit subnet mask can be used (the 32-bit subnet mask is 255.255.255.255).
Use the 32-bit subnet mask.
The addresses on the BTS side are planned independently. Same as in the layer-2
networking, the length of the subnet mask is determined by the number of BTSs in
the same network segment.
16.2 IP Planning on the BTS Side

The IP address planning of the BTS is related to the IP address planning of the BSC. They
need to be implemented jointly.
Device IP address: logical IP address of the BTS (DevIP)
Port IP address: port IP address of the TMU
The same IP address works for O&M, signaling, and services for the BTS. Generally, use the
port IP address.
2015-11-13
On the BTS side, use port IP communication. One BTS uses one IP address.
In layer-2 networking, the port IP address of the BTS is in the same network segment as
the port IP address on the corresponding interface board on the BSC.
In layer-3 networking, the port IP address of the BTS is in the same network segment as
the port IP address of the next-hop gateway.
Huawei Confidential
Page 88 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
16.3 Routing Design on the BSC Side

It is required in IP layer-3 networking. Routes are added on the interface board. Add a host
route for each BTS, or configure a network segment route for the network segment where the
BTS resides. Configure a network segment route.
16.4 Routing Design on the BTS Side

The principles are as follows:
In layer-3 networking, configure the route from the BTS to the corresponding interface
on the BSC and the route from the BTS to the IP CLK server.
In layer-2 networking, a route is not required.
For IP routes based on the source IP address, the port sending packets and the gateway IP
address are queried based on the source IP address in the IP packets. For static route
configurations, if the routes have the same source IP address, mask, and priority but different
next-hop IP addresses, the static routes form equivalent routes. Packet routes can forward
packets based on the flows or packets, and Huawei radio devices support only flow-based
route forwarding. The eNodeB source IP routes configure different priorities to achieve active
and standby routes. However, route load balancing is not supported. The ADD SRCIPRT
command is used to configure the routing entries.
If only one IP address is available on the eNodeB side, the next-hop IP addresses are the
same, and there are multiple destination IP addresses on the peer end, eNodeB source IP
routes are recommended to simplify the route configuration on the eNodeB side, as shown in
the following figure.
Figure 1.1 Scenario where the eNodeB source IP route is recommended
If there are multiple IP addresses on the eNodeB side, the next-hop IP addresses are
different, and there is only one destination IP address on the peer end, the eNodeB source IP
routes are not recommended, as shown in the following figure.
2015-11-13
Huawei Confidential
Page 89 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 1.2 Scenario where the eNodeB source IP route is not recommended
If there is only one IP address on the eNodeB side, the next-hop IP addresses are
different, and there are multiple destination IP addresses on the peer end, the eNodeB source
IP routes are not supported, as shown in the following figure.
Figure 1.3 Scenario where the eNodeB source IP route is not supported
16.5 VLAN Design

The VLAN technology partitions a physical LAN into different broadcast domains (based on
IEEE 802.1Q). The broadcast traffic in one VLAN is not forwarded to another VLAN. This
prevents broadcast storm.
The VLAN technology increases the security because any two VLANs cannot visit each other
in layer-2. A field in the VLAN tag indicates the priority. This can distinguish priorities in
layer-2.
If the BSC is directly connected to the layer-2 interface of the router:
2015-11-13
Huawei Confidential
Page 90 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Generally, no tag is used if sub-interfaces are not used.

If sub-interfaces are used, VLAN tags are required by default. Sub-interfaces share the MAC
address. The router cannot distinguish flows based on the MAC address or based on the IP
addresses in the received IP packets (the IP address in an IP packet is the destination IP
address but not the source IP address). Therefore, VLAN tags are required to distinguish
flows.
Planning principles:
In principle, the VLAN tags are not added on both the BSC and BTS, but are added on
the intermediate devices.
To make VLAN planning logically clear and simplify the configuration, use the number
of BTSs that are connected to one board or one optical interface of the transmission
device (for example, the OSN) at the level-1 convergence point (nearest to the BTS side)
as the number of BTSs that belong to the same VLAN.
Generally, the proper number of BTSs recommended for one VLAN is 20 to 100.
In VLAN planning, use VLAN values segment by segment. For example, 2G and 3G
services use VLAN values in different ranges, and different interfaces use different
VLAN values. This facilitates follow-up management, maintenance, and expansion.
VLAN tags can be added based on the next hop or based on the service flow. Only
VLAN tagging based on the next hop is supported. VLAN tags can be added on the
transmission device side, as listed in Table 1.1.
Table 1.1 Specifications of the VLAN on the interface board

Single Subrack
VLAN ID
8192
ETH board
FG2c/FG2d
GOUc/GOU
d
EXOUa
VLAN /PORT
4094
4094
4094
VLAN /BOARD
4094
4094
4094
16.6 QoS Design

QoS requirements of interfaces for IP transmission:
Table 1.1 Abis IP bearer network QoS requirement
Abis
IP
2015-11-13
Delay (ms)
Jitter (ms)
Packet Loss Rate

(%)
Suggestio
n Value
Max
Value
Suggesti
on Value
Max
Value
Suggesti
on Value
Max
Value
< 15 ms
< 40 ms
< 8 ms
< 15 ms
< 0.05%
< 0.1%
Huawei Confidential
Page 91 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Table 1.2 A/Gb IP bearer network QoS requirement

A/Gb IP
Delay (ms)
Jitter (ms)
Packet Loss Rate (%)
< 15 ms
< 8 ms
< 0.05%
Table 1.3 IPCLK and BTS/NodeB

Input Clock Precision Requirement
IPCLK
< 0.016 ppm
BTS/NodeB
< 0.05 ppm
QoS design principles:

For details, see the Configuration Recommendation of a specific version.
In GSM IP layer-3 network, different DSCP values are set for different services to ensure the
QoS in IP transmission.
Basic principles:
The priority of signaling is the highest.
The priority of the voice service is lower than that of signaling.
The priority of the real-time data service is lower than that of the voice service.
The priority of the non-real-time data service is lower than that of the real-time data
service.
The IP networks of different operators are different. If the number of DSCP values provided
by an operator is smaller than the number of DSCP values recommended by Huawei, DSCP
convergence can be implemented. For example, use the same DSCP for the voice service and
the real-time data service.
2015-11-13
Huawei Confidential
Page 92 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
17
Network Topology Design

The network topology design is based on the unit of area (or city). Within the unit, the
topology details are designed. This is different from the design of the entire target
network.
Design a detailed network topology figure that can help learn the connections between
NEs, transmission type, and geographical location and can guide engineers through
network construction.
Obtain information about the capacities, locations, and manufacturers of the interface
NEs related to the MSC server, MGW, and SGSN, and analyze the possible networking
risks, for example, whether the capacities of the MSC server and SGSN meet the
requirements.
Design the network topology to optimize the resource usage and reduce the invalid load.
Information, such as geographical distribution and the number of NEs, area names,
equipment room names, and NE types
Subscriber transmission information (transmission type, transmission quantity, and

transmission distribution)
17.2 Network Structure Design

17.2.1 Design Guide
The network structure design is based on the unit of area. The input is the pre-sales target
network. Details are designed. In the design, consider the geographical distribution of NEs,
location relationship between the MSC server, MGW, BSC, and BTS, transmission types, and
backbone transmission network.
Design principles:
2015-11-13
Huawei Confidential
Page 93 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The transportation to the equipment room is convenient, and this facilitates maintenance.
Select the equipment room where the power supply is stable and air conditioners/ESD
floor are available.
Consider the distribution of existing transmission nodes when designing the BSC
locations. Use the existing transmission nodes to save the investment.
Abis transmission cost control is important for the network where the number of sites is
large. Ensure that the BSC location and Abis networking are good for transmission
convergence.
Group the sites under the BSC based on geographical areas to facilitate LAC planning.
Plan the BSCs in the same area in the same MSC server. Avoid discontinuous BSC
networking because it increases inter-MSC handovers, inter-MSC signaling load, and the
configuration difficulty, and reduces the handover success ratio.
To improve the network security, connect one BSC to more than two MGWs that belong
to the same MSC server (non-MSC Pool mode).
The SGSN capacity meets the data service requirements of the corresponding PCU. The
networking design needs to collect the information about the capacity and manufacturer
of the SGSN connected to the PCU.
Drawing requirements:
The networking diagram must be drawn based on the standard radio icons.
The networking diagram must include the MSC server, MGW, BSC, and BTS and must
clearly show the connections between the NEs.
The networking diagram must clearly mark the geographical homing areas of the NEs.
Different types of lines must be used in the networking diagram to indicate different
transmission types.
17.2.2 Typical Networking

17.2.2.1 BSC/MGW Single-Homing Networking
Generally, one BSC is connected to one MGW and belongs to one MSC server. See Figure
1.1.
Figure 1.1 Networking of the BSC connected to a single MGW
2015-11-13
Huawei Confidential
Page 94 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
17.2.2.2 BSC/MGW Multi-Homing Networking

If two or more MGWs are connected to the same MSC server, to improve the network
reliability, assign the A interface transmission and signaling to different MGWs. Then, even if
one MGW is faulty, the BSC can provide service continuously. See Figure 1.1.
If the MGW does not possess a signaling point and the SS7 signaling is transparently
transmitted to the MSC server, you do not need to configure the signaling transfer point (STP)
for SS7 on the BSC, and you need only to specify the STP of the MSC server as the
destination STP.
If the MGW possesses signaling points, and the negotiation with the core network indicates
that the STP needs to be configured for the BSC, configure multiple STPs (signaling points of
the connected MGW) for the BSC.
BSC/MGW multi-homing requires the support of the core network and does not have
requirements for the BSC. To use this networking mode, confirm that the core network
supports this mode.
Advantage: The reliability is high and the networking is simple. This facilitates follow-up
expansion and maintenance.
Figure 1.1 BSC/MGW multi-homing networking
17.2.2.3 MSC Pool Networking

Different from traditional networking, in MSC Pool networking, one RNC/BSC can connect
to multiple MSC servers at the same time. This implements the MSC-level backup
mechanism. When one MSC server is faulty, the services of the BSC are automatically
switched to another MSC server. This significantly improves the network reliability. After IP
transmission is widely used, the MSC Pool planning is simplified a lot. In the past, physical
interconnection is required; nowadays, routing configuration is used. It is predicted that the
MSC Pool networking will be widely used after 2011.
Advantage: The reliability is high.
2015-11-13
Huawei Confidential
Page 95 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Disadvantage: The networking on the core network is complicated, and the design and
implementation workload is heavy.
MSC Pool networking mode 1
The MGW implements the A-Flex function.
The MGW implements the NNSF function.
The BSC configuration does not need to be changed. For the BSC, the MSC servers in
the pool work as a large-capacity MSC server.
Use this mode. The BSC does not require special configuration. In addition, the reliability of
the entire network is significantly improved.
Figure 1.1 MSC Pool networking mode 1
MSC Pool networking mode 2
2015-11-13
The BSC implements the NNSF function.
The BSCs are connected to multiple MGWs.
The networking and expansion are complicated, and do not use this mode.
Huawei Confidential
Page 96 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
17.2.2.4 SGSN Pool Networking

The SGSN pool is based on the Gb Flex technology. The SGSN pool area contains the area
served by a group of SGSN nodes. In the pool area, multiple SGSNs run concurrently and
share the service traffic in the pool area. If a single SGSN node is faulty, another SGSN in the
pool can provide services. This significantly improves the network reliability.
Figure 1.1 shows the typical networking of the SGSN pool based on Gb over IP:
2015-11-13
Huawei Confidential
Page 97 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 1.1 Typical networking of the SGSN pool
17.2.2.5 All-IP Networking

The all-IP networking saves the transmission cost a lot and provides better evolution
capability to implement GU and GUL evolution. Generally, use layer-3 networking for all-IP
networking. Figure 1.1 and Figure 1.2 show the typical networking diagrams.
2015-11-13
Huawei Confidential
Page 98 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 1.1 All-IP networking
Figure 1.2 Typical IP-based networking
17.2.2.6 Hybrid Networking

The common networking mode is A interface over IP+Abis interface over TDM, as shown in
Figure 1.1.
2015-11-13
Huawei Confidential
Page 99 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Hybrid networking
17.2.2.7 Transmission Resource Pool Networking

If multiple A interface boards of the BSC form a transmission resource pool, the MGW can
communicate with the BSC through any board in the pool. If an interface board in the pool is
overloaded, the system automatically allocates the newly connected calls to other interface
boards. In addition, the BSC implements the source IP-based route and IP-path-free
configuration, in other words, if IP addresses on the MGW are added or modified, IP route
and IP path configurations on the BSC do not need to be changed, thereby simplifying the
configuration.
The SCTP links on the control plane are deployed on two different interface boards in the
transmission resource pool by dual-homing to improve data protection on the control plane.
The transmission resource pool over the A interface can work in the following modes:
Mode 1: The A interface boards work in active/standby mode.
Mode 2: The A interface boards work independently.
Mode 1 provides high reliability. When A interface boards are faulty, ongoing calls are not
dropped. Mode 2 provides high board usage, but ongoing calls are disconnected once an
interface board is faulty.
This solution requires a three-layer networking between the BSC and the MGW to ensure
interconnection between the MGW and all interface boards of the BSC.
For details, see A Interface Transmission Pool. The transmission resource pool is used on the
A interface over IP network. Figure 1.1 shows the logical networking of the transmission
resource pool.
2015-11-13
Huawei Confidential
Page 100 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Logical networking of the transmission resource pool
As shown in Figure 1.2, two pairs of active/standby FG2c/GOUc boards form a pool. Each
board is configured with a logical board IP (DEVIP) address as the service IP address. These
device IP addresses form a service IP address pool. During a call setup, the BSC selects an IP
address from the service IP address pool to carry the call in a way that ensures load balance.
Figure 1.2 Physical networking of the transmission pool with active/standby boards
As shown in Figure 1.3, four independent FG2c/GOUc boards form a pool. Each board is
configured with a logical board IP (DEVIP) address as the service IP address to comprise a
2015-11-13
Huawei Confidential
Page 101 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
service IP address pool. During a call setup, the BSC selects an IP address from the service IP
address pool to carry the call in a way that ensures load balance. Use this networking.
Figure 1.3 Physical networking of the transmission pool with independent boards
In this networking mode, configure two or more ports for an interface board to form an LAG
to avoid ongoing call drop in case of a port fault. Compared with the networking of
transmission pool with active/standby boards, this networking has the following advantages
and disadvantages:
2015-11-13
Advantages
The maximum payload throughput rate and the connection capability of the interface
board are doubled.
There are no restrictions on the slot configuration of the active/standby board.

Therefore, the interface boards can be evenly distributed to each frame.
The interface board can be independently added.
The router configuration is simple, that is, no static routes with different levels of
priorities are required.
Disadvantages: if the interface board is faulty, ongoing calls are dropped.
Huawei Confidential
Page 102 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
18
Reliability Design

The design is based on product features and transmission routes to improve the network
reliability.
Design a detailed network topology figure that can help learn about the connections
between NEs, transmission type, and geographical location.
Network structure diagram
Transmission routing information
18.2 Network Reliability Design

18.2.1 Design Guide
The reliability design of the radio network is based on the following:
Active/standby feature of boards
Active/standby feature of links
Active/standby feature of transmission resources
MSC pool
SGSN pool
Flow control policy
Device load balancing
Proper data configuration
Design principles:
2015-11-13
1.
Implement the network reliability design based on the board active/standby feature of the
product. For example, board active/standby feature and link active/standby feature of the
A interface, Abis interface, and Gb interface.
2.
Implement port protection on the transmission device to improve reliability. For

example, if you need to improve reliability but do not want to increase investment in
backbone transmission, you can implement transmission port backup by using the
timeslot cross device on designed backbone transmission interfaces.
Huawei Confidential
Page 103 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
3.
Design the networking structure to improve the network reliability. Use BSC/MGW
cross networking, MSC pool networking, and SGSN pool networking.
4.
Distribute transmission resources properly and use proper configuration principles to

improve the network reliability. For example, distribute signaling and calls evenly to
different routes to reduce node cross-connection and avoid concurrent service
interruption.
5.
Learn the port distribution of the remote device. Distribute the transmission resources of
the same BSC to different interface boards to improve the network reliability.
18.2.2 Design Examples

18.2.2.1 Reliability of Active/standby Links on Ports
Based on product features, active/standby links and active/standby boards of the A interface,
Gb interface, and Abis interface can be implemented to improve the network reliability.
If the core network supports the active/standby configuration of the A interface, use the
active/standby configuration function of the A interface of the BSC6910.
If the core network does not support the active/standby configuration of the A interface, the A
interface board of the BSC6910 is configured to work in standalone mode.
For details about APS 1+1, see Figure 1.1.
Figure 1.1 Improving reliability by active/standby links on ports
2015-11-13
Huawei Confidential
Page 104 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.2 Reliability design of the Gb interface
18.2.2.2 Reliability of Load Balancing

Evenly distribute BTSs, TRXs, and SS7 links to the modules of the BSC to balance the load
between the modules. Balanced load can improve the anti-shock capability of the BSC in the
case of burst traffic. The principles are as follows:
Evenly distribute BTSs and TRXs to the modules of the BSC.
Evenly distribute SS7 links (with the same bandwidth) to the modules.
If the traffic of a single BSC reaches 70% of the design specifications, the balance of
SS7 links, BTSs, and RSI links on different CPUs is adjusted to further design the load
balancing. You can apply for R&D support.
18.2.2.3 Reliability of Data Configuration
Evenly distribute SS7 links and bandwidth to the modules of the BSC. The minimum
number of SS7 links is 2.
Configure SS7 links to different transmission resources and boards to reduce node crossconnection and improve the network reliability.
For example, configure SS7 links to different MGW boards, A interface boards, A
interface physical transmission channels, and backbone transmission channels.
18.2.2.4 Reliability of Multiple Transmission Routes

Configure transmission routes to improve the reliability. Configure the transmission resources
of the same NE to different transmission nodes and routes to reduce the possibility of service
interruption caused by the faults in a single transmission node and to improve the network
reliability. As shown in Figure 1.1, the GOUc of the BSC6910 provides four physical
transmission links destined for the remote NEs through two CEs. This implements reliability
of multiple transmission links and prevents service interruption caused by the faults in a
single transmission node.
2015-11-13
Huawei Confidential
Page 105 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Reliability design of IP transmission routes
18.2.2.5 Reliability of the IP Networking VRRP Technology

As shown in Figure 1.1, in Gb over IP mode, the VRRP technology is used to implement the
active/standby mechanism of the routers AR1 and AR2, and the active/standby Gb interface
boards are configured for the BSC. In this manner, a reliable transmission route is established
to improve the network reliability.
Figure 1.1 Reliability design of IP transmission routes
The VRRP protocol is used to dynamically select a master router from a group of VRRP
routers and associate the router to a virtual router as the default gateway of the connected
network segment.
The VRRP router that is selected to associate with the IP address of a virtual router is the
master router. The master router forwards the packets destined for the virtual router.
If the master router is faulty, VRRP selects another VRRP router as the master router to
forward the packets destined for the virtual router.
2015-11-13
Huawei Confidential
Page 106 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The advantage of VRRP is that the reliability of the default gateway is improved for the host.
In VRRP configuration, trunks must be configured between routers. At least two physical links must be
configured, and the bandwidth needs to be twice higher than the service traffic.
18.2.2.6 SCTP Multi-Homing Design

In A over IP mode, the M3UA SCTP multi-homing design implements highly reliable SS7
design. On the GBSS side, only the A interface supports the SCTP configuration. For details,
see section 19.2.3"SCTP Multi-Homing Design."
18.2.2.7 Reliability of BSC Multi-Homing (Connected to Multiple

MGWs)
The BSC6910 supports the configuration of multiple DSPs and STPs. The transmission links
of the A interface of the BSC are connected to two or more MGWs (the MGWs belong to the
same MSC server). Then, if a single MGW is faulty, the services of the BSC are not
interrupted. This improves the BSC reliability. Figure 1.1 shows the typical networking.
18.2.2.8 Reliability Based on the MSC POOL

For details, see the GSM MSC Pool Network Design Guide.
In the traditional mobile communication network, one BSC is connected only to one MSC
server. In the MSC pool network, one BSC can be connected to multiple MSC servers.
Compared with the traditional network, the MSC pool network has the following advantages:
2015-11-13
Load sharing: Multiple MSC servers share the network load to improve the resource
usage of the entire core network and save the device investment.
Disaster tolerance: The MSC-level redundancy is implemented. If a single MSC server is

faulty, the services of the BSC are not interrupted.
Huawei Confidential
Page 107 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Inter-office location updates are reduced, the signaling traffic on the C/D interface is
reduced, and the MSC capacity gain is obtained.
Inter-office handovers are reduced, and the quality of subscriber calls is improved.
In the office that requires high reliability, use the MSC pool solution. The commonly used
networking solution is as follows:
The MGW implements the A-Flex function and NNSF function. The BSC configuration does
not need to be changed. For the BSC, the MSC servers in the pool work as a large-capacity
MSC server. Use this MSC pool networking mode because no special BSC configuration is
required and the reliability of the entire network is improved.
The BSC provides the NNSF function. BSCs are interconnected with multiple MGWs.
Networking and capacity expansion are complicated. Use this networking mode for the IP
networking.
2015-11-13
Huawei Confidential
Page 108 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
18.2.2.9 Reliability Based on the SGSN Pool

The SGSN pool is based on the Iu Flex technology. The SGSN pool area contains the area
served by a group of SGSN nodes. In the pool area, multiple SGSNs run concurrently and
share the service traffic in the pool area. If a single SGSN node is faulty, another SGSN in the
pool can provide services. This improves the network reliability.
For details, see the GSM SGSN Pool Network Design Guide.
This solution can be used based only on Gb over IP.
Figure 1.1 shows the typical networking diagram.
2015-11-13
Huawei Confidential
Page 109 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Typical networking diagram of the SGSN pool
18.2.2.10 Reliability Based on Dynamic Load Balancing of A

Interface Boards
Figure 1.1 shows the IP networking topology of A interface boards based on the dynamic
loading balancing.
Figure 1.1 IP networking topology of A interface boards based on the dynamic loading balancing
If multiple A interface boards of the BSC form a transmission resource pool, the MGW can
communicate with the BSC through any board in the pool. When the load of an interface
board in the pool is high, the BSS automatically distributes calls to interface boards with light
traffic. In addition, the BSC implements the free configurations on IP routes and IP paths.
That is, when the IP address is added or modified on the MGW, the BSC does not need to
change the configuration of the corresponding IP path or IP route. This simplifies the
configurations.
2015-11-13
Huawei Confidential
Page 110 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The SCTP links on the control plane are deployed on two different interface boards in the
transmission resource pool by dual-homing.
The transmission resource pool over the A interface can work in the following modes: Mode
1: The A interface boards work in active/standby mode. Mode 2: The A interface boards work
independently. Mode 1 provides high reliability. When A interface boards are faulty, ongoing
calls are not dropped. Mode 2 provides high board usage, but ongoing calls are disconnected
once an interface board is faulty.
This solution requires a three-layer networking between the BSC and the MGW to ensure
interconnection between the MGW and all interface boards of the BSC.
OM Reliability
The OM reliability design of the BSS is implemented through the active/standby EOMU
configurations and the active/standby EOMU port configurations. Figure 1.2 and Figure 1.3
show the typical networking modes.
Figure 1.2 Standalone EOMU
2015-11-13
Huawei Confidential
Page 111 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.3 Dual EOMUs
Clock Reliability
Clock reliability of the BSC6910
The design of the clock system reliability of the BSC6910 is simple. This section describes
the principle and reliability mechanism of the clock system for better communication with
operators.
GCUa/GCGa boards provide clock information for the BSC6910 system. They are located at
slots 14 and 15 of the GMPS subrack to form the active/standby relationship. Either GCUa
boards or GCGa boards can be installed on the BSC6910 based on the clock type. GCGa
boards have satellite cards and can obtain signals of GPS clock sources. On an all-IP network,
the BSC does not need a clock.
Clock sources of the BSC6910 system are as follows:
Building Integrated Timing Supply System (BITS) clock
External 8 kHz clock
GPS clock
The GCUa/GCGa boards support plane input of the clock sources.

Clock signals provided by the BITS are transmitted to the GCUa/GCGa boards through panel
interfaces.
Figure 1.4 shows the clock subsystem of the BSC6910.
2015-11-13
Huawei Confidential
Page 112 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.4 Clock subsystem of the BSC6910
Clock information generated by the GCUa/GCGa boards is processed as follows:

The clock information is sent to the SCUb board of the local subrack through the backplane
and then the SCUb board of the local subrack sends the clock information to service boards in
the subrack.
The clock information is sent to SCUb boards of GEPS subracks through plane interfaces and
then the GSCU boards of the GEPS subracks send the clock information to service boards in
the subrack.
The clock system of the BSC6910 has the following characteristics:
2015-11-13
Clock holdover function: When external clocks are faulty, local free-run clocks can
continue providing stable clocks for the system.
The clock system of the BSC adopts the digital phase-locked loops and reliable software
phase-locked loops to ensure that the clock of the BSC is synchronized with clock
reference sources.
The BSC clock adopts international standard layer-3 clocks to provide stable and reliable
clock sources for the system.
The clock system provides complete display, alarm, and operation and maintenance
functions. You can set internal parameters of the clock on the maintenance terminal.
If configured clock reference sources are lost, the system traces other available reference
sources. If no clock sources are available and the system has traced other reference
sources for 10 minutes, the system is in the holdover state. If the system does not detect
that frequency deviation is great during this process, software phase-locked loops are
Huawei Confidential
Page 113 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
retained for 10 days and then the system is in the free-run state. Although the system
provides high-precision holdover function, quick restoration of references sources is also
required. In the synchronization networking (non-IP clock sources) mode, services are
affected due to the free-run state. Therefore, time specifications of affected services
cannot be determined.
Reliability of Ethernet Link Aggregation

In link aggregation, multiple physical links are aggregated to form a logical link. Multiple
links are aggregated to an LAG.
The Ethernet link aggregation can increase the bandwidth, improve transmission reliability,
and ensures that traffic is distributed to different links for transmission. When an Ethernet port
in the LAG is faulty, services are not interrupted.
The Ethernet link aggregation implemented by Huawei aims to:
Replace the active/standby port mode. The external networking of the manual
active/standby LAG mode and the active/standby port mode are the same. The manual
active/standby LAG networking mode apples to more scenarios and is more convenient
to interwork with peer equipment. It supports flexible networking mode. For example,
the active/standby LAG can be located on one board, but active/standby ports must be on
active/standard boards respectively.
Reliability and check mechanism for the manual active/standby LAG networking mode
and the active/standby port networking mode are the same.
If the active/standby board has multiple ports, enable multiple manual active/standby
LAGs, which is the same as that of multiple active/standby ports. The local equipment
does not need to interwork with peer equipment through the static LACP.
The Ethernet link aggregation matches the MRFD-210103 Link aggregation feature.
Application of the link aggregation
Typical application scenarios of link aggregation on the BSC side are as follows:
BSC belonging to two layer-2 transmission devices in the dual-homing mode
BSC belonging to one layer-2 transmission device in the single-homing mode
Inter-board link aggregation in the inter-board pool networking scenario
Link aggregation on the BSC side and the interworking router adopting the VRRP
networking mode
BSC belonging to two layer-2 transmission devices in the dual-homing mode

As shown in Figure 1.5, two FE/GE links on the active/standby board of the BSC form an
LAG and are respectively connected to two transmission devices, such as the Optical Switch
Node (OSN)/Packet Transport Network (PTN) device. This LAG crosses two transmission
devices.
2015-11-13
Huawei Confidential
Page 114 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.5 BSC belonging to two layer-2 transmission devices in the dual-homing mode
The LAG that crosses two transmission devices is named Multi-Chassis Link Aggregation
Group (MC-LAG) on the network side. Currently, the MC-LAG works in only the
active/standby mode. Therefore, the working mode of the aggregation group on the BSC side
must be set to the active/standby mode. In addition, the aggregation mode on the BSC side
needs to be set to static aggregation. This ensures that the active/standby properties of links
that transmission devices use to interwork with the BSC through the LACP protocol are
consistent.
BSC belonging to one layer-2 transmission device in the single-homing mode
As shown in Figure 1.6, multiple FE/GE links on the active/standby board of the BSC form an
aggregation group and are connected to one layer-2 transmission device.
Figure 1.6 BSC belonging to one layer-2 transmission device in the single-homing mode
In this application:
2015-11-13
When an aggregation group works in the active/standby mode, the aggregation group
ensures that links are reliable and the active/standby board on the BSC side ensures that
boards are reliable.
When an aggregation group works in the load sharing mode, the aggregation group not
only ensures that links/ports/boards are reliable but also expands the capacity of the link
bandwidth.
Huawei Confidential
Page 115 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Inter-board link aggregation in the inter-board pool networking scenario

As shown in Figure 1.7:
Multiple interface boards form the transmission pool to implement the load balancing
between interface boards.
Multiple FE/GE link groups on interface boards form an aggregation group to implement
the load balancing among multiple interfaces on boards. In this scenario, the aggregation
group needs to work in the load balancing mode.
Figure 1.7 Inter-board link aggregation in the inter-board pool networking scenario
Manual active/standby link aggregation on the BSC side and the interworking router
adopting the VRRP networking mode
As shown in Figure 1.8:
Two FE/GE links on a board on the BSC side are configured into an aggregation group
that connects to two routers.
The aggregating mode of the aggregation group must be manual mode. The working
mode of the aggregation group must be active/standby mode.
The two routers adopt the VRRP to ensure reliability.
On primary links of an aggregation group, enable the BFD check for two actual port IP
addresses of the VRRP device and the ARP check for virtual IP address of the VRRP
device.
On secondary links of an aggregation group, enable the ARP check for virtual IP address
of the VRRP device.
Link aggregation on the BSC side and the interworking router adopting the VRRP
networking mode
2015-11-13
Huawei Confidential
Page 116 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.8 Manual active/standby LAGs on the BSC side+router adopting the VRRP networking
mode
Link aggregation on the BSC side has the following two scenarios:
Two FE/GE links on a board form a link aggregation group and interwork with the
VRRP device. Multiple independent boards share the load. This networking mode
mainly applies to the A interface and GOUa/FG2a boards are used interface boards.
Interface boards form a pool. Two FE/GE links on the interface boards form a link
aggregation group. This implements the load sharing for the interface boards and ensures
reliability.
GOUc/FG2c boards support the transmission resource pool networking mode. Therefore, use
the transmission resource pool networking mode.
Two FE/GE links on the active/standby board form an aggregation group that connects to
VRRP devices.
As shown in Figure 1.9, two FE/GE links on the active/standby board form an aggregation
group that connects to VRRP devices.
Two FE/GE links on the active/standby board form an aggregation group that connects to
VRRP devices.
Figure 1.9 LAG of the active/standby board+router adopting the VRRP networking mode
2015-11-13
Huawei Confidential
Page 117 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
19
Transmission Interface
Design

Properly design the interface networking solution based on the NE geographical

locations and transmission conditions to ensure the reliability and save resources.
Calculate the interface bandwidth and plan the number of transmission links based on the
traffic model and transmission type.
Negotiate interface interworking parameters for follow-up interface interworking to

improve the interworking success ratio.
Specify interface configuration principles to guide LLD design and implementation.
Device BOQ
Networking diagram of the device
Terms about transmission resources in the contract
Transmission type and interface protocol
Feature function application, including the MSC Pool.
19.2 A Interface Design

19.2.1 Interface Description
The A interface is between the BSS and the MSC server, and it implements the interworking
of the products provided by multiple manufacturers.
If A over IP is in use, the core network (CN) side uses the softswitch architecture, the control
plane uses the SIGTRAN protocol, and the user plane uses the RTP protocol.
Figure 1.1 and Figure 1.2 describe the bearer protocol on the A interface.
2015-11-13
Huawei Confidential
Page 118 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Reference protocol model on the control plane of the A interface
Figure 1.2 Reference protocol model on the user plane of the A interface
IP transmission on the A interface allows an operator to construct the IP network between the
BSC and the MSC server. The A interface protocol is the standardized A interface protocol,
and therefore can interwork only with the softswitch devices that also use the standardized A
interface protocol. Therefore, in the interworking with the MSC server, confirm the A
interface protocol supported by the MSC server.
IP transmission on the A interface provides two types of interfaces: FE and GE. The supported
protocol is IPv4. The BSC and the MSC server can be connected through a LAN or WAN
based on the locations of the BSC and the MSC server. The networking mode can be direct
connection or router-based networking. Layer-3 router-based networking is preferred.
Because the A interface supports only the IP bearer, the BSS does not provide the TC
function, and the MGW provides the TC function. Huawei's BSS expands the A interface
protocol to support the TrFO function to reduce the number of coding times and improve the
voice quality.
19.2.2 Networking Design

Design principles:
2015-11-13
The transportation to the equipment room is convenient, and this facilitates maintenance.
Select the equipment room where the power supply is stable and air conditioners/ESD
floor are available.
Consider the distribution of existing transmission nodes when designing the BSC
locations. Use the existing transmission nodes to save the investment.
The A interface supports only the IP bearer and the MGW, instead of the BSS, provides
the TC function.
Huawei Confidential
Page 119 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Plan the BSCs in the same area in the same MSC server. Avoid discontinuous BSC
networking because it increases inter-MSC handovers, inter-MSC signaling load, and the
configuration difficulty, and reduces the handover success ratio.
To improve the network security, connect one BSC to more than two MGWs that belong
to the same MSC server (non-MSC Pool mode).
In BSC/MGW multi-homing mode, the transmission links of the A interface are connected to
two or more MGWs (the MGWs belong to the same MSC server). Then, if a single MGW is
faulty, the services of the BSC are not interrupted. This improves the BSC reliability.
19.2.2.2 Networking Design of IP Transmission on the A

Interface
After an A over IP construction, the TC function is provided by the MGW. Direct connection
networking, layer-2 networking, and layer-3 networking can be used. Generally, layer-3
networking is preferred.
Design guide:
Step 1 Determine the networking mode based on the networking scale and the requirements of the
customer for A interface networking.
Step 2 Design the networking reliability of the interface board based on the support capability of the
interworking device and the requirements of the customer, that is, determine whether interboard active/standby, load sharing, route-based active/standby in standalone mode, or routebased load sharing in standalone mode is used. The inter-board active/standby networking
mode is preferred.
Step 3 If layer-3 networking is in use, design the networking reliability of the router, that is,
determine whether to use VRRP-based route active/standby networking.
Step 4 Determine whether the signaling and service are separated on the A interface on the bearer
network. The optional modes are as follows:
2015-11-13
The signaling and service share the same port. (Recommended)
Huawei Confidential
Page 120 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
If the customer requires the signaling and service to be separated, use different ports for
the signaling and service. Then, they are mapped to different virtual private networks
(VPNs) based on different ports. In this way, signaling and service separation on the
bearer network is implemented. (Recommended)
Configure multiple IP addresses for the physical port. Then, the signaling and service are
mapped to different VPNs based on different IP addresses. In this way, signaling and
service separation on the bearer network is implemented. (Not recommended)
----End
Figure 1.1 shows the typical networking.
Figure 1.1 Typical A over IP networking mode (pool of standalone boards)
On the bearer network side: Layer-3 router networking is in use and a pair of independent
routers is deployed.
On the BSC side: Independent interface boards are used, service logical IP addresses form an
IP pool, and pooled interface boards work in load-sharing mode.
IP route: The next hop is the IP address of the router.
The BSC does not require the configuration of IP path. It requires the configuration of a local
end IP pool, instead of the configuration of peer IP addresses.
Detection mechanism
After a transmission resource pool is deployed on the A interface, pooled interface boards in
multiple pairs and pooled ports in multiple pairs protect each other.
IP pool fault detection and switchover triggering mechanism: Each address in the IP pool on
the BSC side automatically starts the ICMP ping detection. If the ICMP ping detection of a
pooled IP address fails, services on the faulty links are allocated to other pooled IP addresses.
Direct connection networking without routers (not recommended)

Advantages:
2015-11-13
If no datacom device is used, the reliability is improved, the network construction

cost is low, and the maintenance is simple.
Huawei Confidential
Page 121 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The datacom device is uncontrollable in some aspects. If no datacom device is used,

the QoS can be easily guaranteed to facilitate bandwidth call access control.
The compatibility problem can be prevented in the interworking with the devices
provided by other manufacturers.
Disadvantages:
The networking is not open.
It is not applicable to large-scale networking.
Follow-up expansion is inconvenient, and the evolution capability is low.
Router networking (recommended)

Advantages:
This solution provides a high bandwidth and reliable transmission bearer for the A
interface.
This solution meets the requirements for the transmission bearer network in GSM
network evolution.
This solution can protect the network from burst data services.
The networking is open and supports large-scale networking.
Disadvantages:
The best effort feature of the IP network causes low QoS, and the end-to-end QoS
mechanism is required to ensure the QoS.
In this solution, the operator must provide an IP bearer network (that consists of
devices such as routers) as the transmission bearer network. This means that certain
investment is required.
Networking reliability design

Analyze and design the networking scenario based on the requirements of the customer and
the project condition. Use either or both of the direct connection networking mode (without a
router) and the router networking mode. The router networking mode is preferred. After you
determine the networking mode, design the networking reliability.
If A over IP (FE/GE) is in use, the active/standby and load-sharing modes are supported for
the reliability of A interface networking.
In addition, if only one interface board is used (that is, the standalone mode), port
active/standby or port load-sharing can be implemented on the interface board using route
configuration on condition that the interface adopts the device IP communication mode.
Intra-board port active/standby: Configure active and standby routes (with different priorities)
from different ports to the same destination address to implement route-based port
active/standby within the board. (Not recommended)
Intra-board port load sharing: Configure equivalent routes (with the same priority) from
different ports to the same destination address to implement route-based port load sharing
within the board. (Not recommended)
The requirements of the customer and the capability of the interworking device determine the
networking reliability mode on the A interface (the active/standby mode, load-sharing mode,
route-based active/standby in standalone mode, or route-based load sharing in standalone
mode). Adopt the mode recommended by Huawei.
Promoted networking schemes for the GBSS15.0 BSC6910:
2015-11-13
Promoted scheme 1: using a transmission resource pool
Huawei Confidential
Page 122 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Networking: Pool of standalone boards

Interface boards on the BSC side are standalone and are divided into two groups to
connect to two routers.
Logical IP addresses (such as IP170 and IP180 in Figure 1.2) of multiple interface boards
of the BSC form an IP pool, and loads are evenly distributed among the pooled interface
boards.
Each SCTP link on the control plane is configured with inter-board dual-homing
protection (two IP addresses such as IP150 and IP160 in Figure 1.2).
The BSC does not require the configuration of IP path. It requires the configuration of a
local end IP pool, instead of the configuration of peer IP addresses.
Packets on the control plane and the service plane are transmitted and received through
the port on the active board.
Figure 1.2 Typical A over IP networking mode (pool of standalone boards)
On the bearer network side: Layer-3 router networking is in use and a pair of
independent routers is deployed.
On the BSC side: Interface boards adopt the IP pool comprised by logical IP addresses of
the independent service plane of the board. Device IP addresses are used for
communication.
Route configuration examples
Devic Source
e
IP
Next
Hop
BSC
IP150
IP110
IP170
IP130
IP160
IP120
IP180
IP140
Standby Next
Hop
Detection mechanism
2015-11-13
Huawei Confidential
Page 123 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
After a transmission resource pool is deployed on the A interface, pooled interface

boards in multiple pairs and pooled ports in multiple pairs protect each other.
IP pool fault detection and switchover triggering mechanism: Each address in the IP pool
on the BSC side automatically starts the ICMP ping detection. If the ICMP ping
detection of a pooled IP address fails, services on the faulty links are allocated to other
pooled IP addresses.
Analysis of the fault switchover mechanism (only on single-fault scenarios)
Fault 1: On the BSC side, the interface board on which IP170 is configured is faulty.
On the control plane: Packet receiving by the SCTP links times out and the packets are
resent through the standby path. Upper layer signaling indicates that no packets are lost
but a time delay of about 1s occurs. The SCTP working links switch to the standby path
about 15s after the packets are resent.
On the user plane: ICMP ping detection on IP170 in the IP pool fails (5 x 5s). IP170 is
blocked and subsequent services are provided through IP180.
Optional scheme 1: using a transmission resource pool

Networking: Pool of active/standby interface boards+dual-active ports
Logical IP addresses on the service plane of the active/standby interface boards on the
BSC side form a pool. The BSC is directly connected to the dual routers through two
independent ports on the active/standby interface boards. The pooled interface boards
work in load-sharing mode.
Device IP addresses are configured only on the logical active board, that is, IP150,
IP160, IP170, and IP180 are configured only on the active board. With the active/standby
MGW features of the source IP enabled, the active/standby paths are bound to the
outgoing ports of the active/standby boards to achieve active/standby routes of the
active/standby boards. Besides, the outgoing port routes of the active/standby boards are
configured and active/standby routes are configured for routers, so that the ports of the
active/standby boards can protect each other.
SCTP links must be configured to inter-board dual-homing mode to provide inter-board
protection in the IP pool.
Local IP addresses instead of IP paths are configured on the BSC.
Packets on the control plane and the service plane are transmitted and received through
the port on the active board.
2015-11-13
Huawei Confidential
Page 124 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.3 Typical A over IP networking mode (pool of active/standby interface boards+dualactive ports)
In actual application, IP addresses of more than two boards form an IP pool.
On the BSC side: Interface boards adopt the IP pool comprised by logical IP addresses of
the active/standby service planes of the board. Device IP addresses are used for
communication.
Devic Source
e
IP
Next
Hop
Standby Next
Hop
BSC
IP150
IP110
IP120
IP170
IP130
IP140
IP160
IP120
IP110
IP180
IP140
IP130
Detection mechanism
The active port of each board enables two BFD sessions to detect the IP addresses of the
two routers. The BSC performs a BFD detection every 100 ms for three times. Configure
the delay enabling BFD on CE1 and CE2 to avoid service interruption of CE1 and CE2
due to a reset upon power-off. If BFD is deployed between interface boards and the peer
routers, the BSC triggers a switchover of the active/standby gateways in the source IP
address routing table. The BSC can detect board faults and if it detects a board fault, the
2015-11-13
Huawei Confidential
Page 125 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
active/standby boards of the BSC switch over, and the logical IP address is migrated to a
normal board from the faulty board. Accordingly, the source IP address route mapping
the logical IP address is switched between the active/standby gateways.
Fault 1: On the BSC side, the interface board on which IP170 is configured is faulty.
On the control plane: Packet receiving by the SCTP links times out and the packets are
resent through the standby path. Upper layer signaling indicates that no packets are lost
but a time delay of about 1s occurs. The SCTP working links switch to the standby path
about 15s after the packets are resent. The destination is reachable all the time.
On the user plane: ICMP ping detection on IP170 in the IP pool fails (5 x 5s). IP170 is
blocked and subsequent services are provided through IP180.
Optional scheme 2: using a transmission resource pool

Networking: Pool of active/standby boards+manual active/standby LAGs
The BSC is directly connected to the dual routers through the active/standby ports on the
active/standby interface boards. All data is sent and received through the active port.
VRRP IP addresses are configured between the dual routers, which function as the next
hops of the BSC. Heartbeat messages are transmitted over the trunk between CE1 and
CE2. Configure multiple VRRP IP addresses between CE1 and CE2 to share loads. As
shown in 19.2.7.1.1Step 5Figure 1.1, CE1 functions as the high priority router of VRRP1
and CE2 functions as the high priority router of VRRP2.
Logical IP addresses of multiple pairs of active/standby interface boards of the BSC form
an IP pool. Multiples pairs of active/standby ports of a pair of active/standby boards form
a pool. In this networking, because the ports on the interface boards of the BSC side do
not support adding ports to the boards of the LAG, new IP addresses must be configured
and added to the pool. New IP addresses and new VRRP IP addresses must be also
configured on peer devices.
The BSC, MSC server, and the MGW are deployed in layer-3 networking mode. The
MSC server and the MGW connect to the BSC in loose coupling mode and the BSC can
be configured to work in either active/standby mode or load-sharing mode. If an IP pool
is deployed on the BSC side, contact peer maintenance engineers to configure the peer
device to work in load-sharing mode, which is consistent with the IP pool on the BSC
side.
The control plane and the user plane use the same physical port. However, the bearer
network adopts different VPN isolation for the control plane and the user plane of the A
interface. Therefore, different VLANs and ETHIPs must be designed on the BSC side
(configure separate ETHIP for the control plane on one pair of interface boards: IP111
and the mapping VLAN, and separate ETHIP for the user plane: IP131 and the mapping
VLAN; configure separate ETHIP for the control plane on the other pair of interface
boards: IP121 and the mapping VLAN, and separate ETHIP for the user plane: IP141
and the mapping VLAN).
Figure 1.4 Typical A over IP networking mode (pool of active/standby boards+manual

active/standby LAGs)
2015-11-13
Huawei Confidential
Page 126 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)

Devic Source
e
IP
Next
Hop
BSC
IP150
IP110
IP170
IP130
IP160
IP120
IP180
IP140
Standby Next Hop
Transmission fault detection scheme

The active port of a certain pair of active/standby interface boards on the BSC enables
two BFD sessions to detect the physical IP addresses (IP110 and IP112) of the two
routers, and Whether affect the port swapping is set to YES. The standby port of a
certain pair of active/standby interface boards on the BSC enables an ARP detection
session to detect the VRRP virtual IP addresses (IP114 and IP124 are configured when
the ARP detection is enabled). If faults are detected during both BFD sessions and the
ARP detection indicates that the standby port is normal, the active/standby ports switch
over.
Fault 1: On the BSC side, a pair of active/standby interface boards is faulty.
SACK message receiving by data blocks on the active path of the SCTP links times out
and the SACK messages are resent through the standby path. Upper layer signaling
indicates that no SACK messages are lost but a time delay of about 1s occurs. The SCTP
working links switch to the standby path about 15s (1 + 2 + 3 + 3 + 3 + 3) after the
SACK messages are resent. Then SACK messages are not resent.
ICMP ping detection on an IP address in the IP pool fails (5 * 5s). This IP address is
blocked and subsequent services are provided through other IP addresses.
2015-11-13
Huawei Confidential
Page 127 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The processing after a pair of active/standby interfaces on the BSC side is faulty is
similar to the processing after a pair of active/standby interface boards on the BSC side
is faulty.
IP Interface Boards
For GBSC15.0, only FG2c/FG2d/GOUd/GOUc/EXOUa interface boards support the
transmission interface Pool feature.
Compared with the network where the FG2c interface boards are in use, the network where
the GOUc/GOUd/EXOUa interface boards are deployed has no limitation on transmission
distance and the networking is more flexible. However, the peer end must support GE optical
interfaces. If the peer end has GE optical interfaces, use the GOUc/GOUd/EXOUa interface
boards.
If the FG2c/FG2d interface boards are in use, only GE/FE electrical interfaces are available
and the transmission distance is less than 100 meters (328 ft). Therefore, if the peer end has
sufficient GE optical interfaces, use the GOUc/GOUd/EXOUa interface boards.
Restrictions and Constrains of the Transmission Pool Network

Requirements and specification for the transmission pool networking:
The transmission pool can be used only on layer-3 Ethernet network (an layer-3 route is
deployed before the BSC).
Only the FG2c/FG2d/GOUd/GOUc/EXOUa interface boards support the transmission

interface Pool feature.
The IP addresses in an IP address pool must be device IP addresses and be configured

with the source IP routing. Device IP addresses without source IP routing cannot be
added to the IP address pool.
Each IP address can be in only one pool.
Interfaces of different systems cannot share the transmission pool.
Impact and constrains:

The load-sharing boards in the pool and the boards that have logical ports configured are
mutually exclusive. If boards with logical ports are configured, these boards are preferentially
selected. If these boards are faulty or the CPU of these boards is overloaded, other boards are
selected.
Output: A over IP networking design
BSC Subr
Nam ack
e
No.
Sl
ot
No
.
Boa
rd
Typ
e
Net
Mo
de
Po
rt
Ty
pe
P
or
t
N
o.
Config
ure All
Max
Trans
fer
Unit
Auto
Negotiat
ion
Mode
Port
Rate(
M)
Dupl
ex
Mod
e
Parameter description:
2015-11-13
Board Type: board type.
Huawei Confidential
Page 128 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Net Mode: inter-board mode (inter-board active/standby mode, inter-board load-sharing

mode, or standalone mode).
Port Type: port type (GE/10 GE or FE port).
Max Transfer Unit, Auto Negotiation Mode, Port Rate(M), and Duplex Mode: must
be consistent with those of the directly connected devices.
19.2.3 SCTP Multi-Homing Design

In SCTP multi-homing mode, multiple IP addresses on the local and peer end are configured
to form multiple IP paths. This can improve the networking reliability and is used in A
interface signaling networking (IP over FE/GE).
Implement SCTP multi-homing design based on the networking plan and configuration of the
peer MSC server. The scenarios are as follows (scenario 1 is the most recommended one):
19.2.3.1 Scenario 1
In A over IP mode, the BSC is configured with two pairs of A interface boards in subrack 0
and subrack 1. Each pair works in active/standby mode. The service and signaling are
separated on physical ports. On each pair of boards, one port is allocated for signaling, that is,
totally two physical ports are allocated for signaling. Two IP addresses (Local IP1 and Local
IP2) are allocated to the two physical ports. The peer MSC server provides two pairs of IP
interface boards and two IP addresses (Peer IP1 and Peer IP2) to implement SCTP multihoming with the BSC. A router is deployed between the BSC and the MSC server. Each pair
of interface boards on the BSC must be configured with routes destined for both pairs of
interface boards on the peer MSC server to implement four-homing.
On the BSC side, one physical interface board provides two physical ports for signaling. On the peer
MSC server side, two pairs of interface boards are available, and each pair provides one physical port for
signaling. Alternatively, on the MSC server side, one physical board provides two physical ports for
signaling. In both of the preceding cases, the following networking mode shown in Figure 1.1 can be
used to implement SCTP multi-homing.
2015-11-13
Huawei Confidential
Page 129 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 SCTP four-homing between the BSC and the MSC server
If bit 1 of P42 on the MSC server is set to 1 (default value), two paths are formed, that is, dual-homing
can be implemented. Set bit 1 of P42 to 0 to implement four-homing.
Bit
Bit 1
Description
It controls whether to enable the new path management mechanism.

= 0: The original path management mechanism is used, that is, two
source IP addresses, two destination IP addresses, and four paths are
available.
= 1: The new path management mechanism is used, that is, two source
IP addresses, two destination IP addresses, and two paths are available.
Default value: 1
Application
Scenario
It is used to select a path management mechanism in SCTP multihoming.
Impact on the None.

System
2015-11-13
Related
Software
Parameter
None.
Public or Not
For internal use only
Huawei Confidential
Page 130 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
19.2.3.2 Scenario 2
An M3UA link is added compared with the networking in scenario 1. Configure local address
1 to Local IP1, peer address 1 to Peer IP2, local address 2 to Local IP2, and peer address 2 to
Peer IP1. The other configurations remain unchanged.
Figure 1.1 Two M3UA links and SCTP four-homing between the BSC and the MSC server
If bit 1 of P42 on the MSC server is set to 1 (default value), two paths are formed, that is, dual-homing
can be implemented. Set bit 1 of P42 to 0 to implement four-homing. For detailed information about
P42, see the parameter description in scenario 1.
19.2.3.3 Scenario 3
In A over IP mode, the BSC is configured with a pair of A interface boards that work in
active/standby mode. The signaling and service are separated on physical ports. One physical
port is allocated for signaling, and one IP address is allocated to this physical port. The peer
MSC server provides two pairs of IP interface boards and two IP addresses to implement
SCTP dual-homing on the MSC server side. A router is deployed between the BSC and the
MSC server.
2015-11-13
Huawei Confidential
Page 131 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 SCTP dual-homing on the MSC server side and SCTP single-homing on the BSC side
(1)
Bit 1 of P42 on the MSC server is set to 1 by default. Set bit 1 of P42 to 0 to implement SCTP multihoming on the MSC server side. For detailed information about P42, see the parameter description in
scenario 1.
19.2.3.4 Scenario 4
In A over IP mode, the BSC is configured with a pair of A interface boards that work in
active/standby mode. The signaling and service are separated on physical ports. One physical
port is allocated for signaling, and two IP addresses are allocated to this physical port (multiIP function on the port). The peer MSC server provides two pairs of IP interface boards and
two IP addresses to implement SCTP dual-homing on the MSC server side. A router is
deployed between the BSC and the MSC server.
2015-11-13
Huawei Confidential
Page 132 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 1.1 SCTP dual-homing on the MSC server side and SCTP single-homing on the BSC side
(2)
19.2.3.5 Scenario 5
In A over IP mode, the BSC provides two pairs of A interface boards, and each pair works in
active/standby mode. The service and signaling are separated on physical ports. Each pair of
interface boards provides one port for the signaling, that is, two IP addresses (in different
network segments) are configured to the two pairs of interface boards on the BSC. The peer
MSC server provides one pair of IP interface boards and one IP address to implement SCTP
dual-homing on the BSC side. A router is deployed between the BSC and the MSC server.
2015-11-13
Huawei Confidential
Page 133 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 SCTP single-homing on the MSC server side and SCTP dual-homing on the BSC side
Summary
By default, the duration of detection on a single-homing link is 3s during data services and 7s
when no service is being processed. The detection duration doubles for a dual-homing link.
For GBSS15.0: The RTO min value can be set to 50 ms. However, if the system performs
detection frequently, the SCTP links may repeatedly switch over if transmission condition is
poor.
First local IP address and Second local IP address of an SCTP link must be service IP
addresses (device IP addresses are recommended). Two pairs of the A interface boards provide
a port respectively to implement SCTP multi-homing.
Output: SCTP link
Linkset
name
Local
Port
No.
Local
Address
1
Local
Address
2
Peer
Address
1
Peer
Addres
s2
Peer
Port
No.
2015-11-13
Linkset name: name of the link set to which the M3UA link belongs. It is planned on the
BSC internally.
Local Port No.: local port number. It must be negotiated with the peer end.
Local Address 1: first local IP address. It must be negotiated with the peer end.
Local Address 2: second local IP address. It is required only when SCTP multi-homing
is enabled. It must be negotiated with the peer end.
Huawei Confidential
Page 134 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Peer Address 1: first peer IP address. It must be negotiated with the local end.
Peer Address 2: second peer IP address. It is required only when SCTP multi-homing is
enabled. It must be negotiated with the local end.
Peer Port No.: peer port number. It is planned on the BSC internally.
19.2.4 Signaling Bandwidth Calculation

Calculation method
Use the GSM NEP tool to calculate the bandwidth.
For details, see section Error: Reference source not found.
BSC
Traffi
Name c
7000
Signaling
link type
Bandwidth
per link
Signaling
quantity
Subrac
k
Numbe
r0
Subrac
k
Numbe
r1
Subrac
k
Numbe
r2
64 kbit/s/2
Mbit/s
64
kbit/s/256/512/1
024/2 Mbit/s
19.2.5 Signaling Configuration Principles

If the A interface adopts IP transmission, M3UA signaling links are used, and broadband
signaling and narrowband signaling are not distinguished. Each SCTP/M3UA link supports a
maximum of 4,000 Erlang (it is an experience-based estimated value and not presented to the
customer) traffic. Each SCTP link maps an M3UA link. In signaling design, configure one
SCTP link to bear 4,000 Erlang traffic.
In terms of reliability, configure at least two links regardless of the traffic load. If the A
interface adopts IP transmission, M3UA signaling links are used, and broadband signaling and
narrowband signaling are not distinguished.
Take the signaling balance and reliability of the BSC into account when configuring
signaling links. The principles are as follows:
2015-11-13
In A over IP mode, bandwidth for M3UA links needs not to be configured. Each M3UA
link supports about 4,000 Erlang traffic. To improve reliability, configure an M3UA link
set for each BSC, and the M3UA link set contains at least four M3UA links that are
distributed to boards in different subracks.
In A over IP mode, use the SCTP multi-homing for signaling links to improve reliability.
For details, see section 19.2.3"SCTP Multi-Homing Design."
Huawei Confidential
Page 135 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
19.2.6 Traffic Bandwidth Calculation

This chapter provides engineers with a guide to the calculation of the number of CICs and
transmission bandwidth of the A interface. For calculation methods, see section Error:
Reference source not found.
Methods and principles of bandwidth calculation
Method 1:
Cell channel configuration -> Checking the ErlangB table -> Cell traffic -> Sum of the traffic
of all the cells in the BSC -> Total BSC traffic -> Checking the ErlangB table -> Number of
CICs of the A interface
Method 2:
Traffic per subscriber x Number of subscribers -> Total BSC traffic -> Dividing by 0.7 (traffic
per CIC) -> Number of CICs of the A interface
Generally, the network design tool uses the first calculation method (this method is adopted in
pre-sales marketing). The redundancy of A interface bandwidth calculated using this method
is sufficient, and this method does not cause transmission bottleneck of the BSC. The result
obtained using the second calculation method is precise, and the second method is applicable
to the scenario where transmission resources are insufficient to meet the minimum
transmission requirements.
Table 1.1 Calculation result of A interface bandwidth in IP transmission mode
BSC
Name
Traffic
(Erlan
g)
System
Congestion
Ratio
7000
10-6
Number of
CICs of the A
Interface
IP
Transmission
Bandwidth
19.2.7 IP Address Planning (A over IP)

The IP addresses of the interface boards of the BSC include device IP addresses (logical IP
addresses) and physical port IP addresses. Physical ports support the configuration of multiple
IP addresses. In the GBSS9.0 and later versions, the IP interface board supports multiple
device IP addresses. The following uses the PIU as the common name of the A interface
board, Gb interface board, and Abis interface board. The device IP communication mode is in
use. The device IP address uses the 32-bit mask to save IP resources.
Design principles:
2015-11-13
The planned IP addresses must facilitate follow-up maintenance.
The planned IP addresses must meet the expansion requirements in a certain period in
future.
Plan VLAN tagging based on the next hop or service type to facilitate follow-up
maintenance and expansion.
Huawei Confidential
Page 136 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Design guide:
Step 1 Allocate device IP addresses based on the A interface networking design and number of
FE/GE/10 GE ports calculated in the bandwidth design.
Step 2 Allocate port IP addresses. In the active/standby boards+manual active/standby LAGs,
configure an IP address for the active port. Configure an IP address for the standby port only
when the ARP detection is performed on the standby port. The IP addresses of the
active/standby ports must be in the same network segment.
Step 3 If layer-3 networking is used and the device IP address is used for communication, configure
a route from the intermediate router to the device IP address of the BSC. If layer-3 networking
is used and the port IP address is used for communication, a route from the intermediate
router to the port IP address of the BSC is not required.
Step 4 If the end-to-end solution requires that VLAN tags be added on the BSC side based on
different destination IP addresses, add VLAN tags based on the next hop or service type on
the BSC side.
Step 5 If the control plane and the user plane are separated using multiple IP addresses on the
physical port, configure the BSC to add different VLAN tags based on different next hops.
----End
The device IP address of a board is the logical IP address that the board uses for
communication. The device IP address is valid for all the port IP addresses of the board.
Use the pool of independent board networking mode.
For the FG2c board, if the FE interface mode is adopted, one board can be configured
with 12 port IP addresses that are in different network segments. If the GE interface
mode is adopted, one board can be configured with four port IP addresses. In addition,
the port IP addresses and the device IP address must be in different network segments. If
the PIU adopts the active/standby configuration, only the port on the active board is
allocated with an IP address.
The gateway IP address must be in the same network segment as the port IP address of
the PIU.
One physical port can be configured with a maximum of six IP addresses. The multiple
IP function is supported. The IP addresses of the same physical port must be in different
network segments.
The BSC can add VLAN tags based on the next hop or service type. The VLAN ID
If the interface communication mode is device IP communication, the port IP address works
as the gateway IP address used to communicate with other external devices. If the interface
communication mode is port IP communication, the port IP address works as the IP address
used to communicate with other external devices. In addition, each PIU can be configured
with a maximum of eight service logical IP addresses. A service logical IP address works as
the source or destination IP address used to communicate with other external devices.
For the PIU, the 12 FE ports and four GE ports adopt the router mode, that is, the IP addresses
of the FE ports must be in different network segments. In addition, to simplify the
implementation, the device IP address (logical IP address) of each PIU and the port IP address
of the PIU must be in different network segments.
The PIU interface board of the BSC adopts the routing mode. The IP addresses of the FE ports
on the same interface board must be in different network segments. In addition, the IP
addresses of the FE ports on different interface boards must be in different network segments.
2015-11-13
Huawei Confidential
Page 137 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Example:
Figure 1.1 shows the IP network topology of the BSC.
Figure 1.1 IP network topology of the BSC
In Figure 1.1:
IP_1 to IP_4 (in yellow) are the IP addresses used only for the internal communication in the
BSC. This type of address is generated automatically and does not need to be configured.
Users do not sense and do not need to pay attention to this type of IP address.
IP1 to IP8 are the port IP addresses that need to be planned on the BSC side. The port IP
addresses must be in different network segments.
IP_L1 and IP_L2 are the device IP addresses (logical IP addresses) to be planned for the PIU
on the BSC side. Each PIU can be configured with a maximum of eight logical IP addresses.
19.2.8 Routing Planning (A over IP)

Design guide:
Step 1 Plan BSC routing based on the A interface networking design, IP address of the MSC server,
IP address of the MGW, and current IP address planning of the A over IP interface board. The
BSC needs to be configured with routes to the MGW and MSC server.
Step 2 If the A interface adopts the device IP address for communication, routes to the device IP
address of the BSC need to be configured on the MGW and MSC server.
Step 3 If the A interface adopts the device IP address for communication and layer-3 networking is
used, routes to the device IP address of the BSC need to be configured on the intermediate
router.
----End
Principles of routing planning:
2015-11-13
Huawei Confidential
Page 138 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
If the router adopts the VRRP+VLANIF networking mode, the gateway IP address in the
routing information configured for the A interface board of the BSC is the VRRP IP
address or the port IP address of the device directly connected to the port.
Routes to the MGW and MSC server need to be configured on the BSC. In layer-2
networking mode, routes to the MGW and MSC server are not required.
In BSC active/standby mode, configure routes only on the active board to the MGW and
MSC server. In load-sharing mode, however, configure routes on both load-sharing
boards to the MGW and MSC server.
If the A interface of the BSC adopts the device IP address for communication, configure
routes on the MGW and MSC server to the device IP address of the BSC regardless of
whether the layer-3 or layer-2 networking mode is used. In layer-3 networking mode,
configure a route on the intermediate router to the device IP address of the BSC.
Output:
Inde
x
No.
Subra
ck No.
Sl
ot
No
.
Outgoi
ng Port
No.
Destinati
on IP
Address
Subn
et
Mask
Gatew
ay
To
MSC/MG
W
Outgoing Port No.: port number of the outgoing port from the A interface board of the
BSC to the peer MGW or MSC server.
Destination IP Address: network IP address of the device IP address of the peer

MGW/MSC server (destination of the data from the A interface board of the BSC). If the
peer MGW or MSC server does not have a device IP address (logical IP address), this
parameter indicates the network IP address of the port IP address. The network IP
address is obtained by performing the AND operation on the device IP address (or port
IP address if no device IP address is available) of the MGW or MSC server and the
subnet mask.
Subnet Mask: subnet mask of the IP address of the peer MGW or MSC server.
Gateway: port IP address of the device directly connected to the outgoing port (indicated
by Outgoing Port No.) of the A interface board of the BSC. This IP address must be in
the same network segment as the IP address of the outgoing port of the A interface board.
To MSC/MGW: The routes to the MSC server and MGW need to be configured
separately because the control plane and service plane are separated for the A interface.
19.2.9 QoS Design (A over IP)

Design principles:
Port link detection
2015-11-13
The BFD detection and ARP link detection cannot be enabled at the same time on the
interface board. The BFD detection supports SBFD detection (single-hop BFD
detection, recommended) and MBFD detection (multiple-hop BFD detection), and
the BFD detection requires the next-hop device to support the BFD detection. (The
function that the ARP link detection implements is similar to the single-hop BFD
detection. Generally, it takes several seconds to detect a fault in the ARP link
detection, but takes only tens of milliseconds to detect a fault in the BFD detection.
Huawei Confidential
Page 139 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
However, the ARP link detection can be implemented as long as one end supports the
ARP detection, but the BFD detection can be implemented only if both ends support
the BFD detection.)
One port can be configured with only one detection mode. If a port is not configured
with the BFD detection or ARP link detection, the physical-layer detection is adopted
by default.
The following detection modes are supported: BFD detection on the active port and
ARP detection on the standby port (the standby port does not support BFD detection),
ARP link detection on the active and standby ports, BFD detection on the active port
and physical-layer detection on the standby port, ARP detection on the active port and
physical-layer detection on the standby port, and physical-layer detection on the
active and standby ports. Do not use the MBFD detection. If the customer requires
the MBFD detection, the Huawei headquarters need to work out a solution.
The commonly used detections are the ARP detection and physical-layer detection.
The physical-layer detection does not require data configuration, and the ports
support this detection by default. The configurations recommended for the ARP
detection are as follows: number of retries: 3, interval: 300 ms.
Figure 1.1 shows the promoted detection mode.

Figure 1.1 Promoted detection mode in active/standby mode
Logical port
The logical port bandwidth is different from other types of bandwidth. For the logical
port bandwidth, 1 represents 64 bit/s.
Reserved bandwidth of the logical port = Reserved bandwidth threshold of the logical
port x Logical port bandwidth
Congestion bandwidth of the logical port = Congestion bandwidth threshold of the

logical port x Logical port bandwidth
Congestion clearance bandwidth of the logical port = Congestion clearance threshold

of the logical port x Logical port bandwidth
QoS parameters
See the Configuration Recommendation of the specific version.
2015-11-13
VLAN
Huawei Confidential
Page 140 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
In A over IP mode, only several pairs of boards are used, and the volume of broadcast
packets is small. Therefore, VLAN is not necessary. Add VLAN tags based on the next
hop or add VLAN tags on the intermediate transmission devices.
Output: A over IP QoS design
ARP link detection
Subra
ck No.
Slo
t
No.
Por
t
No.
IP
Addre
ss
Index
Peer
IP
Addre
ss
Arp
Retry
Attemp
ts
Arp
Timeo
ut
VLA
N
Flag
VLA
N ID
Peer IP
Address of
the
Standby
Board
Port No.: port number of the A interface board that requires the physical link detection.
IP Address Index indicates the IP address index. The system supports the configuration
of multiple IP addresses for a port.
Peer IP Address: port IP address of the device that is directly connected to the physical
port.
Arp Retry Attempts: number of ARP detection times in a period. The default value is 3.
ARP Timeout: ARP response timeout interval (after an ARP request is sent) in the ARP
detection. The default timeout interval is 3 seconds. Use the default value 3. The
software has a bug, and do not change the value.
VLAN Flag: whether VLAN tags are added to ARP packets when the BSC implements
ARP detection. If the VLAN function is enabled on the device port that possesses the
gateway IP address of the route configured on the port, this parameter must be enabled,
and the VLAN ID must be the same as the VLAN ID configured for the device port that
possesses the gateway address; otherwise, the route is unreachable.
VLAN ID: VLAN ID in the ARP detection packets when VLAN Flag is set to Enable.
Peer IP Address of the Standby Board: physical IP address of the peer port directly
connected to the physical port of the standby board.
BFD detection
Subrack
No.
Slo
t
No.
Port
No.
IP
Addres
s Index
Peer
IP
Addre
ss
MinTxIn
terval(
ms)
MinRxI
nterval
(ms)
Dete
ct
Mult
Peer IP
Address of
the
Standby
Board
2015-11-13
Peer IP Address: peer IP address in the BFD session. The BFD detection supports only
the next hop detection. Therefore, the peer IP address in the BFD session is the port IP
address of the device that is directly connected to the port.
Huawei Confidential
Page 141 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
MinTxInterval(ms): minimum interval between the BFD control packets that the local
system sends.
MinRxInterval(ms): minimum interval between the BFD control packets that the local
system receives.
Detect Mult: number of detection times, that is, the link is considered disconnected after
the detection fails for the specified number of times.
For details about the other parameters, see the parameter description in "ARP link detection."
Logical port
Subra
ck No.
Sl
ot
No
.
Physi
cal
Port
No.
Log
ic
Por
t
No.
Bandwidth
of the
Logical
Port(32Kpb
s)
Reserved
Bandwidth
Threshold
(%)
Congestio
n
Bandwidth
Threshold
(%)
Congestion
Clear
Bandwidth
Threshold(%)
Physical Port No.: physical port number of the interface board to which the logical port
belongs.
Bandwidth of the Logical Port(32Kpbs): fixed bandwidth of the logical port. It ranges
from 32 kbit/s to 64 kbit/s. The sum of the bandwidths of all the logical ports bound to
the same physical port cannot exceed the bandwidth of the physical port.
For details about the other parameters, see the parameter description in "IP Path."
19.2.10 Interface Interworking

In interface interworking, the following must be taken into consideration:
Protocol type
Protocol phase identifier (Phase2, or Phase2+, GSM_PHASE_2+ is recommended)
OSP
DSP
Whether to use signaling point transfer, and whether to adopt a single signaling point or
multiple signaling points
The preceding information needs to be negotiated between the core network personnel and the
customer.
Table 1.1 A interface interworking parameters
2015-11-13
Parameter
Name
Recommende
d Value
Description
OSP Name
BSC
It is consistent with the BSC name in the BSC

attribute.
OSP Code
None
Set it to the hexadecimal signaling point code

(SPC) that is actually planned by the customer.
Huawei Confidential
Page 142 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Parameter
Name
Recommende
d Value
Description
OSP Code Bit
None
Set it to the actual encoding rule of the country. In

China, it is 14bit.
Network ID
None
Set it to the actual network indicator of the

country. In China, it is NATB.
DSP Name
MSC
Use MSC as the name or use the actual NE name.
DSP Code
None
Set it to the hexadecimal SPC of the interworking

MSC server.
Is using STP?
None
If the interworking is implemented using the

MGW, enable this switch; if the BSC interworks
with the MSC server directly, disable this switch.
STP
MGW
Use MGW as the name or use the actual NE name.
STP Code
None
Set it to the hexadecimal SPC of the signaling

transfer MGW.
MCC
Planned on the
customer network
It uniquely identifies the country to which the

mobile subscribers belong. The value range is 000
to 999. For example, the MCC of China is 460.
MNC
Planned on the
customer network
It identifies the public land mobile network

(PLMN) to which the mobile subscribers belong.
The value range is 000 to 999.
LAC
Planned on the
customer network
It indicates the local area code. The value range is

0 to 65535.
CI
Planned on the
customer network
It indicates the cell identification. The value range

is 0 to 7999.
A interface tag
Negotiated with
the peer end
It indicates the GSM protocol phase identifier

supported by the A interface. Set it based on the A
interface phase identifier provided by the MSC
server.
Generally, it is set
to
GSM_PHASE_2+
.
Value:
GSM_PHASE_1
GSM_PHASE_2
GSM_PHASE_2+
If the system supports the GPRS and AMR

services, the A interface phase identifier must be
set to GSM_PHASE_2+.
2015-11-13
Speech Version
Negotiated with
the peer end
Full-rate version 1, full-rate version 3, half-rate

version 1, and half-rate version 3 are supported.
Set it based on the actual voice version.
Encryption type
Negotiated with
the peer end
The bits, from the most significant bit to the least

significant bit, respectively indicate whether the
A5/0, A5/1, A5/2,..., and A5/7 algorithms are
supported. The value 1 indicates that the BSS
Huawei Confidential
Page 143 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Parameter
Name
Recommende
d Value
Description
supports the encryption algorithm. The value 0
indicates that the BSS does not support the
encryption algorithm. Confirm the supported
encryption algorithms. It cannot be set to all 0s.
The most significant bit must be set to 1, that is,
the A5/0 algorithm must be supported.
A over IP interworking instance in the P project on the live network:

The following document provides the interworking parameter planning of the BSC in A over
IP transmission mode on the live network.
A over IP (A
interface active/standby+VRRP).xls
19.3 A Interface Design (TDM)

The A interface is between the BSS and the MSC. In versions earlier than BSC6910
V100R015C01, the BSC6910 only supports A over IP transmission. In BSC6910
V100R015C01 and later, the BSC6910 also supports A over TDM transmission (only the
optical interface is supported, and the interface board is POUc).
The protocols over the A interface are as follows:
1.
TDM-based Interface
Physically, the A interface is a trunk circuit and trunk interface between the BSS and the
MSC. The following figure illustrates the reference protocol model on the control plane over
the A interface.
2015-11-13
Huawei Confidential
Page 144 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Reference protocol model on the control plane of the A interface
BSSAP: Base Station Subsystem Application Part

BSSMAP: Base Station Subsystem Management Application Part
DTAP: Direct Transfer Application Part
MTP: Message Transfer Part
SCCP: Signaling Connection Control Part
Protocol and specifications that the A interface complies with are as follows:
Physical layer: complies with ITU-T G.703, G.704, G.705, and G.732.
MTP: complies with ITU-T Q.701-Q.704, Q.706, and Q.707.
SCCP: complies with ITU-T Q.711-Q.714 and Q.716.
BSSAP: complies with 3GPP TS 48.008, 3GPP TS 24.008, and 3GPP TS 44.018.
BSC6910 V100R015C01 has the following restrictions:
Only the POUc interface board is supported in A over TDM transmission mode.
The Ater interface is not supported. That is, BM/TC separated mode, remote TC, TC
pool, and Ater over IP are not supported.
TDM/IP dual-stack is not supported.
The control plane can use IP transmission and the user plane can use TDM transmission.

The A over TDM transmission is a traditional networking mode over the A interface on a
GSM network. The following figure shows a typical topology.
2015-11-13
Huawei Confidential
Page 145 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 BSCs connected to an MGW
Generally, a BSC is connected to an MGW and both of them are homed to the same MSC.
To improve network reliability, distribute A interface traffic and signaling on the BSC to
different MGWs if two or more MGWs are available under an MSC. This enables the BSC to
properly provide services even if any of the MGWs is faulty, as shown in the following figure.
Figure 1.2 BSCs connected to multiple MGWs
19.3.3 Transmission Bandwidth Design

For details about how to calculate the transmission bandwidth, see section 10.2.3.
2015-11-13
Huawei Confidential
Page 146 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
19.3.4 Signaling Configuration Principles

In an SS7 network, two types of signaling links are used: 64 kbit/s links and 2 Mbit/s highspeed signaling links. The 2-Mbit/s signaling links consist of the standard signaling links at
the rate of 2048 kbit/s and signaling links at the rate of N x 64 kbit/s (N ranges from 2 to 31).
Specifications of signaling points:
A BSC can be configured with a maximum of four local signaling points.
Specifications of SS7 signaling links:
Narrowband signaling link: The bandwidth of each signaling link is 64 kbit/s. A single
signaling point can be configured with a maximum of 16 narrowband signaling links, and a
BSC can be configured with a maximum of 64 (16 x 4) narrowband signaling links.
2 Mbit/s signaling link: The bandwidth of each signaling link is 2 Mbit/s. The maximum
bandwidth for a single signaling point is 16 Mbit/s. A BSC can be configured with 32 2 Mbit/s
signaling links. That is, the maximum bandwidth for the BSC is 64 Mbit/s.
The bandwidth calculation method and processing for a narrowband signaling link are the
same as those for a high-speed signaling link. The only difference between a narrowband
signaling link (64 kbit/s) and a high-speed signaling link (2 Mbit/s) lies in the bandwidth. The
bandwidth of a narrowband signaling link is 64 kbit/s while that of a high-speed signaling link
is N x 64 kbit/s (N ranges from 2 to 31).
As stipulated in GSM protocols Q.703 and Q.704, a maximum of 16 signaling links can be
configured between two signaling points. If 64 kbit/s signaling links are used, the maximum
bandwidth for signaling links between two signaling points can be 1 Mbit/s (16 x 64 kbit/s).
In A over TDM transmission mode, the BSC6900 supports the following two methods for
increasing the capacity of signaling links:
Using local multiple signaling points: The BSC6900 supports a maximum of four local
signaling points. The maximum number of signaling links that can be configured over
the A interface increases to 64.
Using 2 Mbit/s signaling links: The maximum bandwidth for each 2 Mbit/s signaling
link can reach 1984 kbit/s.
The MTP3 signaling link set and signaling route mask are designed as follows:
Check whether the signaling link mask for the MTP3 signaling link set is appropriately
configured.
Assume that the number of digits 1 in the signaling link mask (in binary) for the
signaling link set is n. If 2n is greater than or equal to the number of signaling links
contained in the signaling link set, the configuration is appropriate. Otherwise, the
configuration is inappropriate.
For example, assume that the signaling link mask is B0001 (the digits after B are binary
numbers). If the maximum number of signaling links contained in the signaling link set
is 2, the configuration is appropriate. If the maximum number of signaling links
contained in the signaling link set is greater than 2, the configuration is inappropriate,
and the signaling link mask must be reconfigured.
2015-11-13
Check whether the signaling route mask for the destination signaling point is
appropriately configured.
Huawei Confidential
Page 147 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Assume that the number of digits 1 in the signaling route mask (in binary) for the
destination signaling point is n. If 2n is greater than or equal to the number of routes
contained in the destination signaling point, the configuration is appropriate. Otherwise,
the configuration is inappropriate.
For example, assume that the signaling route mask is B0001 (the digits after B are binary
numbers). If the maximum number of routes contained in the destination signaling point
is 2, the configuration is appropriate. If the maximum number of routes contained in the
destination signaling point is greater than 2, the configuration is inappropriate, and the
signaling route mask must be reconfigured.
In versions earlier than GBSS14.0, the result of the "and" operation performed between
the signaling route mask and signaling link mask must be 0. Otherwise, signaling load
may be unbalanced. In GBSS14.0 and later, this restriction no longer applies, and the
default signaling route mask is B1111.
Local signaling points of the BSC6900 are not bound to the BM subrack. Therefore, a single
subrack can be configured with four local signaling points. If the CN does not support
wideband signaling, it is recommended that local multiple signaling points are configured for
the BSC (in version earlier than GBSS9.0, source signaling points are bound to the BM
subrack and TCS; in GBSS9.0 and later, source signaling points are no longer bound to the
subracks. Therefore, a single subrack can be configured with multiple source signaling points
even if only an A interface board is available). When local multiple signaling points are
configured, the signaling point binding relationship between cells and CICs must be
configured. It is recommended that cells be homed to different signaling points based on their
areas. (The inter-cell handover procedure between different signaling points is similar with
the inter-BSC handover procedure. Therefore, cell homing based on areas can decrease the
number of such handovers.)
In A over IP transmission mode, the signaling links are M3UA links which do not distinguish
wideband and narrowband signaling links. Each SCTP or M3UA link supports maximum
traffic volume of 4000 Erlang (this value is based on experience and not provided for
customers. It is equivalent to the number of speech channels supported by 16 64 kbit/s SS7
signaling links. Each 64 kbit/s signaling link supports 256 CIC speech channels). During the
signaling design, you are advised to set an SCTP link per 4000 Erlang.
To ensure reliability, a minimum of two SCTP links must be configured regardless of traffic
volume.
In A over IP transmission mode, the signaling links are M3UA links which do not distinguish
wideband and narrowband signaling links.
In TDM transmission mode, the configuration of signaling links must comply with the
following rules to ensure the signaling load balancing and reliability:
2015-11-13
When high-speed signaling links are used, the restrictions on the number of signaling
links and bandwidth supported by the PARC platform must be considered. A maximum
of eight high-speed signaling links can be configured in each GMPS/GEPS, and the total
bandwidth for each GMPS/GEPS cannot exceed 4 Mbit/s. In GBSS8.1 and earlier
versions, this restriction applies. In GBSS9.0 and later, this restriction does not apply.
A maximum of 32 2-Mbit/s signaling links can be configured.
High-speed signaling links and 64 kbit/s signaling links cannot be simultaneously

configured between the BSC and the same destination signaling point.
The BSC6900 supports simultaneous configuration of high-speed signaling links and

local multiple signaling points.
Huawei Confidential
Page 148 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Different bandwidths can be configured for each high-speed signaling link. Since the
SS7 signaling links use load sharing, the same bandwidth can be set for multiple SS7
high-speed signaling links.
Any combination of timeslots (except timeslot 0) on an E1 link can form a high-speed

signaling link.
Under the same signaling point, at least two high-speed signaling links must be
configured. The two signaling links are distributed to different STM-1 ports of an A
interface board or to different A interface boards to improve the reliability.
The BSC supports MSC Pool. A BSC can be connected to multiple MSCs. In TDM
transmission mode, the SS7 configuration is determined according to the proportion of
the number of A CICs between the BSC and MSC to the total number of A CICs.
19.4 Gb Interface Design

In new offices of the BSC6910, the embedded PCU is used; therefore, the external PCU
solution is not taken into consideration in the Gb interface design.

The Gb interface connects the BSC (including the PS service) and the SGSN after the PCU is
embedded, and it is the standard interface defined in the protocol.
The Gb interface of the GBSS15.0 BSC6910 supports only IP networking mode.
Figure 1.1 shows the protocol model of the Gb interface.
Figure 1.1 Gb over IP protocol stack
The Gb interface implements the communication between the SGSN and the BSS system and
between the SGSN and MSs, transmits packet data, manages the mobility, and manages
sessions. The Gb interface is mandatory in GPRS networking.
Physical-layer protocol L1
The physical-layer configurations and protocols defined in GSM 08.14 can be used.
Physical resources are configured in the O&M process.
2015-11-13
Network service layer (NS)
Huawei Confidential
Page 149 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
NS refers to the network service control part in the NS protocol. The NS-layer protocol
transmits service data unit (SDU) data, manages NS-VC links, shares subscriber data in
load-sharing mode, and provides the report of network congestion status indication and
the network status report for the upper-layer service module.
BSSGP protocol layer

In the BSS, the BSSGP is the interface between the LLC frame and the RLC/MAC
block. In the SGSN, the BSSGP is the interface between the RLC-/MAC-originated
information and the LLC frame.
The BSSGP provides radio-related data, QoS, and routing information to meet the
requirements for subscriber data transmission between the BSS and the SGSN. The
BSSGP transmits signaling information and subscriber data, performs flow control for
downlink data, blocks and unblocks the BVC, dynamically configures and manages the
BVC, and detects errors in interface messages.
In the Gb over IP function, the Gb interface uses the IP protocol to provide the lower-layer
transmission service for the NS. The IP transmission module implements the interworking
between sub-networks so that the PCU and the SGSN can directly connect to each other
(direct connection mode) or connect to each other through the IP transmission network
(routing mode).
With the Gb over IP function, IP headers are compressed, and the data on the Gb interface can
share the transmission bandwidth to improve the transmission efficiency and save the
transmission cost. After the Gb over IP function is used, the Gb interface maintenance
commands are simplified, and the maintenance work is simplified.
Generally, the Gb interface adopts IP transmission.
Figure 1.2 shows the logical networking diagram of the embedded PCU.
Figure 1.2 Embedded PCU networking
2015-11-13
Huawei Confidential
Page 150 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)

In IP networking mode, the PCU and the SGSN can be connected in either of the following
modes:
Direct connection (direct connection mode)
IP transmission network connection (routing mode)
In Gb over IP direct connection mode, the PCU and the SGSN are directly connected to each
other without any intermediate IP network. In this connection mode, a switch can be deployed
to provide the layer-2 switching service for the PCU and the SGSN, as shown in Figure 1.1.
Figure 1.1 Direction connection (Gb over IP)
In Gb over IP routing connection mode, the PCU and the SGSN are connected through an
intermediate IP network. In this connection mode, routers are used to provide the layer-3
routing service for the PCU and the SGSN, as shown in Figure 1.2.
Figure 1.2 IP transmission network connection (Gb over IP)
Design guide:
Step 1 Determine the networking mode (direction connection networking or layer-2/layer-3
networking) based on the requirements of the customer for the Gb interface networking. The
active/standby boards and manual active/standby LAGs+layer-3 router VRRP networking
mode is recommended. Do not use the layer-2 networking or direct connection mode.
Step 2 Design the networking reliability of the interface boards based on the capability of the
interworking device and the requirements of the customer. The networking reliability of the
Gb interface is the same as that of the A interface. The optional modes are active/standby
boards, route-based active/standby boards in standalone mode, and route-based boards load
sharing in standalone mode. Use the active/standby boards+active/standby ports.
Step 3 If the layer-2/layer-3 networking is adopted, design the layer-2/layer-3 networking reliability
and use layer-3 router VRRP+VLANIF to ensure the router reliability.
----End
Design principles:
2015-11-13
Huawei Confidential
Page 151 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
If the IP transmission conditions are met, use IP transmission because the IP transmission
bandwidth is sufficient; this facilitates follow-up expansion and SGSN pool
implementation.
In IP transmission mode, configure the Gb interface boards to work in active/standby

mode.
Connect the active and standby ports to the two VRRP routers respectively.
The typical Gb over IP networking modes are as follows:

Gb interface networking schemes for GBSS15.0 BSC6910:
Promoted scheme
Networking: Pool of active/standby boards+manual active/standby LAGs
CE2.
The BSC and the SGSN are deployed in layer-3 networking mode. The SGSN connects
to the BSC in loose coupling mode and the BSC can be configured to work in either
active/standby mode or load-sharing mode.
Figure 1.1 Typical Gb over IP networking mode (active/standby boards+manual active/standby

LAGs)
Ports configured on the two routers CE1 and CE2 and used by the VRRP IP addresses
must be configured to layer-2 networking mode, including the ports connecting the BSC
and the trunk ports between routers.
Bandwidth of trunks must be greater than 50% of the total data volume of the BSC and
at least two GE interfaces must be converged.
The ports in even-numbered slots of the BSC connect to a high-priority VRRP router,
which improves the probability that the same active path is used by the BSC and the
router.
Ports of the BSC do not support layer-2 exchange, and peer devices are not required to
be configured with the STP protocol. If peer devices are configured with the STP
protocol, mode of the ports connecting to the BSC must be modified. If peer devices are
configured with the RSTP/MSTP, set the port mode to the STP edge port. If peer devices
are configured with the STP (802.1D-1998), set the port mode to PortFast. If peer
devices do not support the preceding modes, set the port mode to STP Disable.
2015-11-13
Huawei Confidential
Page 152 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)

Devic Destination Next
e
IP
Hop
Priorit
y
BSC
IP311
IP119
Default
CE1
IP150
IP111
Default
CE2
IP150
IP111
Default

The active ports of the active/standby interface boards on the BSC enable two BFD
sessions to detect the physical IP addresses of the two routers. The standby ports of
active/standby interface boards on the BSC enable an ARP detection session. The BSC
performs a BFD detection every 100 ms for three times and performs an ARP detection
every 10s for three times. Configure the delay enabling BFD on CE1 and CE2 to avoid
service interruption of CE1 and CE2 due to a reset upon power-off.
2015-11-13
CE1 is faulty (suspended or powered off): Heartbeat detection on the VRRP port of
CE2 fails and the standby VRRP port of CE2 becomes active. If the BSC sends
packets to the SGSN through CE1, the two BFD detections detect faults, manual
active/standby LAGs switch over, and the BSC sends packets to the SGSN through
CE2. If the BSC sends packets to the SGSN through CE2, only one BFD session on
the BSC fails and the active/standby ports are not switched over. After CE1 is faulty,
the OSPF re-converges, and the SGSN sends packets to the BSC through CE2.
Regardless of whether the BSC sends packets to the SGSN through CE1 or CE2, the
BSC and the SGSN exchange messages through CE2. In this case, services are not
interrupted and the switchover duration is less than 1s.
Connection between the BSC and CE1 is faulty: Heartbeat communication on the
VRRP ports is correct and the VRRP ports are not switched over. If the BSC sends
packets to the SGSN through CE2, only the standby port is faulty, the active/standby
ports are not switched over, and the original path remains unchanged. If the BSC
sends packets to the SGSN through CE1, it then sends packets through CE2: The two
BFD sessions on the active port of the BSC fail, the active/standby ports switch over,
and the path of uplink packets of the BSC is BSC => CE2 => CE1 => SGSN. When
the active/standby ports switch over, the BSC sends free ARP to update the ARP
entries on CE1 and CE2. The SGSN sends packets to the BSC through CE2. In this
case, packets are sent to the BSC through CE1 and CE2, services are not interrupted
and the switchover duration is less than 1s.
Connection between two routers is faulty: BFD detection of heartbeat messages on

VRRP1 fails and the standby port on CE2 becomes active, then VRRP1 has two
active ports. One BFD detection on the active port of the BSC succeeds and the other
BFD detection fails, then the active/standby ports are not switched over. If the BSC
sends packets to the SGSN through CE1, the original path remains unchanged. If the
BSC sends packets to the SGSN through CE1, then the BSC sends packets to the
SGSN through CE2: BSC <=> CE2 <=> SGSN. In this case, packet sending from the
BSC to the SGSN is not interrupted. If the SGSN sends packets to the BSC through
CE1, packet sending is not interrupted. If the BSC sends packets to the SGSN
through CE2, packet sending is interrupted. That is, there is 50% probability that
packet sending from the SGSN to the BSC may be interrupted (see the following
Note).
Huawei Confidential
Page 153 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Connection between CE1 and the intermediate network is faulty: The OSPF reconverges, services are not interrupted, and the switchover duration is less than 1s.
Manual switchover of ports on the BSC: VRRP routes and intermediate network
routes are not affected. When the active/standby ports switch over, the BSC sends
free ARP to update the ARP entries on CE1 and CE2. The SGSN sends packets to the
BSC through CE2. In this case, services are not interrupted and the switchover
duration is less than 1s.
Interface boards of the BSC are faulty: The active/standby boards switch over, and
the active/standby ports switch over. In this case, services are not interrupted and the
switchover duration is 1 to 3s.
Intermediate network is faulty, which causes a service IP address of the BSC

unreachable (rarely occurs): If only one service IP address is configured on the BSC
(if only a pair of ports is configured, only one service IP address is configured by
default), no redundancy service IP address is available. In this case, all services on
this interface are interrupted. In actual commercial deployment, multiple pairs of
ports are configured on the BSC to bearer multiple service IP addresses, and one pair
of ports can bearer multiple service IP addresses. In this case, if ping detection
indicates that the network service virtual connection (NSVC) of an IP address is
faulty (service restoration requires 5s x 5 = 25s by default), the BSC allocates
subsequent services to another IP address. That is, if the Gb interface on the BSC has
only one IP address, all services on this interface are interrupted. If the Gb interface
on the BSC has more than one IP address but services on one IP address are
interrupted, service access success rate is low within 25s but services restore after
25s. However, because the intermediate network has self-healing function, IP address
reachable rarely occurs on VRRP network. This problem can be ignored.
Logical link configuration:

NSVL configuration
Devic NSVL IP
e
Addr
UDP Port
No
Weight
BSC
NSVL 1 IP150
Port 1
100 (see the following Note)
SGSN
NSVL 1 IP311
Port 2
100 (see the following Note)
The weight here is not a percentage, the value range is from 1 to 255, and the default value 100
is used (the same for each NSVL). If the bandwidth of intermediate paths is different or the
multiple SGSNs are configured on the peer end, modify the value of the weight parameter.
Modifying the value of the weight parameter affects the load-sharing effect of this scheme.
Therefore, contact R&D engineers if the value of the weight parameter needs to be modified.
NSVC configuration
NSVC BSC IP
Addr
BSC UDP Port

No
SGSN IP
Addr
SGSN UDP Port

No
NSVC 1 IP150
Port 1
IP311
Port 2
Optional scheme
Networking: Active/standby boards+dual-active ports
2015-11-13
Huawei Confidential
Page 154 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The BSC is directly connected to the dual routers through two independent ports on the
active/standby interface boards.
Device IP addresses are configured on the logical active board. With the active/standby
Figure 1.2 Typical Gb over IP networking mode (active/standby boards+dual-active ports)
The Gb interface occupies an Ethernet port. The control plane and user plane are not
distinguished, therefore, no VLAN needs to be configured.
Configure IP addresses in the same network segment for each sub-interface, which
facilitates route combination and simplifies intermediate network route. To facilitate
network expansion, service IP addresses of the BSC use logical IP addresses (device IP
addresses).
e
IP
Hop
Priorit
y
BSC
CE1
CE2
2015-11-13
IP311/32
IP110
High
IP331/32
IP130
High
IP311/32
IP130
Low
IP331/32
IP110
Low
IP150/32
IP111
High
IP170/32
IP111
Low
IP150/32
IP131
Low
IP170/32
IP131
High
Huawei Confidential
Page 155 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Local service IP addresses and peer service IP addresses are grouped in two. With route
priority configuration, IP addresses in the two groups have different priorities, thereby
implementing load-sharing. In addition, backup routes are configured to ensure
reliability.
Dynamic route protocols (OSPF/ISIS) need to be configured between CE1 or CE2 and
intermediate bearer networks. Static routes also need to be configured. In addition, route
priorities need to be configured to ensure that the route from CE1 to IP150 has higher
priority than the route from CE2 to IP150, and the route from CE2 toIP170 has higher
priority than the route from CE1 to IP170. In this case, data is sent to IP150 through CE1
and sent to IP170 through CE2.
Detection mechanism
2015-11-13
Connection between the BSC and CE1 is faulty: SBFD detection on the route from
the BSC to IP110 of CE1 fails and the active route whose next hop is IP110 becomes
invalid. Then the active/standby routes switch over. The standby route whose next
hop is IP130 becomes valid and services migrated to this route. At the same time, the
static route that bound to the SBFD on CE1 becomes invalid, and the OSPF reconverges. The next hop of the route between intermediate network to IP150 switches
to CE2 from CE1 and the SGSN sends packets to the BSC only through CE2. In this
case, services are not interrupted and the switchover duration is less than 1s.
CE1 is faulty (suspended or powered off): SBFD detection on the route from the BSC
to CE1 fails and the active route whose next hop is IP110 becomes invalid. Then the
active/standby routes switch over. The standby route whose next hop is IP130
becomes valid and services migrated to this route. At the same time, the BFD for
OSPF detection on the intermediate network indicates that CE1 is faulty, and the
OSPF re-converges. The next hop of the route between intermediate network to
IP150 switches to CE2 from CE1 and the SGSN sends packets to the BSC only
through CE2. In this case, services are not interrupted and the switchover duration is
less than 1s.
Main interface boards of the BSC are faulty: In most cases, the BSC can detect board
faults and if it detects a board fault, the active/standby boards of the BSC switch over,
and the active/standby routes switch over. In this case, services are not interrupted
and the switchover duration is less than 1s. In rare cases, the BSC cannot detect board
faults. If NSVC detection detects a board fault, the active/standby boards of the BSC
switch over. The board switchover triggering mechanism is similar to that when the
intermediate network is faulty.
Intermediate network is faulty, which causes a service IP address of the BSC

unreachable (rarely occurs): An NSVC detection takes 15s (3s x 5 = 15s) to 45s (30s
+ 15s = 45s) (The BSC starts NSVC detection after the Tns-test times out and the
system determines that the NSVL is faulty after NS-ALIVE sends an NS-ALIVERETRIES message for five times). After the BSC detects an NS-VC fault, the BSC
switches the uplink data on this NSVC to a normal NSVC (the BSC replaces only
local NSVL) and informs the peer end of the NSVC fault in any of the following
ways:
Huawei Confidential
Page 156 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
a)
The NS STATUS informs the peer end of the local NSVL fault.
b)
The NS BLOCK blocks the NSVC.
c)
R-BIT in NS UNITDATA of uplink data is set to 1, requesting the SGSN to

replace the peer NSVL.
In this case, data on the faulty path and to be sent to the BSC is sent through a normal
NSVL (source NSVL in NS UNITDATA). That is, services over the Gb interface
restore after 15 to 45s.
Capacity restriction after the switchover is not taken into consideration.
In this scheme, the assumed convergence time of the intermediate network is less than 1s.
Output: Gb over IP networking design

Subrac
k No.
Slo
t
No
.
Boar
d
Type
Net
Mod
e
Por
t
Typ
e
Po
rt
No
.
Configu
re All
Max
Transf
er
Unit
Auto
Negotiati
on Mode
Port
Rate(
M)
Flow
Contr
ol
Duple
x
Mode
Board Type: board type.
Net Mode: inter-board mode (inter-board active/standby mode, inter-board load-sharing

mode, or standalone mode).
Port Type: port type (GE or FE port).
Flow Control: whether the flow control is enabled.
Max Transfer Unit, Auto Negotiation Mode, Port Rate(M), and Duplex Mode: must
be consistent with those of the directly connected devices.
19.4.3 Bandwidth Calculation

Output: Gb over IP bandwidth design
BSC
Nam
e
Configur
ed BTS
Number
active
subscribe
rs
Average
traffic in
BH/sub
(bps)
Gb IP
throughp
ut
(Mbps)
GE Link
Numbe
r
Board
Numbe
r
IP
Segmen
t
Number
19.4.4 IP Address Planning

The BSC interface boards mainly contain the device IP address (logical IP address) and
physical port IP address. A physical port can be configured with multiple IP addresses. The
following uses the PIU as the common name of the A interface board, Gb interface board, and
Abis interface board.
2015-11-13
Huawei Confidential
Page 157 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Design principles:
The planned IP addresses must meet the expansion requirements in a certain period
(determined by the customer) in future.
The planned IP addresses must facilitate follow-up maintenance.
The Gb interface supports VLAN tagging based on the next hop. Add VLAN tags on the
intermediate transmission devices.
Design guide:
Step 1 The device IP address communication mode (recommended) or port IP address
communication mode can be used.
Configure and use the port IP address or device IP address. The system automatically
identifies the communication mode (port IP communication or device IP communication). The
processing of the A interface is the same as the processing of the Abis interface.
Step 2 Allocate port IP addresses and device IP addresses based on the available IP address
resources, Gb over IP networking design, and number of boards and number of FE/GE ports
calculated in the bandwidth design provided by the customer. Use device IP addresses.
Step 3 In layer-3 networking mode, configure a route on the intermediate router to the device IP
address of the BSC.
The service address and port address are separated on the Gb interface board of the
BSC6910. The NSVL uses the service address, and the port address is used to forward
data for the communication between the external device and the service address.
Therefore, on the router that connects to the BSC, configure a route whose destination
address is the service address of the Gb interface and the next hop is the port IP address
on the Gb interface board. On the SGSN, configure a route destined for the service
address of the Gb interface board.
In active/standby mode, the device IP addresses of the active and standby boards must be
the same.
The subnet mask of the device IP address can be 255.255.255.255.
If one board is configured with different port IP addresses, the port IP addresses must be
in different network segments. In addition, the port IP addresses and the device IP
address must be in different network segments. In active/standby configuration mode,
configure the active port IP address. When the ARP detection is enabled on the standby
port, configure an IP address for the standby port. In other cases, the standby port does
not require an IP address. The IP addresses of the active/standby ports must be in the
same network segment.
When you configure a BSC route, the destination IP address must be configured to the
network address and cannot be in the same network segment as the port IP address of the
board. Otherwise, the destination IP address is invalid.
Network address is a technical term. A network address refers to an address that is used for addressing
the peer device when two devices communicate with each other. The network address is the AND result
of the peer IP address and subnet mask. For example, the peer IP address is 192.168.80.2 and the subnet
mask is 255.255.255.128; the AND result of them is 192.168.80.0 (network address).
2015-11-13
When you configure a BSC route, the gateway IP address must be in the same network
segment as the port IP address of the board.
The IP address cannot be all 0s or all 255s.
The IP address cannot be a loopback address whose network number is 127.x.x.x.
Huawei Confidential
Page 158 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The IP address cannot be the multicast IP address of category D, that is, address in the
range of 224.0.0.0 to 239.255.255.255.
The IP address cannot be the reserved IP address of category E, that is, address in the
range of 240.0.0.0 to 247.255.255.255.
19.4.5 Routing Planning (Gb over IP)

Design principles:
The designed routing solution must facilitate follow-up maintenance.
The designed routing solution must facilitate follow-up expansion.
Design guide:
Step 1 Plan BSC routing based on the Gb interface networking design. On the BSC, configure a
route to the SGSN. The routing information configured on the BSC includes the destination IP
address, subnet mask, and gateway address.
Step 2 On the SGSN, configure a route to the device IP address of the BSC.
Step 3 In layer-3 networking mode, configure a route on the intermediate router to the device IP
address of the BSC.
----End
On the Gb interface board, add a BSC route, that is, plan a route to the service IP address
of the SGSN. Destination IP is the network address of the service address of the SGSN.
Subnet Mask is the subnet mask of the network where the service address of the SGSN
resides. Next Gateway is the IP address of the port (connected to the BSC) on the first
router on the way from the BSC to the SGSN; this address must be in the same network
segment as the port IP address connected to the BSC.
When the IP subnet configuration mode is dynamic configuration, if the control-plane

address and service address of the SGSN are not in the same network segment, configure
a route to the network segment of the control-plane address based on the preceding
principle. The gateway configuration in control-plane route configuration is the same as
the gateway configuration in service-plane route configuration.
Output: Gb over IP routing design

NE
name
Boar
d
Type
Board
Numb
er
Po
rt
ID
SGS
N
Nam
e
Destinati
on IP
Subn
et
Mask
Next
Gateway
2015-11-13
Destination IP Address: network IP address of the destination IP address of the peer

SGSN (destination of the data from the Gb interface board of the BSC). If the peer
SGSN does not have a device IP address (logical IP address), this parameter indicates the
network IP address of the port IP address. The network IP address is obtained by
performing the AND operation on the device IP address (or port IP address if no device
IP address is available) of the SGSN and the subnet mask.
Huawei Confidential
Page 159 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Subnet Mask: subnet mask of the IP address of the peer SGSN.
Gateway: port IP address of the device directly connected to the outgoing port of the Gb
interface board on the BSC. This IP address must be in the same network segment as the
IP address of the outgoing port of the Gb interface board.
Route Priority: route priority, and the default value is 1.
19.4.6 QoS Design (Gb over IP)

Design principles:
Port link detection
The BFD detection and ARP link detection cannot be enabled at the same time on the
interface board.
One port can be configured with only one detection mode. If a port is not configured
with the BFD detection or ARP link detection, the physical-layer detection is adopted by
default.
The following detection modes are supported: BFD detection on the active port and ARP
detection on the standby port (the standby port does not support BFD detection), ARP
link detection on the active and standby ports, BFD detection on the active port and
physical-layer detection on the standby port, ARP detection on the active port and
physical-layer detection on the standby port, and physical-layer detection on the active
and standby ports.
Design guide:
Step 1 Design the port QoS attribute parameters based on the capability of the interworking device.
Step 2 Determine the port link detection mode (BFD detection, ARP link detection, or physical
detection) based on the capability of the interworking device. The physical detection is
supported by default and does not require configuration.
----End
Output: Gb over IP QoS design
ARP link detection
Subra
ck No.
Sl
ot
No
.
Po
rt
No
.
IP
Addres
s Index
Peer
IP
Addre
ss
Arp
Retry
Attempt
s
Arp
Timeo
ut
VLA
N
Flag
VLA
N ID
Peer IP
Address of
the Standby
Board
2015-11-13
Port No.: port number of the Abis interface board that requires the physical link
detection.
IP Address Index: IP address index. The system supports the configuration of multiple
IP addresses for a port.
port.
Huawei Confidential
Page 160 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
BFD detection
Subrac Slo
k No.
t
No
.
Por
t
No.
IP
Addre
ss
Index
Peer
IP
Addre
ss
MinT
xInt
erva
l(ms
)
Min
RxIn
terv
al(m
s)
Dete
ct
Mult
Peer IP
Address
of the
Standby
Board
system sends.
system receives.
19.4.7 Configuration Principles
One BSC can connect to multiple SGSNs.
One BSC can be configured with multiple network service entities (NSEs).
NSEs belong to the same SGSN must be different.

One SGSN can correspond to multiple NSEs in a BSC, but all point-to-point BSSGP
virtual connections (PTP BVCs) in an NSE can provide services only for one SGSN.
One cell corresponds to one PTP BVC (except for the SGSN pool scenario).
One BSC can belong to different SGSN cells.
19.4.8 Interface Interworking

The Gb interface interworking needs to be negotiated with the SGSN. In the negotiation,
prepare the template of interface interworking parameters, parameter description, networking
diagram, and precautions to improve the interworking efficiency.
2015-11-13
Huawei Confidential
Page 161 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
If the Gb interface adopts the IP protocol, the Gb interface link needs to be configured with
the NSE, local NSVL, remote NSVL, and PTP BVC objects.
Table 1.1 Parameters of Gb interface link configuration
Object
Description
Recommended
Configuration
NSE
On the BSSGP layer, the NSE functions as

a BVC collection; on the NS layer, the NSE
functions as an IP NSVC collection. An
NSEI uniquely identifies an NSE. The
NSEIs configured on the BSC are the same
as the NSEIs configured on the SGSN.
Non-SGSN pool
networking: Configure
each BSC with an NSE (or
multiple NSEs).
Select the IP protocol for the NSE.

Configure the local and remote NSVL
objects.
Local and
remote NSVLs
The local NSVL determines the location

information about the Gb interface board,
establishes mappings between NSEs and
device IP addresses/port numbers, and
determines the ports through which the
NSE cell data is transmitted. The remote
NSVL establishes mappings between NSEs
and device IP addresses/port numbers on
the SGSN side and determines the ports
through which the NSE cell data is
transmitted.
SGSN pool networking:

The number of NSEs is
equal to the number of
SGSNs in the SGSN pool
to which the BSC belongs.
The NSEs on the BSC map
the NSEs on the SGSN
based on the one-to-one
relationship, that is, the
NSEs at the two ends must
be consistent.
Each NSE must be
configured with at least two
NSVLs.
The number of NSVLs
must be equal to or larger
than the number of physical
transmission links.
Use logical IP addresses
(device IP addresses) as the
IP addresses of NSVL.
The NSVL is used in Gb over IP mode, and

it is equivalent to the NSVC in Gb over FR
mode.
On the IP network, an IP NSVC is uniquely
identified a quadruple of the local IP
address, local port number, peer IP address,
and peer port number.
PTP BVC
The PTP BVC establishes the point-to-point

virtual connection on the BSSGP layer.
Each cell corresponds to one PTP BVC.
Non-SGSN pool
networking: One cell
corresponds to one PTP
BVC.
SGSN pool networking: A
PTP BVC must be
configured between each
cell and each SGSN in the
Pool.
2015-11-13
Huawei Confidential
Page 162 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.2 shows the logical connection at the NS layer and BSSGP layer between the BSC
and the SGSN.
Figure 1.2 Logical connection between the NS layer and the SSGP layer
As shown in Figure 1.2, on the BSSGP layer, the NSE functions as a BVC collection
(equivalent to a cell collection); on the NS layer, the NSE functions as an IP NSVC group (IP
NSVC collection). The NS layer provides data transmission channels for the BSSGP layer.
The channel used to transmit the data of the cells in the same NSE must be an IP NSVC in the
IP NSVC collection of the NSE. The channel selection principle is that the traffic between IP
NSVCs is balanced.
In Gb over IP mode, the user needs to configure IP NSVCs by using the local NSVL and
remote NSVL objects
Key interworking parameters in IP networking mode:
Table 2.1 Gb over IP interworking parameters
2015-11-13
Parameter
Description
NSEI
Value range: 0 to 65534. It is the same as the

NSEI on the SGSN.
Subnet configuration mode
Static or Dynamic.
Service address of the SGSN
It indicates the IP address and subnet mask of

the remote NSVL.
Service address of the BSC (device IP

address)
It indicates the IP address and subnet mask of

the local NSVL. The device must inform the
SGSN of this parameter.
IP address on the server (control plane)

and UDP port number on the server
This parameter needs to be negotiated when
Huawei Confidential
Page 163 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
(control plane) of the SGSN
the subnet configuration mode is Dynamic.
Address and UDP port number of the

remote NSVL
This parameter needs to be negotiated when

the subnet configuration mode is Static.
IP address and UDP port number of the

local NSVL
The device needs to inform the SGSN of this

parameter when the subnet configuration mode
is Static.
Port IP address
IP address of the port on the board. The

routing gateway IP address on the SGSN side
is the port IP address of the BSC when the
networking mode is direct connection.
In Gb over IP mode, the NSVL configuration modes are static configuration and dynamic
configuration. In static configuration mode, the remote NSVL is configured by using the local
maintenance terminal (LMT) or man-machine language (MML) commands. In dynamic
configuration mode, the remote NSVL is obtained from the SGSN by using the subnet service
process (SNS process) in the 48018 protocol. The main differences between the dynamic
configuration process and the static configuration process are as follows:
When you configure the NSE in dynamic configuration mode, set Subnet configure mode to
Dynamic and configure Server IP and Server UDP Port. Server IP and Server UDP Port
specify the interface corresponding to the remote NSVL (used to communicate with the BSC)
on the SGSN. In static configuration mode, set Subnet configure mode to Static.
In static configuration mode, configure the remote NSVL, whereas in dynamic configuration
mode, the remote NSVL is not required.
Local NSVL:
NSVLI: indicates the local NSVL ID. The value range is 0 to 65534. It needs to be
negotiated with the SGSN.
NSEI: indicates the NSE ID. The value range is 0 to 65534. The NSEs of the same
SGSN are numbered in a unified manner. This parameter needs to be negotiated with the
SGSN.
IP: indicates the IP address. The value must comply with IPv4. It needs to be negotiated
with the SGSN.
UDPPN: indicates the UDP port number. The value range is 0 to 65534. It needs to be
Remote NSVL:
2015-11-13
NSVLI: indicates the local NSVL ID. The value range is 0 to 65534. It needs to be
NSEI: indicates the NSE ID. The value range is 0 to 65534. The NSEs of the same
SGSN are numbered in a unified manner. This parameter needs to be negotiated with the
SGSN.
IP: indicates the IP address. The value must comply with IPv4. It needs to be negotiated
with the SGSN.
UDPPN: indicates the UDP port number. The value range is 0 to 65534. It needs to be
Huawei Confidential
Page 164 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
An NSE specifies a network service entity; it is a collection of the BVC and the IP NSVC. Determine a
protocol type Gb over IP for an NSE. In IP mode, configure the local and remote NSVL objects.
19.4.9 Interworking Instances

Gb over IP instances
Design the basic Gb over IP parameters that need to be negotiated with the SGSN based on
the Gb over IP networking design and IP address planning. The basic negotiation parameters
are as follows:
Table 1.1 SGSN node parameters
SGSN Node
Parameter
Description
SGSN Node ID
It indicates the SGSN number.
Configure Capacity
It indicates the number of subscribers that can access the

NSE. In the case that one BSC can connect to multiple
SGSNs, a subscriber determines the SGSN to access based on
the NSE capacity.
SGSN Management Status
It indicates that when an MS accesses the network initially,

the MS selects the SGSN randomly, causing some SGSNs
overloaded. The three SGSN management states are
uninstalled, normal, and prohibited. In the uninstalled state,
the SGSN cannot provide access for new users; in the normal
state, the SGSN can be used normally; in the prohibited state,
the SGSN cannot be used.
Table 1.2 NSE attribute parameters
2015-11-13
NSE Attribute
Parameter
Description
NSE Identifier
It indicates the NSE ID. An NSE manages a group of NSVCs.
Subrack Number
It indicates the number of the subrack where the end-to-end

communication NSE is located.
Subnet protocol type
In Gb over IP mode, the Gb interface is connected to the IP

network. The IP network in Gb over IP mode takes the place
of the FR connections in Gb over FR mode. In Gb over IP
mode, the subscriber data and signaling data of the Gb
interface are transmitted in UDP packets. In Gb over IP mode,
the functions of the protocol stacks of the NS and upper layers
are completely the same as the functions in Gb over FR mode.
Subnetwork Configure
Mode
It indicates the IP subnet configuration mode. The two

optional modes are static configuration and dynamic
configuration. In static configuration mode, the remote NSVL
can be configured on the LMT. In dynamic configuration
mode, the remote NSVL can be obtained from the SGSN by
Huawei Confidential
Page 165 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
NSE Attribute
Parameter
Description
using the protocol process.
Server IP
It indicates the IP address of the SGSN. If the IP subnet

configuration mode is static configuration, the server IP
address is not required. If the IP subnet configuration mode is
dynamic configuration, configure the server IP address.
Server UPD Port
It indicates the UDP port number of the SGSN. If the IP

subnet configuration mode is static configuration, the server
UDP port number is not required. If the IP subnet
configuration mode is dynamic configuration, configure the
server UDP port number.
RIM Support
It indicates whether the NSE supports the RIM function. If the

RIM function is enabled, the SIGBVC needs to be reset. The
RIM function can be implemented only if the SGSN also
supports the RIM function. Default value: NO.
PFC Support
It indicates whether the NSE supports the PFC function. If the

PFC function is enabled, the SIGBVC needs to be reset. The
PFC function can be implemented only if the SGSN also
supports the PFC function. Default value: NO.
Table 1.3 Local NSVL parameters (configured on the Gb interface board)
2015-11-13
Local NSVL
Parameter
Description
Local NSVL ID
It indicates the NSVL ID on the BSC side. An NSVL ID

uniquely identifies an NSVL. An NSVL is a network service
virtual link.
NSE Index
It indicates the NSE index.
IP Address
It indicates the IP address in the dotted decimal format of the

Gb interface board.
UDP Port
It indicates the UDP port number. It must be consistent with

the UDP port number configured on the SGSN.
Signaling load weight
It indicates the signaling data load sharing. In Gb over IP

mode, the uplink signaling message load sharing involves the
selection of the local IP endpoint and the remote IP endpoint.
The protocol does not describe the selection of the local IP
endpoint in detail, and this process is simplified. All the
available local signaling endpoints are evenly polled to select
a local signaling endpoint.
User Data load weight
The service data load sharing involves the selection of the

local IP endpoint and the remote IP endpoint. All local IP
endpoints are polled to select a local IP endpoint. The value
range is 0 to 255.
Huawei Confidential
Page 166 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Table 1.4 Remote NSVL parameters

Remote NSVL
parameter
Description
Remote NSVL ID
It indicates the NSVL ID on the SGSN side. An NSVL ID

uniquely identifies an NSVL.
NSE Index
It indicates the NSE index.
IP Address
It indicates the IP address of the remote NSVL. You need to

specify the peer IP address only when configuring the remote
NSVL.
UDP Port
It indicates the UDP port number. It must be consistent with

the UDP port number configured on the SGSN.
Signaling load weight
It indicates the signaling data load sharing.

In Gb over IP mode, the uplink signaling message load
sharing involves the selection of the local IP endpoint and the
remote IP endpoint. The protocol does not describe the
selection of the local IP endpoint in detail, and this process is
simplified. All the available local signaling endpoints are
evenly polled to select a local signaling endpoint.
User Data load weight
It indicates the service data load sharing. The service data

load sharing involves the selection of the local IP endpoint
and the remote IP endpoint. All local IP endpoints are polled
to select a local IP endpoint. The value range is 0 to 255.
If the NSE attribute Subnet configure mode is set to Static, configure the remote NSVL; if it is set to
Dynamic, do not configure the remote NSVL.
Table 1.5 PTP BVC attribute parameters

PTP BVC Attribute
Parameter
Description
NSE Index
It indicates the NSE ID.
PTP BVC Identifier
It identifies a unique PTP BVC with NSE.
Cell Name
It indicates the BSC cell name.
Table 1.6 SGSN attribute parameters:
2015-11-13
SGSN Attribute
Parameter
Description
SGSN Name
It indicates the SGSN name.
Huawei Confidential
Page 167 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
IP Address
It indicates the IP address of the SGSN.
In active/standby mode, the IP address of the standby board must be in the same network segment as the
IP address of the active board.
19.5 Abis Interface Design

The Abis interface in an internal interface. The BTS that are provided by different
manufacturers cannot interwork through the Abis interface.
Therefore, in terms of Abis interface interworking, the BSC6910 can interwork only with
Huawei's BTS. The BSC6910 supports only BTS3012 and BTS3900, but does not support the
co-locate deployment of BTS3012 and BTS3900, or the BTSs of BTS3X serials.
The protocols and standards that the Abis interface complies with are as follows:
Basic principles of the Abis interface: 3GPP 48.052
Physical layer: 3GPP 48.054
Data link layer: 3GPP 48.056
Layer 3 signaling procedure: 3GPP 48.058
O&M message transfer mechanism: 3GPP 52.021
BSC code converter/rate adaptation in-band control protocol: 3GPP 48.060
The Abis interface of the GBSS15.0 BSC6910 supports IP over FE/GE, and TDM over STM1. The interface bandwidth and networking mode are closely related to the interface type.
IP transmission mode
Basic principle: UDP/IP bears the CS and PS service, signaling, and O&M messages.
Implementation method:
Packet interfaces boards are added to the BTS and the BSC. On the Abis interface, the PS and
service service/signaling messages are transmitted in IP over FE/GE mode.
Each BTS is configured with an independent logical IP address. Each CS service channel,
RSL, OML, and ESL is allocated with a UDP port number. For the PS service, each TRX is
allocated with a UDP port number.
On the BSC side, a fixed UDP port number is used, and the UDP port number on the BTS side
is used to distinguish CS and PS signaling/O&M messages.
2015-11-13
Huawei Confidential
Page 168 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Abis over HDLC interface protocol

19.5.2.1 Basic Procedure
1. Transmission mode
selection
Abis over TDM
Abis over IP
2. Transmission type
selection
STM-1
IP or FE
3. Networking design
Design the networking based on the site configuration, number of

transmission resources of the operator, and distribution of the
transmission backbone network.
For the operator who provides sufficient TDM transmission

resources, use the simple star networking.
For some VIP BTSs where transmission resources are sufficient

and the security requirements are high, use the ring
networking.
For the operator who purchases the BTS local exchange and
Flex Abis functions and adopts TDM transmission, use the
star or chain (determined by the technical personnel based on
the onsite conditions) networking.
19.5.2.2 Transmission Mode

The Abis interface supports two transmission modes: Abis over TDM and Abis over IP.
2015-11-13
Huawei Confidential
Page 169 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Abis over TDM

In Abis over TDM mode, the Abis interface adopts the traditional TDM mode for
transmission. In this mode, the Abis interface board of the BSC6910 is the POUc, and
the SDH/PDH network is deployed between the BSC and the BTS to provide the
transmission service.
Advantages: The networking solution is proven with the perfect QoS guarantee
mechanism. This mode is secure and reliable. The operator can fully use the existing
PDH/SDH transmission resources.
Disadvantage: The cost of this solution is higher than the cost of the IP transmission
solution.
Abis over IP
In Abis over IP mode, layer 2 of the protocol stack of the Abis interface complies with
the IP protocol. The Abis interface board of the BSC6910 is the
FG2c/GOUc/EXOUa/FG2d/GOUd, and the IP network is deployed between the BSC
and the BTS to provide the transmission service.
Advantages: The bandwidth is sufficient; the cost is low; the evolution capability is high.
Disadvantage: The QoS is difficult to guarantee.
The following is the design guide to the Abis over IP mode:
Design guide:
Step 1 Determine the networking mode (direction connection networking or layer-2/layer-3
networking) based on the requirements of the customer for the Abis interface networking and
the adopted transmission backbone network (MSTP and MPLS/IP). Use the active/standby
boards+manual active/standby LAGs (Pool of single IP address)+layer-3 router networking
mode. Do not use the layer-2 or direct connection mode.
Step 2 If the layer-2/layer-3 networking mode is used, determine the transmission mode. The
optional transmission modes are MSTP transmission, layer-2/layer-3 data network
transmission, satellite link transmission, and XDSL transmission.
Step 3 Design the networking reliability of the interface board based on the capability of the
interworking device and the customer requirements. Use the mode of active/standby
boards+manual active/standby LAGs (Pool of single IP address) for the Abis interface
networking, and do not use the load sharing+router VRRP mode.
Step 4 If the layer-3 data transmission networking mode is used, determine whether to enable the
router to adopt VRRP+VLANIF to improve the reliability in the mode of active/standby
boards.
----End
The Abis over IP mode supports two application scenarios: direct connection and
switch/router networking.
Figure 1.1, Figure 1.2, and Figure 1.3 show the typical networking diagrams.
2015-11-13
Huawei Confidential
Page 170 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 TDM networking when the Abis interface adopts STM-1 transmission
Figure 1.2 IP networking when the Abis adopts MSTP transmission
Figure 1.3 IP networking when the Abis adopts data network transmission
19.5.2.3 Networking Design (TDM)

The main BTS networking types are chain, star, ring, and tree.
2015-11-13
Huawei Confidential
Page 171 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 BTS networking diagram
Each networking type has its own advantages and disadvantages. Configure the networking
based on the requirements of the operator based on the following principles:
Star networking
Advantages: The network structure is simple; expansion is

convenient; the reliability is high. If one BTS is faulty, the other
BTSs are not affected. Therefore, the star networking is widely
used. However, for small-scale BTSs, the transmission usage of the
star networking is low. The timeslot consolidation device can be
used to address this issue.
Chain networking
If the coverage area is in the form of a band (for example, the

highway), and the traffic in the coverage area is light, the chain
networking can be used because if the star networking is used in
such a coverage area, transmission resources are wasted. In chain
networking, the BTSs are cascaded. A BTS on a cascading link
processes only its own timeslots and transparently transmits the
timeslots of downstream BTSs.
The clock preciseness decreases with the number of cascading
levels, and this may degrade the BTS performance. Therefore, the
number of cascading levels cannot be too large. Disadvantage:
Expansion is inconvenient, and the reliability is low. If a
transmission fault occurs on a BTS in the chain, the services of the
downstream BTSs are affected.
Tree networking
In tree networking mode, the BTS that is directly connected to the

BSC is the parent BTS, and the BTSs subordinated to the parent
BTS are children BTSs. The parent BTS completes timeslot
interchange of the children BTSs. Timeslot interchange can be
controlled by using the LMT.
The tree networking combines the characteristics of the star
networking and chain networking, and the reliability of the tree
networking is between these two networking modes. The
disadvantage is that the structure is complicated, which causes
inconvenience for expansion and maintenance.
Ring networking
2015-11-13
The ring networking is a special chain networking. In normal cases,

the BTSs form an ordinary chain, and the last BTS is connected
back to the BSC. In this way, a ring is formed. If a transmission
fault occurs on a BTS on the ring, the networking mode of the
upstream BTSs before the faulty BTS does not change, and the
Huawei Confidential
Page 172 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
downstream BTSs after the faulty BTS form a new chain in a

reverse manner.
The reliability is high, but the transmission resource cost is high. It
is recommended for some VIP sites.
Star networking and configuration principles
In star networking mode, any TDM BTS is directly connected to the Abis TDM
interface board (POUc) on the BSC through fiber and does not need another BTS for
transfer.
A transmission-optimized BTS is directly connected to the Abis interface board on

the BSC through fiber and does not need another BTS for transfer.
An IP BTS is connected to the router or switch through one or two FE ports and does
not need another BTS for forwarding.
Chain networking and configuration principles

In chain networking mode, each BTS receives information from the upper-level BTS and
transmits the information to the lower-level BTS in cascading mode. The chain
networking mode can be single-chain cascading or multi-chain cascading:
Single-chain cascading: If the total number of cascading BTSs is less than 5 and the
total number of TRXs is less than 15, the single-chain cascading mode can be used.
Multi-chain cascading: If the total number of cascading BTSs is less than 5 and the
total number of TRXs is greater than 15, the multi-chain cascading mode can be used.
A BTS supports a maximum of eight E1 ports; therefore, the multi-chain cascading
mode supports a maximum of eight active/standby links.
If the number of TRXs of each BTS is great (greater than six), the chain networking
mode can be used. The number of TRXs allowed in multi-chain cascading mode is
greater than that in single-chain cascading mode.
TDM BTSs are directly connected to the Abis interface board on the BSC through
fiber.
Tree networking and configuration principles
If two BTSs are subordinated to another BTS, the networking mode is the tree
networking mode. The tree networking mode combines the characteristics of the star
and chain networking modes.
If a BTS adopts the single-chain mode to connect to the Abis interface board directly,
this root can connect to a maximum of 15 TRXs. Due to the E1 port restriction (each
TMU provide only eight E1 ports) of the TMU on the BTS, each root can possess a
maximum of seven branches (in the case that two TMUs that work in active/standby
mode are configured for the root BTS).
If a BTS adopts the multi-chain mode to connect to the Abis interface board directly,
the number of TRXs that can connect to this root is determined by the number of E1
timeslots. Due to the E1 port restriction (each TMU provide only eight E1 ports) of
the TMU on the BTS, each root can possess a maximum of six branches (in the case
that two TMUs that work in active/standby mode are configured for the root BTS).
TDM BTSs are directly connected to the Abis interface board on the BSC through E1
links.
Ring networking and configuration principles

In ring networking mode, the Abis interface board and the BTSs form a ring, as shown in
Figure 1.1. In normal cases, the BTSs form a ring network. If a BTS is faulty, the
networking mode of the upstream BTSs before the faulty BTS does not change, and the
2015-11-13
Huawei Confidential
Page 173 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
downstream BTSs after the faulty BTS form a new chain in a reverse manner. Compared
with the chain networking mode, the ring networking mode is more robust. If the ring is
disconnected at a point, it is automatically split into two chains, and the BTSs before and
after the faulty point can work properly.
In ring networking mode, if the BSC6910 is connected to BTSs through

active/standby Abis interface boards, the ports of the Abis interface may be
insufficient. In this case, configure the Abis interface boards of the BSC6910 to work
in non-active/standby mode. Then, the interface boards are counted in single-board
mode instead of active/standby mode, that is, each interface board is counted (not
recommended).
One ring supports the BTS cascading of a maximum of five levels. Each BTS on the
ring can support tributary BTSs. The BTSs on the ring and their tributary BTSs form
a tree network. This networking mode is the same as the tree networking mode. The
number of BTS cascading levels between a tributary BTS and the BSC cannot exceed
five.
OML backup networking

Principles:
On Huawei's 2G site, the OML link can be configured only on the timeslot of port 0. If
the OML link fails, the services of the entire site are interrupted. Therefore, in the
V900R011 version, the OML backup function is planned.
If the user uses the OML backup function when configuring the BTS, the BSC
configures an OML link on timeslot 31 of port 0 and port 1. After the BTS is reset, the
BTS attempts to establish a link on the two ports one by one. Once the OML link is
established on either port, this OML link is always used unless the BTS is reset or the
OML link is disconnected. If the established OML link is disconnected, the BTS
automatically switches to the other port and attempts to establish a link on that port. If
the link is established successfully, the BSC triggers an OML link switchover and
switches the related ESL/EML links to the port where the currently available OML link
is located.
In the OML backup function, the processing of OML/EML/ESL links is similar to the
processing on the ring network, that is, link switchovers can be performed. However, the
processing of carrier RSLs, service channels, idle timeslots, and monitoring timeslots is
different from the processing on the ring network, that is, switchovers are not performed.
In other words, for port 0 and 1 of the BTS, if the port where the currently available
OML link is established is faulty, all the carriers, channels, idle timeslots, and monitoring
timeslots configured on this port become invalid, whereas the OML/ESL/EML links can
be switched over to the other port to ensure that the BTS is not out-of-service and that
the normal port can provide services.
Restrictions:
2015-11-13
The new BTS types (such as the 3900 series) later than the BTS3012 series of
double-density BTSs support this function.
Only the TDM (including the 16 kbit/s and Flex Abis scenarios) supports this
function. The IP networking mode does not support this function.
The OML backup function is mutually exclusive with the ring network and Abis
bypass functions.
OML backup can be implemented only between port 0 and port 1 in the primary
cabinet group and cannot be implemented between other ports or between the
primary cabinet group and the secondary cabinet group.
The two mutually backup OML links (including EML/ESL) between the BSC and the
BTS cannot be located on the same E1/T1 (or on the same E1/T1 on the upper-level
Huawei Confidential
Page 174 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
BTS). That is, if the upper-level BTS has only one E1/T1 connected to the BSC, the
OML backup function can be used only if either of the following conditions is met:
A secondary link to the upper-level BTS is added.
A secondary link (directly connected to the BSC) to the lower-level BTS is added.
Figure 1.2 and Figure 1.3 show the typical networking diagrams.
Figure 1.2 Two E1s connected to different interface boards
Figure 1.3 Two E1s connected to different ports on the same interface board
The reliability of the networking mode where the E1s of a BTS are connected to two pairs of
interface boards is high. Therefore, use this mode.
However, do not connect the E1s of the BTS to the interface boards in different subracks.
19.5.2.4 Networking Design (IP)

Interface networking schemes for GBSS15.0:
Promoted scheme
Networking: Pool of active/standby boards+manual active/standby LAGs+single IP
address
2015-11-13
Huawei Confidential
Page 175 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
CE2.
When ports are added to the internal interface boards of the BSC in this networking
mode, new IP addresses must be added. New IP addresses and VRRPs must be also
Logical IP addresses of each pair of active/standby interface boards of the BSC form an
IP pool.
The BTS is connected to the Ethernet network through a single Ethernet port.
The BSC uses layer-3 networking.
Figure 1.1 Typical A over IP networking mode (pool of active/standby boards+manual
active/standby LAGs+single IP address)
Layer-2 ports connecting CE1 and the BSC and connecting CE2 and the BSC are
configured to the Access mode, and the BSC does not require VLAN tags.
For standalone NEs, O&M data and service data is separated on the interface and a
logical IP address must be configured for O&M data. Non-standalone NEs do not have
this requirement.
e
IP
Hop
Priorit
y
BSC
IP151
IP19
Default
BTS
IP200
IP119
Default
CE1
IP200
IP11
Default
CE2
IP200
IP11
Default

IP pool fault detection and switchover triggering mechanism: The IP pool on the BSC
side starts the UDP ping detection.
The active ports of the active/standby interface boards on the BSC enable two BFD
sessions to detect the physical IP addresses of the two routers. The standby ports of
active/standby interface boards on the BSC enable an ARP detection session. The BSC
2015-11-13
Huawei Confidential
Page 176 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
performs a BFD detection every 100 ms for three times and performs an ARP detection
every 10s for three times. Configure the delay enabling BFD on CE1 and CE2 to avoid
service interruption of CE1 and CE2 due to a reset upon power-off.
CE1 is faulty (suspended or powered off): Heartbeat detection on the VRRP port of
CE2 fails and the standby VRRP port of CE2 becomes active. If the BSC sends
packets to the BTS through CE1, the two BFD detections detect faults, manual
active/standby LAGs switch over, and the BSC sends packets to the BTS through
CE2. If the BSC sends packets to the BTS through CE2, only one BFD session on the
BSC fails and the active/standby ports are not switched over. After CE1 is faulty, the
OSPF re-converges, and the BTS sends packets to the BSC through CE2. Regardless
of whether the BSC sends packets to the BTS through CE1 or CE2, the BSC and the
BTS exchange messages through CE2. In this case, services are not interrupted.
Connection between the BSC and CE1 is faulty: Heartbeat communication on the
VRRP ports is correct and the VRRP ports are not switched over. If the BSC sends
packets to the BTS through CE2, only the standby port is faulty, the active/standby
ports are not switched over, and the original path remains unchanged. If the BSC
sends packets to the BTS through CE1, it then sends packets through CE2: The two
BFD sessions on the active port of the BSC fail, the active/standby ports switch over,
and the path of uplink packets of the BSC is BSC => CE2 => CE1 => BTS. When the
active/standby ports switch over, the BSC sends free ARP to update the ARP entries
on CE1 and CE2. The BTS sends packets to the BSC through CE2. In this case,
packets are sent to the BSC through CE1 and CE2, and services are not interrupted.
Connection between two routers is faulty: BFD detection of heartbeat messages on

VRRP1 fails and the standby port on CE2 becomes active, then VRRP1 has two
active ports. One BFD detection on the active port of the BSC succeeds and the other
BFD detection fails, then the active/standby ports are not switched over. If the BTS
sends packets to the BSC through CE1, the original path remains unchanged. If the
BTS sends packets to the BSC through CE2, then the BTS sends packets to the BSC
through CE2 and CE3: BSC <=> CE2 <=> CE3 <=> BTS. In this case, packet
sending from the BTS to the BSC is not interrupted. If the BTS sends packets to the
BSC through CE1, packet sending is not interrupted. If the BTS sends packets to the
BSC through CE2, packet sending is interrupted. That is, there is 50% probability that
packet sending from the BTS to the BSC may be interrupted (see the following Note).
Connection between CE1 and the intermediate network is faulty: The OSPF reconverges, and services are not interrupted.
Manual switchover of ports on the BSC: VRRP routes and intermediate network
routes are not affected. When the active/standby ports switch over, the BSC sends
free ARP to update the ARP entries on CE1 and CE2. The BTS sends packets to the
BSC through CE2. In this case, services are not interrupted.
Interface boards of the BSC are faulty: The active/standby boards switch over, and
the active/standby ports switch over. In this case, services are not interrupted.
Optional scheme
Networking: Pool of active/standby boards+dual-active ports+single IP address
Logical IP addresses of each pair of active/standby interface boards of the BSC form an
IP pool. The BSC is directly connected to the dual routers through two independent ports
on the active/standby interface boards.
Device IP addresses are configured on the logical active board. With the active/standby
2015-11-13
Huawei Confidential
Page 177 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
When ports are added to the internal interface boards of the BSC in this networking
mode, new IP addresses must be added. New IP addresses and VRRPs must be also
Figure 1.2 Typical A over IP networking mode (pool of active/standby boards+dual-active
ports+single IP address)
Configure IP addresses in the same network segment for each sub-interface, which
facilitates route combination and simplifies intermediate network route. To facilitate
network expansion, service IP addresses of the BSC use logical IP addresses (device IP
addresses).
Devic Destination IP
e
Next Hop
Priority
BSC
IP151
IP10
High
IP151
IP20
Low
IP161
IP20
High
IP161
IP10
Low
BTS1
IP210
IP119
Default
BTS6
IP220
IP129
Default
CE1
IP210
IP11
High
IP220
IP11
Low
IP210
IP21
Low
IP220
IP21
High
CE2
Local service IP addresses and peer service IP addresses are grouped in two. With route
priority configuration, IP addresses in the two groups have different priorities, thereby
2015-11-13
Huawei Confidential
Page 178 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
implementing load-sharing. In addition, backup routes are configured to ensure

reliability.
Dynamic route protocols (OSPF/ISIS) need to be configured between CE1 or CE2 and
intermediate bearer networks. Static routes also need to be configured. In addition, route
priorities need to be configured to ensure that the route between CE1 and IP150 has
higher priority than the route between CE2 and IP150, and the route between CE2 and
IP170 has higher priority than the route between CE1 and IP170. In this case, data is sent
to CE1 through IP150 and sent to CE2 through IP170.
Detection mechanism
IP pool fault detection and switchover triggering mechanism: The IP pool on the BSC
side starts the UDP ping detection.
Data transmission path in normal cases for BTS1: BSC -> CE1 -> CE3 -> BTS1; Data
transmission path in normal cases for BTS6: BSC -> CE2 -> CE4 -> BTS6.
Connection between the BSC and CE1 is faulty: SBFD detection on the route from
the BSC to IP10 of CE1 fails and the active route whose next hop is IP10 becomes
invalid. Then the active/standby routes switch over. The standby route whose next
hop is IP20 becomes valid and services migrated to this route. At the same time, the
static route that bound to the SBFD on CE1 becomes invalid, and the OSPF reconverges. The next hop of the route between CE3 and IP210 switches to CE2 from
CE1 and the BTS sends packets to the BSC only through CE2. In this case, services
are not interrupted.
Connection between two routers is faulty: Services are not affected because no data is
transmitted between CE1 and CE2 in normal cases.
Connection between CE1 and CE3 is faulty: The OSPF re-converges. The next hop of
the route between CE3 and IP210 switches to CE2 and the next hop of the route
between CE1 and IP111 switches to CE3 from CE2. That is, the BSC sends packets
to the BTS through this path: BSC -> CE1 -> CE2 -> CE3 -> BTS1. In this case,
services are not interrupted.
CE1 is faulty (suspended or powered off): SBFD detection on the route from the BSC
to CE1 fails and the active route whose next hop is IP10 becomes invalid. Then the
active/standby routes switch over. The standby route whose next hop is IP20 becomes
valid and the BTS sends packets to the BSC through CE2. At the same time, the BFD
for OSPF detection on CE3 indicates that CE1 is faulty, and the OSPF re-converges.
The next hop of the route to IP210 switch to CE2 from CE1 and the BTS sends
packets to the BSC through CE2. In this case, services are not interrupted.
Therefore, data sent to and received by one BTS is always through a certain port. The IP
PM function can be used correctly.
Capacity restriction after the switchover is not taken into consideration.
In this scheme, the assumed convergence time of the intermediate network is less than 1s.
2015-11-13
Huawei Confidential
Page 179 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Active/standby scheme on the BTS side (do not use this scheme unless it is approved by
the R&D department):
Basic principle:
In IP transmission mode, configure two FE/GE ports on the TMU of the BTS to work in
active/standby mode. This scheme is specific to the customer who has high reliability
requirements.
In the active/standby port mode, use the logical IP address for BTS communication.
Configure two IP addresses for the Ethernet ports, and configure active and standby
routes for the uplink outgoing interface of the BTS.
Configure the ARP sessions for the two next hops of the BTS, and set the session type to
reliable session (enabling and disabling the related route based on the ARP session).

When designing Abis interface bandwidth, take certain redundancy into consideration for
subsequent expansion. Determine the redundancy with the operator. Use a redundancy
capacity of about 20%.
19.5.3.1 Abis over TDM

For the detailed bandwidth calculation formula, see section Error: Reference source not
found.
The following description will help you understand timeslot allocation principles for the Abis
interface.
The formula for calculating interface bandwidth is as follows:
Total BTS bandwidth = OML bandwidth + RSL bandwidth + TCH bandwidth + PDCH
bandwidth + idle timeslot bandwidth
Timeslot multiplexing of the Abis interface belongs to statistical multiplexing. Multiple links
share the 64 kbit/s bandwidth.
Basic principles are as follows:
2015-11-13
A TCH occupies a 16 kbit/s timeslot.
A PDCH occupies a 16 kbit/s timeslot.
Each idle timeslot is 16 kbit/s.
Signaling timeslots and service timeslots (TCH timeslot, PDCH timeslot, and idle
timeslot) are multiplexed on a 64 kbit/s timeslot.
Huawei Confidential
Page 180 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Based on different multiplex ratios of the Abis interface, the bandwidth calculation formulas
are as follows:
1:1 OML = 64 kbit/s
RSL = 64 kbit/s
2:1 OML + RSL = 64 kbit/s 2 x RSL = 64 kbit/s Remaining RSL = 64 kbit/
3:1 OML + 2 x RSL = 64 kbit/s 3 x RSL = 64 kbit/s Remaining RSLs (less than 3) =
64 kbit/s
4:1 OML + 3 x RSL = 64 kbit/s 4 x RSL = 64 kbit/s Remaining RSLs (less than 4) =
64 kbit/s
5:1 OML + 2 x RSL + ESL = 64 kbit/s 5 x RSL = 64 kbit/s Remaining RSLs (less
than 5) = 64 kbit/s
(the Flex Abis function must be enabled.)
6:1 OML + 2 x RSL + ESL = 64 kbit/s 6 x RSL = 64 kbit/s Remaining RSLs (less
than 6) = 64 kbit/s
(the Flex Abis function must be enabled.)
16 kbit/s mode: OML = 16 kbit/s RSL = 16 kbit/s
For the detailed timeslot distribution, see the BTS timeslot distribution on the LMT.
For a single site, use the GSM ENP to calculate Abis transmission timeslots.
19.5.3.2 Abis over IP

For the detailed bandwidth calculation formula, see section Error: Reference source not
found.
Design of the Multiplex Ratio of Abis Signaling Links (TDM)
Aiming to improve the utilization rate of transmission resources over the Abis interface,
statistical multiplexing is introduced to Abis RSLs. This technology allows signals over
multiple RSLs to be multiplexed into one LAPD link for transmission. The ratio of the
number of multiplexed RSLs to the number of actually occupied LAPD links is called the
multiplex ratio. For example, if signals over N RSLs are transmitted over one LAPD link, the
multiplex ratio is N:1. The multiplex ratio is configured as a parameter for the BTS. The ratio
of the multiplex ratio to the multiplex ratio related to the BTS is called the BTS multiplex
ratio.
2015-11-13
Huawei Confidential
Page 181 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The multiplex ratio of Abis signaling links determines the bandwidth of each Abis signaling
link. A proper multiplex ratio for a network helps improve the utilization rate of transmission
bandwidth. The recommended BTS multiplex ratio used in common situations is as follows:
If the half-rate transmission is disabled, the recommended ratio is 4:1 or 2:1.
If the half-rate transmission is enabled, the recommended ratio is 2:1 or 1:1.
To calculate the multiplex ratio, use the following formula:

BTS multiplex ratio = 64 x 1024 x Payload rate of LAPD links/8/RSL bandwidth
Round off the calculation result of the preceding formula to the nearest integer. The result is
the BTS multiplex ratio. For example, if the calculation result value obtained is 2.6, the BTS
multiplex ratio is 2:1. Note that when 16 kbit/s signaling links are used on the Abis interface,
multiplexing is not applicable, and the multiplex ratio is fixedly 1.
The following describes the fields in the preceding formula:
2015-11-13
The number 64 indicates that the bandwidth of each LAPD link is 64 kbit/s.
The number 1024 indicates that 1024 bits equal to one k.
The number 8 indicates that eight bits equal to one byte.
The payload rate of LAPD links ranges from 70% to 75%.
The RSL bandwidth is calculated as follows: RSL bandwidth = Average number of peakhour messages for a TRX/3600.
Average number of peak-hour messages for a TRX = [a x (Number of bytes in peakhour call-related messages + Number of bytes in handover-related messages +
Number of peak-hour messages for call measurement reports) + b x Number of bytes
in peak-hour messages for location updates + c x Number of bytes in received and
sent peak-hour messages + d x Average number of bytes in peak-hour pagings] x
Number of subscribers supported by each TRX
Number of bytes in peak-hour call-related messages = Number of originating peakhour calls x Average number of bytes in messages for an originating call + Number of
received peak-hour calls x Average number of bytes in messages for a received call
Number of bytes in handover-related messages = Number of intra-BSC handovers x

Average number of bytes in messages for an intra-BSC handover + Number of interBSC handovers x Average number of bytes in messages for an inter-BSC handover
Number of peak-hour messages for call measurement reports = (Number of

originating peak-hour calls + Number of received peak-hour calls) x Average number
of measurement reports for a call x Average number of bytes in a measurement report
Number of bytes in peak-hour messages for location updates = Number of peak-hour

location updates x Average number of bytes in messages for a location update
Number of bytes in received and sent peak-hour messages = Number of sent peakhour short messages x Average number of bytes in messages for a mobile originating
short message + Number of received peak-hour short messages x Average number of
bytes in messages for a mobile terminating short message
Average number of bytes in peak-hour pagings = Number of received peak-hour calls

+ Number of received peak-hour short messages + Average number of re-pagings) x
Average number of bytes in messages for a paging
Number of subscribers supported by each TRX = Traffic volume of the site/Peakhour traffic volume for each subscriber/Number of TRXs for a site
Average number of measurement reports for a call = Average duration of a call/0.5
Huawei Confidential
Page 182 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Average number of re-pagings = (Number of received peak-hour calls + Number of

received peak-hour short messages) x Average ratio of re-pagings to total pagings.
The value of the average ratio of re-pagings to total pagings ranges from 0.2 to 0.4
and the most used value is 0.35.
Table 1.1 describes the performance test results of parameters a, b, c, and d.

Table 1.1 Performance test results of parameters a, b, c, and d
Service
a: Call (No Paging)
b: Location Update
c: Short Message (No Paging)
d: Paging
Weight
100%
60%
80%
1%
Table 1.2 lists estimates of data related to parameters a, b, c, and d.

Table 1.2 Estimates of data related to parameters a, b, c, and d
Parameter
Value
Average number of bytes in messages for an originating call
230
Average number of bytes in messages for a received call
240
Average number of bytes in messages for an intra-BSC handover
80
Average number of bytes in messages for an inter-BSC handover
80
Average number of bytes in a measurement report
60
Average number of bytes in messages for a location update
90
Average number of bytes in messages for a mobile originating short message
220
Average number of bytes in messages for a mobile terminating short message
220
Average number of bytes in messages for a paging
16
19.5.4 IP Address Planning

The interface board of the BSC is configured with the device IP address (also called logical IP
address) and port IP address. Multiple port IP addresses can be configured. The packet
interface unit (PIU) is the common name of the A-interface board, Gb interface board, and
Abis interface board.
Design principles:
IP addresses can facilitate future maintenance.
The planned IP addresses can meet future expansion requirements.
The planned IP addresses are added with different VLAN tags based on different
destination IP addresses for easy maintenance and expansion.
Design guide:
2015-11-13
Huawei Confidential
Page 183 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Step 1 Select a communication mode (port IP communication or device IP communication). For the
best results, adopt device IP communication for the Abis interface board on the BSC side (if a
device IP address is configured) and port IP communication for the interface board (TMU
board) on the BTS side.
Configure and use the port IP address or device IP address. The system automatically
identifies the communication mode (port IP communication or device IP communication). The
processing of the A interface and Gb interface is the same as the processing of the Abis
interface.
Step 2 Allocate device IP addresses according to the networking design scheme for the Abis interface
and the calculated number of FE/GE ports of the Abis interface on the BSC side in the
bandwidth design.
Step 3 Allocate port IP addresses. In active/standby boards+manual active/standby LAGs, configure
a group IP address (IP address of the active port) for the trunk. In addition, configure an IP
address for the standby port during ARP detection. The IP address of the active port must be
in the same network segment as that of the standby port.
Step 4 If the layer-3 networking and device IP communication are adopted, configure a route to the
device IP address of the BSC for the intermediate router. If the layer-3 networking and port IP
communication are adopted, do not configure a route to the port IP address of the BSC for the
intermediate device.
Step 5 If the BSC is required to add VLAN tags according to the next hop or service type in the E2E
solution, configure the function of adding VLAN tags according to the next hop or service
type on the BSC side.
----End
2015-11-13
The device IP address is a logical IP address that a board uses for communication. The
device IP address is valid for all the port IP addresses of the board. Use the
active/standby board networking mode. Use the device IP communication mode for the
Abis interface board on the BSC side, and the port IP communication mode for the Abis
interface board on the BTS side.
When the FE interface is used, one board can be configured with eight port IP addresses
that are in different network segments. When the GE interface is used, one board can be
configured with two port IP addresses. Port IP addresses must be in different network
segments from device IP addresses. When the PIU works in active/standby mode, the
port IP addresses of the active and standby PIUs must be in the same network segment.
One physical port can be configured with a maximum of six IP addresses that must be in
different network segments.
If the physical address of the PIU on the BSC side is in the same network segment as that
of the PTU/GTMU on the BTS side, layer-2 interworking is allowed.
If the physical address of the PIU on the BSC side is in a different network segment from
that of the PTU/GTMU on the BTS side, layer-2 interworking is not supported, and a
layer-3 device (for example, layer-3 switch or router) is required for routing. This is
called layer-3 networking.
The gateway IP address must be in the same network segment as the port IP address of
the board.
The IP address cannot be all 0s or all 255s.
The IP address cannot be a loopback address whose network number is 127.x.x.x.
Huawei Confidential
Page 184 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The IP address cannot be the multicast IP address of category D, that is, address in the
range of 224.0.0.0 to 239.255.255.255.
The IP address cannot be the reserved IP address of category E, that is, address in the
range of 240.0.0.0 to 247.255.255.255.
Output: IP address planning in Abis over IP mode

BSC Attribute (for Abis over IP communication method)
Abis IP Type
Abis IP Type: communication mode (logical IP communication or port IP communication)
that the Abis interface adopts in Abis over IP mode.
Device IP (logical IP) of BSC side board
Subrack
No.
Slot
No.
Sub System
No.
Device IP
Address
Subnet
Mask
Ethernet port IP of BSC side board

Subra
ck No.
Slot
No.
Por
t
No.
IP
Address
Index
Port IP
Addres
s
Port
Standby IP
Address
Subnet
Mask
IP VLAN
Subrack No.
Slot No.
Dest IP Address
VLAN ID
Dest IP Address: next hop IP address of the destination BTS.
VLAN ID: VLAN ID of the next hop of the BSC port corresponding to the destination
IP address, that is, VLAN ID carried in the IP packet.
BTS Information
2015-11-13
Huawei Confidential
Page 185 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Site
Inde
x
Site
Nam
e
Site
Typ
e
Upperlevel
Port No
Servi
ce
Mode
Site IP
Addre
ss
Site IP
Subne
t
Auto
Negotiati
on Mode
Mask
Port
Rate(M)
Slot
No.
(BSC)
Dupl
ex
Mode
Port
No.
(BSC)
MTU
BSC
IP
BTS Bar
Code
Reference
Clock
Source
type
BSC IP
Mask
Activity
State
IP
Clock
Port
Subrac
k No.
(BSC)
Longitu
de
Latit
ude
2015-11-13
Site Index: index number of a site. This parameter uniquely identifies a site in a BSC.
Site Name: name of a site.
Site Type: BTS type of a site.
Upper-level Port No: upper-level port number. In Abis over IP mode, only the star
networking is supported, and therefore this parameter is set to the outgoing port number
of the directly connected BSC.
Service Mode: service type. In Abis over IP mode, this parameter is set to IP.
Site IP Address: IP address allocated to a site.
Site IP Subnet Mask: subnet mask allocated to a site.
Auto Negotiation Mode, Port Rate(M), Duplex Mode, and MTU: must be negotiated
with the device directly connected to the IP interface board of the BTS.
BTS Bar Code: electronic label of the BTS. This parameter must be correctly set;
otherwise, the BTS cannot be started.
Reference Clock Source type: type of the reference clock source of the BTS. Values are
IP Time, Trace Transport Clock, Transport Clock, Internal Clock, External sync.
Clock, and Trace GPS Clock. Set this parameter to IP Time if the IP clock server is
used to provide clocks and Trace GPS Clock if the GPS is used to provide clocks.
IP Clock Port: indicates that the BTS uses the IP clock server as the clock source. When
the BTS functions as the client, this parameter is constantly set to 33003 and cannot be
modified.
Subrack No.(BSC): subrack number of the interface board of the BSC connected to the
BTS.
Slot No.(BSC): slot number of the interface board of the BSC connected to the BTS.
Huawei Confidential
Page 186 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Port No.(BSC): port number of the interface board of the BSC connected to the BTS.
BSC IP: port IP addresses (destination IP address) of the interface board of the BSC
connected to the BTS.
19.5.5 Routing Planning

In Abis over IP mode, only static routes are supported. All routes need to be configured
manually.
Design principles:
The designed routing solution must facilitate follow-up maintenance.
The route to the BTS needs to be configured on the BSC.
When the BTS adopts the device IP address for communication, a route to the BTS needs
to be configured on the BSC.
When the BTS adopts the port IP address for communication and the layer-2 networking
is available, no route to the BTS needs to be configured on the BSC.
When the BTS adopts the port IP address for communication and the layer-3 networking
is available, a route to the BTS needs to be configured on the BSC.
The route to the BSC cannot be configured on the BTS. The BTS calculates the route to
the BSC according to the information delivered by the BSC.
Design guide:
Step 1 Plan routes for the BSC according to the networking design scheme for the Abis interface and
the reliability design.
Step 2 If the layer-3 networking is adopted and the BSC uses the device IP address for
communication, configure the route to the device IP address of the BSC on the intermediate
router.
Step 3 If the layer-3 networking is adopted and the BTS uses the device IP address for
communication, configure the route to the device IP address of the BTS on the intermediate
router.
----End
When the BTS adopts port IP addresses to provide services, the BTS does not support
load sharing. Instead, only one FE port IP address can be configured, and the port IP
address is the same as the logical IP address. In layer-2 networking, no route to the BTS
needs to be configured on the BSC because the physical address of the FE port of the
PIU (IP interface board on the BSC side) is in the same network segment as that of the
FE port of the PTU (IP interface board on the BTS side).
When the BTS is configured with logical IP addresses for providing services, the BTS
can be configured with two FE port IP addresses for load sharing. In this case, configure
two FE port IP addresses (they must be in different network segments), one logical IP
address (that is, the IP address of the BTS), and a route from the BSC to the logical IP
address of the BTS, with the next hop being the physical IP address and the destination
address being the logical address of the PTU.
Set Destination IP to a network address in the network segment of the IP address of the
BTS.
During communication, a network address is used to address the peer device. The
network address is obtained by performing the AND operation on the peer IP address and
2015-11-13
Huawei Confidential
Page 187 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
the subnet mask. For example, if the peer IP address is 192.168.80.2 and the subnet mask
is 255.255.255.128, the network address is 192.168.80.0.
Set Subnet Mask to the subnet mask of the destination device.
Set Gateway Address to an address that is in the same network segment as the IP
address of the related port of the interface board on the BSC side.
The DHCP relay is configured on the router connected to the BTS, and no configuration
is required on the BSC port.
Output: Abis over IP routing design

Subrack
No.
Slot No.
Destination
IP Address
Subnet
Mask
Gateway
Route
Priority
Destination IP Address: network IP address of the destination IP address of the peer

BTS (destination of the data from the Abis interface board of the BSC). If the peer BTS
does not have a device IP address (logical IP address), this parameter indicates the
network IP address of the port IP address. The network IP address is obtained by
performing the AND operation on the device IP address (or port IP address if no device
IP address is available) of the BTS and the subnet mask.
Subnet Mask: subnet mask of the IP address of the BTS of the peer device.
Gateway: port IP address of the device directly connected to the outgoing port of the Gb
interface board on the BSC. The IP address specified by this parameter must be in the
same network segment as the IP address of the outgoing port of the Gb interface board
on the BSC.
Route Priority: route priority. The default value is 1.
19.5.6 QoS Design

Design principles:
Port link detection
The BFD detection of the interface and ARP link detection cannot be enabled at the same
time.
One port can be configured with only one detection mode. When a port is configured
with neither BFD detection nor ARP link detection, physical layer detection is adopted.
The following detection modes are supported: BFD detection for the active port, ARP
link detection for both active and standby ports, BFD detection for the active port and
physical layer detection for the standby mode, ARP link detection for the active port and
physical layer detection for the standby mode, and physical layer detection for both
active and standby ports.
ARP link detection and physical layer detection are mainly used. Physical layer detection
does not need to be configured and is supported by all ports by default. Configure the
retry attempts for ARP link detection to 3 with 300 ms per attempt.
Design guide:
Step 1 Design the port QoS attribute parameters based on the capability of the interworking device.
2015-11-13
Huawei Confidential
Page 188 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Step 2 Determine the port link detection mode (BFD detection, ARP link detection, or physical layer
detection) according to the support capabilities of the interconnected device.
Step 3 Design IP addresses and VLANs.
Step 4 Design VLAN priorities and DSCP mappings.
Huawei GBSS provides the same QoS assurance mechanism for Abis over IP transmission
and A over IP transmission to provide E2E QoS assurance, including the physical layer, link
layer, IP layer, and application layer.
----End
Output: Abis over IP QoS design
ARP link detection
Subrac
k No.
Slo
t
No
.
Por
t
No.
IP
Addre
ss
Index
Peer
IP
Addre
ss
Arp
Retry
Attem
pts
Arp
Timeo
ut
VLA
N
Flag
VLA
N ID
Peer IP
Address of
the
Standby
Board
Port No.: port number of the Abis interface board that requires the physical link
detection.
IP Address Index: index of an IP address. The system supports the configuration of

multiple IP addresses for a port.
port.
BFD detection
2015-11-13
Huawei Confidential
Page 189 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Subrac
k No.
Slot
No.
Port
No.
IP
Addres
s Index
Peer IP
Addres
s
MinTx
Interv
al(ms)
MinRx
Interv
al(ms)
Detec
t Mult
Peer IP
Address of
the
Standby
Board
system sends.
system receives.
Logic Port
Subrac Slo
k No.
t
No
.
Physic
al Port
No.
Logi
c
Port
No.
Bandwidt
h of the
Logical
Port(32Kp
bs)
Reserved
Bandwidth
Threshold(
%)
Congestio
n
Bandwidth
Threshold(
%)
Congestion
Clear
Bandwidth
Threshold(
%)
Physical Port No.: physical port number of the interface board to which the logical port
belongs.
Bandwidth of the Logical Port(32Kpbs): fixed bandwidth of the logical port. It ranges
from 32 kbit/s to 64 kbit/s. The sum of the bandwidths of all the logical ports bound to
the same physical port cannot exceed the bandwidth of the physical port.
Reserved Bandwidth Threshold(%): reserved threshold of the logical port, that is, the
percentage of the logical port reserved bandwidth to the logical port bandwidth. The
default value is 95.
Congestion Bandwidth Threshold(%): congestion threshold of the logical port, that is,
the percentage of the logical port congestion bandwidth to the logical port bandwidth.
The default value is 85.
Congestion Clear Bandwidth Threshold(%): congestion clearance threshold of the

logical port, that is, the percentage of the congestion clearance bandwidth to the logical
port bandwidth. The default value is 75.
BSC ABIS MUX
2015-11-13
Huawei Confidential
Page 190 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
ABIS
MUX
Stat
e
Subr
ack
No.
Slo
t
No.
Sit
e
ID
Servi
ce
Type
Multiplexi
ng
SubFrame
Threshold
Multiplexi
ng Packet
Length
Threshold
Time
Out(0.
1ms)
ABIS MUX State: whether to enable the Abis MUX function.
Site ID: ID of a site, which is unique in a BSC.
Service Type: service type of the Abis MUX function on the IP-based interface board.
Service types are as follows: OML service, RSL service, EML service, ESL service, CS
voice service, CS data service, PS service (high priority), and PS service (low priority).
Multiplexing SubFrame Threshold: multiplexing subframe threshold. If the data

streams on the same MUX channel are multiplexed, their multiplexing types must be the
same and the packet length before the multiplexing cannot exceed the multiplexing
subframe threshold.
Multiplexing Packet Length Threshold: threshold for the length of the multiplexed
packet. The packet length after the multiplexing cannot exceed this parameter value. If
the packet length after the multiplexing exceeds this parameter value, the packet is
directly sent and no subracks are added. This parameter value refers to the payload,
excluding the IP/UDP header.
Time Out(0.1ms): maximum multiplexing waiting time. When no content is added to

the multiplexed packet within the time specified by this parameter, the timer expires and
the packet is directly sent. The duration of the timer depends on the average number of
packets to be multiplexed within the timer. The more the number of packets to be
multiplexed, the longer the duration of the timer.
BSC ABIS MUX is optional. Set it when the Abis MUX technology is adopted to improve the IP
transmission efficiency of the Abis interface. The Abis MUX function is available when only the
GFGUB board is configured with the BTS supporting IP transmission, the PTU of the BTS is configured
with the BTS Abis MUX function, and the GFGUB board is configured with the BSC Abis MUX
function. The Abis MUX function is valid only when the Abis MUX function is enabled on the BSC and
BTS sides at the same time. The parameters for the BTS Abis MUX function are the same as those for
the BSC Abis MUX function.
19.5.7 Abis Port Allocation Design
Subrack-based Abis port planning by LAC

To minimize the inter-subrack signaling traffic caused due to inter-cell handover and
paging forwarding, plan the BTSs in the same LAC to the same BM subrack as possible
as you can.
Discontinuous BTS distribution in a subrack (optional)

The BTSs in a subrack can be distributed between boards in a discontinuous manner.
Overlapping coverage exists between adjacent cells. Based on discontinuous BTS
distribution, adjacent BTSs are distributed to different Abis interface boards. When a
board is faulty, the BTS under it is out of service but the overlapping coverage of the
peripheral cells can still ensure services to a certain degree. This can minimize the
impacts of board faults.
2015-11-13
Huawei Confidential
Page 191 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The BTS must be evenly distributed to different interface boards based on the station
module to ensure load balance among boards.
Continuous BTS distribution in a subrack

BTSs are continuously distributed in a BM subrack according to the longitude and
latitude. Select this principle or the principle of discontinuous BTS distribution in a
subrack according to the actual situation.
The BTS must be evenly distributed to different interface boards based on the station
module to ensure load balance among boards.
Batch site establishment

For certain projects, sites need to be established in batches due to transmission providing
capabilities, engineering implementation capabilities and customer requirements. Based
on special requirements, the site distribution strategy can be adjusted. For the best
results, abide by the principle of subrack-based Abis port planning by LAC.
19.6 Abis Interface Design (IP over E1)

The interface description is the same as that in section 19.5.1"Interface Description."
Only the POUc board (IP over STM-1) of the BSC6910 supports IP over E1 over the A
interface.

The IP over E1 networking over the Abis interface can be classified into two types: noncascading IP over E1 networking and cascading IP over E1 networking.
Non-Cascading IP over E1 Networking

In the non-cascading IP over E1 networking mode, the BTS directly uses the IP over E1 transmission to
communicate with the BSC. This networking includes the following scenarios:
2015-11-13
Direct connection between the BTS and BSC using IP over E1 (the BTS is considered
directly connected to the BSC although an SDH network is deployed between the BTS
and BSC)
Huawei Confidential
Page 192 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Direction connection between the BTS and BSC using IP over E1
Channelized STM-1 ports are the only available physical ports on the BSC side.
Cascading IP over E1 Networking

In the cascading IP over E1 networking mode, BTSs in IP over E1 mode may be cascaded with those in
TDM mode. This networking includes the following scenarios:
Chain networking when all the BTSs use IP over E1 transmission
Figure 1.2 Chain networking when all the BTSs use IP over E1 transmission
PPP links are terminated between two BTSs. The intermediate BTSs work as the routes for
forwarding the traffic of lower-level BTSs to the BSC, and the routes to destination IP
addresses of all BTSs must be configured on the BSC.
2015-11-13
Tree networking when all the BTSs use IP over E1 transmission
Huawei Confidential
Page 193 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.3 Tree networking when all the BTSs use IP over E1 transmission
PPP links are terminated between two BTSs. The intermediate BTSs work as the routes for forwarding
the traffic of lower-level BTSs to the BSC, and the routes to destination IP addresses of all BTSs must be
configured on the BSC.
The BTSs that use IP over E1 over the Abis interface can be cascaded. In this situation, the
upper-level BTSs must provide the DHCP relay function for lower-level BTSs so that device
IP addresses (logical IP addresses) can be obtained using the DHCP and used for
communication after PPP or MP negotiation is successful.
Among cascaded BTSs using IP over E1, the upper-level BTSs must work as the routes for
forwarding packets of lower-level BTSs. Otherwise, when the BTSs are cascaded, leaf or
intermediate BTSs cannot establish communications with the BSC.
19.6.3 Transmission Bandwidth Design

For details about how to calculate the transmission bandwidth, see 10.2.2 Abis Interface.
19.6.4 Configuration Principles

In IP over E1 mode, E1 links can be configured as PPP links or an MP. Generally, a single E1
link can be configured as a PPP link. Multiple E1 links can be configured as an MP. A single
E1 link which may be expanded subsequently can be configured as an MP. This helps
subsequent capacity expansion adjustment.
The Abis on the BSC6900 uses a resource pool and therefore IP paths are not required.
2015-11-13
Huawei Confidential
Page 194 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
19.6.5 IP Planning
In IP over E1 transmission mode, both the BTS and BSC support the device IP (using the
local device IP) and port IP addresses.
The device IP address is recommended on the BSC side (the MP or PPP links use the local
device IP). This saves IP addresses. The port IP address is recommended on the BTS.
19.6.6 Route Planning

The BSC is directly connected to the BTS over the PPP/MP.
If the BTS IP address and the BSC IP address are included in the MP group. Routes are not
required on both the uplink and downlink.
If the BTS uses a logical IP address for communication, the BSC must be configured with a
route from the BTS to the logical IP address (the next-hop IP address is the PPP or MP IP
address on the BTS side).
If the BSC IP address is the same as a specified DEVIP, the BTS must be configured with a
route to the BSC DEVIP by running the ADD BTSIPRT command with Route Type set to
OUTIF.
19.6.7 QoS Planning

For details, see Abis Interface Configuration Specification_IP(GBSS17.0) and A&GB
Interface Configuration Specification_IP(GBSS17.0).
http://support.huawei.com/support/pages/kbcenter/view/product.do?
actionFlag=detailProductSimple&web_doc_id=SC0000783702&doc_type=1232&doc_type=123-2/support/pages/kbcenter/view/product.do?
actionFlag=detailProductSimple&web_doc_id=SC0000783702&doc_type=123-2
19.6.8 Clock Synchronization

In IP over E1 transmission mode, the BTS can extract a line clock from E1 links, and the line
clock is used as the clock reference source.
2015-11-13
Huawei Confidential
Page 195 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
19.7 Lb Interface Design

Lb is a standard interface between the BSC and the Serving Mobile Location Center (SMLC).
The BSC provides the LCS for subscribers over the Lb interface by using the external SMLC.
LCSs can increase operators' revenues. Operators can provide various LCSs for subscribers
based on subscribers' locations. LCSs include weather forecasts, trip scheduling, emergency
assistance, stock information, business planning, and transportation information.
With the Lb Interface feature, Huawei GSM BSS equipment can be connected to the SMLC
and LMU (TypeB) of other vendors to provide the LCS in CellID+TA, or AGPS mode. The
Lb interface complies with the 3GPP TS 48.071, 3GPP TS 49.031, 3GPP TS 44.031, and
3GPP TS 03.71.
The SMLC selects a positioning mode, manages the positioning process, and estimates the
location of an MS based on the measurement results reported by the MS.
Figure 1.1 shows the SMLC-based network topology for the Lb interface.
Figure 1.1 SMLC-based network topology for the Lb interface
Huawei's BSS supports message tracing over the Lb interface and can provide LCS
performance measurement entities.
Huawei BSS supports flow control on LCS services. When the external SMLC is overloaded
or the number of LCS requests received by the BSC exceeds the maximum limit, the BSC
rejects some LCS requests to ensure the correct running of the GPS.
19.7.2 Function Interaction

The LCS service is mutually exclusive with the following functions:
GBFD-115401 NSS-Based LCS (Cell ID+TA)
GBFD-115402 BSS-Based LCS (Cell ID+TA)
GBFD-115403 Simple Mode LCS(Cell ID+TA)
The AGPS positioning method requires the support the cell phone and the core network must
support the LCS service.
2015-11-13
Huawei Confidential
Page 196 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
19.7.3 Constraints and Limitations
When the external SMLC uses the IP transmission mode, the Lb interface can be
configured only on an IP-based A or Abis interface board because the A interface of the
BSC6910 supports only the IP transmission mode.
When the BSC does not use RAN Sharing, each BSC can be connected to only one
SMLC.
When the BSC uses RAN Sharing, each BSC can be connected to a maximum of four
SMLCs. One operator can be configured with only one SMLC.
Each SMLC can be configured with at most one DSP or M3UA destination entity. The
Lb interface and the A interface must use the same network indicator.
When the Lb interface uses the IP transmission mode, each SMLC can be configured
with a maximum of 16 MTP3-User Adaptation Layer (M3UA) links.
When the Lb interface uses the IP transmission mode and the STP, the Lb interface must
use a different STP from the A interface.
The Lb interface on the BSC must be configured in the same subrack as the inter-BSC
connection if the BSC is configured with the IP-based Lb interface and the IP-based
inter-BSC connection.

The SMLC and BSC use IP to communicate with each other. In IP transmission mode, set
DPCT of the destination signaling point (DSP) to LB(LB), set OPNAME to specify the
name of the operator to which the SMLC belongs, and set DESTSSN to specify the
subsystem number of the peer SMLC.
The SMLC can be connected to the BSC either directly or through the signaling transfer point
(STP).
Direct connection
The BSC is directly connected to the SMLC in IP transmission mode, as shown in Figure
1.1.
Figure 1.1 Direct connection between the BSC and the SMLC
2015-11-13
Huawei Confidential
Page 197 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Connection through STP

The BSC is connected to the SMLC through the STP, and IP can be used for the
intermediate transmission, as shown in Figure 1.2.
Figure 1.2 Connection through STP
19.7.5 Positioning Modes
CellID+TA mode
The CellID+TA positioning method estimates the location of an MS based on the timing
advance (TA) value reported by the MS. In CellID+TA mode, the SMLC needs to
exchange BSSAP-LE-layer signaling with the BSC. The signaling interaction procedure
is as follows:
AGPS mode
In the AGPS positioning method, the SMLC locates an MS by using GPS and exchanges
the positioning assistance information with the MS.

When the Lb interface uses the IP transmission mode, each SMLC can be configured with a
maximum of 16 M3UA links.
The LCS service is initiated and controlled by the SMLC, such as the number of LCS
services, the number of M3UA links need to be configured, and the positioning method.
The peer SMLC provides the bandwidth required by the Lb interface.
Calculation formula for reference only:
Lb interface bandwidth (kbit/s) = Average signaling traffic per LCS service (byte) x 8 x
Number of LCS services per second/1024
2015-11-13
Huawei Confidential
Page 198 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
However, because the link bandwidth is small, and the number of initiated LCS services
varies, design the bandwidth with a redundancy capacity of about 50%.
The digit 8 in the preceding formula indicates that one byte consists of eight bits.
Average signaling traffic per LCS service is about 120 bytes.
19.7.7 Parameter Design

A BSC can be connected to an SMLC in the following ways:
By adding an Lb interface board

IP bearer can be used over the Lb interface.
Without adding an Lb interface board

If the Lb interface uses the IP bearer, the Lb interface must be configured on the A
interface board or Abis interface board of the BM subrack.
Table 1.1 lists the data to be planned and negotiated for the Lb interface (in IP bearer mode).
Table 1.1 Parameters to be planned for the Lb interface (in IP bearer mode)
Parameter
Example
Value
How to Obtain
OSP Code
Negotiated with the peer end
DSP Code
Network ID
NATB(NATB)
OSP code bits
BIT14
SS7 protocol type
ITUT(ITUT)
Local entity type
M3UA_IPSP(M
3UA_IPSP)
Destination Entity
Type
M3UA_IPSP(M
3UA_IPSP)
Traffic mode
M3UA_OVERR
IDE_MOD(Acti
ve/Standby
Mode)
Work mode
M3UA_IPSP(M
3UA_IPSP)
In multiple local signaling points scenarios, configure the Lb interface according to the following
operations. Specifically, establish the relationship between the signaling points and the SMLCs and
configure related link and route information for the signaling points and the SMLCs.
2015-11-13
Huawei Confidential
Page 199 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
For details about signaling link design of the Lb interface, see section 19.2.5"Signaling Configuration
Principles."
19.8 BTS Homing Allocation

BTS homing and TRX homing are designed according to the network planning design made
by network planning personnel. When performing network planning, network planning
personnel plan the allocation of BTSs and TRXs in various BSCs but do not complete the
module-level planning. To balance the processing capabilities of various BSC modules and
improve anti-impact and anti-risk capabilities, allocate BTSs among modules properly to
implement load balancing. That is, distribute BTSs in a continuous manner between BSCs but
distribute BTSs in a discontinuous manner within a BSC and between boards in large sites.
Design principles:
The number of TRXs needs to meet the designed specifications of GMPS and GEPS
subracks.
The number of TRXs needs to meet the required board processing specification.
The traffic carried by each module and BHCA do not exceed 60% of the designed
specification.
Certain redundant ports and capacity need to be reserved for each Abis interface board
for subsequent small-scale adjustment and expansion. The recommended redundancy is
20%, and the actual redundancy depends on the BOQ.
Plan the BTSs connected to the BSC continuously in the coverage area (unless
transmission conditions do not permit). Avoid discontinuous BTS distribution in different
BSCs; otherwise inter-MSC handovers increase.
Allocate the BTSs in the same LAC to the same subrack to reduce inter-module
signaling traffic.
Allocate the VIP sites (hot-spot areas with heavy traffic) in an area to different Abis
interface boards in a subrack in a discontinuous manner. Overlapping coverage exists
between adjacent cells. Therefore, this allocation mode can minimize the impacts due to
out-of-service of partial VIP sites in the same area.
In a multi-chain site (ring networking), distribute multiple chains to different boards to

prevent the entire site from being out of service due to board faults. Whether to adopt
this mode depends on the actual situation.
For an office that is constructed by phase, there may be many site re-homing
requirements. Therefore, during initial site allocation, allocate the sites that have such a
re-homing requirement to several Abis interface boards in a module and adopt
centralized cabling on the DDF to reduce the workload during re-homing.

HLD output
BSC Name
Module No
BSC1
BTS quantity
TRX quantity
1
2
3
2015-11-13
Huawei Confidential
Page 200 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
LLD output (work out LLD based on the LLD template)

BTS name BTS configuration
Module
No
Board No
Port No
S2/2/2
2015-11-13
Huawei Confidential
Page 201 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
20
Clock Synchronization
Design

Select a proper clock source according to the situation of the customer.
Design a networking diagram and a proper clock synchronization route according to the
clock source position, NE position, and transmission environment.
Design a clock connection diagram, instructing project implementation.

Type of the clock source
20.2 Clock Description

20.2.1 Definition of Synchronization
Synchronization indicates that two or more signals keep a specific relationship in

frequency or phase. That is, the phase difference or frequency difference of two or more
signals at the same moment keeps within a tolerable range.
Clock synchronization generally refers to frequency synchronization. The frequency of a

signal is on the reference frequency. Initial moment does not require consistency.
Time synchronization is also called moment synchronization, indicating an absolute time

synchronization. This means that the initial moment of a signal is consistent with the
Universal Time Coordinated (UTC).
20.2.2 SyncE
2015-11-13
The SyncE technology is defined in the ITU-T G.8262 protocol. This technology inherits
the basic clock synchronization theory of the SDH and PDH networks. The downstream
NEs obtain and trace the clock of the upstream NEs by restoring the clock from the serial
Huawei Confidential
Page 202 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
data streams received on the physical layer. The clock is extracted and restored from the
Ethernet physical layer, irrelevant to the specific service of the upstream NEs.
SyncE is compliant with the constraints and requirements of SyncE specified in the
G.8261, G.8262, and G.8264 protocols.
SyncE is available for commercial usage in GBSS15.0 or later. Intermediate transmission

devices are required to support the SyncE.
20.2.3 IEEE 1588 V2
The initial edition of the IEEE 1588 (Precision Time Protocol (PTP) used in
measurement and control systems) was developed by John Edison from the Agilent
Laboratories and 12 persons from other companies and organizations. It was approved
by the IEEE in November 2002.
The IEEE 1588 defines the PTP protocol for the standard Ethernet. The accuracy reaches
microsecond level. In 2008, the second edition of PTP focused on improving the
accuracy of frequency synchronization and minimizing forward delay between the
intermediate devices.
The IEEE 1588 is intended to synchronize the independent clocks running in

measurement and control systems. This protocol applies to the IP RAN and can
implement high-precision frequency synchronization even time synchronization between
the clock server (for example, IP Clock 3000) and the NodeB.
The second edition of the IEEE 1558 is compliant with the G.8265.1 and IEEE 1588 V2.
The G.8265.1 is released by the ITU for the synchronization of the layer-3 unicast
frequency in IEEE 1588. It supports interworking with servers of other vendors. The
IEEE 1588 V2, but does not support interworking with servers of other vendors.
Huawei's clock over IP proprietary protocol is not described here. Do not use it.
20.2.4 Advantages and Disadvantages of Clock

Protocols
Clock Source
The following clock sources are available:
Building Integrated Timing Supply System (BITS) clock
Clock obtained from the A interface
Local free-run clock
Clock over IP (Huawei proprietary protocol)
SyncE
IEEE 1588 V2
Comparison of synchronization technologies

Technolo
gy
Frequen
cy
Synchro
nization
Time
Synchro
nization
Advantage
Disadvantage
Remarks
Clock over
IP (Huawei
proprietary
Supports
transparent
transmission
Not support time

synchronization.
The IP Clock 1000

serves as a clock server
and the NodeB/BTS
2015-11-13
Huawei Confidential
Page 203 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Technolo
gy
Frequen
cy
Synchro
nization
Time
Synchro
nization
protocol)
Advantage
Disadvantage
across the bearer

network and has
low requirements
on the
intermediate
devices.
Not
recommende
d
serves as a client. Clock

information can be
transmitted on the IP
bearer network,
reflecting high
adaptability. Besides
network QoS, the Clock
over IP does not have
other special
requirements on the IP
bearer network and
therefore network
reconstruction is not
required.
3900 series
base stations
later than
GBSS9.0 do
not support
this function.
This technology
is mature and
application in
market is in a
long time.
The
BTS3012
still supports
this function.
Remarks
The clock recovery

quality is
vulnerable to delay,
jitter, and packet
loss on the bearer
network.
This protocol is a
Huawei proprietary
protocol and does not
support time
synchronization. In the
same condition, use the
IEEE 1588 V2.
Occupies Iub
downlink
bandwidth
resource. Normal,
30 kbit/s; max: 50
kbit/s.
SyncE
2015-11-13
This clock is
obtained from
the physical
layer, irrelevant
to upper layer
services.
Connectivity is
good.
Not support time

synchronization.
The Ethernet physical

media conversion
devices or packet
switched (PS) devices
located between the
Ethernet clock source
and the NodeB/BTS
(namely, the client)
must support SyncE.
Otherwise, interruption
occurs and the clock
cannot be allocated to
the lower layer NEs.
This technology
is mature and
clock recovery
quality is good
and is not
vulnerable to
packet loss and
In addition to the
RNC and NodeB,
the RAN network
also requires
intermediate
devices, such as the
hub and LAN
The NodeB/BTS and

other intermediate NEs
(for example,
TGW1000) support
SyncE. The
NodeB/BTS supports
not only hub cascade on
Huawei Confidential
Page 204 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Technolo
gy
Frequen
cy
Synchro
nization
Time
Synchro
nization
Advantage
Disadvantage
Remarks
jitter.
switch to support
clock transparent
transmission on the
physical layer or
clock regeneration.
the Iub interface but

also allocation of the
SyncE downstream
clock. However, if other
NEs exist between the
NodeB/BTS and the
downstream clock, the
intermediate NEs must
support SyncE.
Not occupy radio

bandwidth.
IEEE 1588
V2
2015-11-13
SyncE is not supported

when the transmission
rate is set to 10 Mbit/s.
If frequency
synchronization
is used,
transparent
transmission
across the bearer
network is
supported and
the requirement
on the
intermediate
devices is low.
If time
synchronization is
used, all
intermediate
devices must be
upgraded to
support IEEE 1588.
The clock server (for

example, Huawei IP
Clock 1000) that
supports IEEE 1588 V2
serves as a clock source
and the NodeB/BTS
serves as a client. If
frequency
synchronization is used,
either the clock server
or the NodeB/BTS
needs to support IEEE
1588 V2. Besides
network QoS, the IEEE
1588 V2 does not have
other special
requirements on the IP
bearer network and
therefore network
reconstruction is not
required.
Supports
frequency
synchronization
and time
synchronization
and meets the
requirements of
the LTE TDD on
clock.
The clock recovery

quality is
vulnerable to delay,
jitter, and packet
loss on the bearer
network.
If time synchronization
needs to be used, in
addition to the clock
server and NodeB/BTS,
all intermediate devices
(including microwave
devices, routers, and L2
switches) must support
IEEE 1588 V2.
IEEE 1588 V2 is
a standard
protocol,
supporting
interconnection
Occupies Iub
downlink
bandwidth
resource. Normal,
20 kbit/s; max: 40
Time synchronization is
not planned for the
NodeB/BTS.
Huawei Confidential
Page 205 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Technolo
gy
Frequen
cy
Synchro
nization
Time
Synchro
nization
Advantage
Disadvantage
of devices of
manufacturers.
kbit/s.
Remarks
20.2.5 QoS Requirements of Clock Protocols

Table 1.1 Requirements on clock accuracy
Clock
Synchronizati
on Over IP
Index
Index
Value
Remarks
Clock over IP
(Huawei
proprietary
protocol)
Jitter
< 20 ms
Packet loss rate
< 1%
SyncE
Frequency
accuracy of the
input clock
(+ 4.6 ppm, 4.
6 ppm)
Similar to E1/T1, SyncE is

obtained from the physical
layer and therefore SyncE
does not have special
requirements on the QoS of
the data bearer network.
According to the G.8262
protocol, the frequency
accuracy of the input SyncE
clock must be between + 4.6
ppm and 4.6 ppm.
IEEE 1588 V2
Jitter
20 ms
Packet loss rate
1%
If the jitter is great, the

frequency deviation of the
BTS and the clock source is
great. When the frequency
deviation is greater than 0.05
ppm, the clock is unlocked.
1. When the packet loss rate
is great, clock packets of the
timestamp are lost. This
results in great frequency
deviation.
2. When the packet loss rate
is great, negotiated packets
and the great period packet
are lost. This results in
interruption of clock links.
2015-11-13
Huawei Confidential
Page 206 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Clock
Synchronizati
on Over IP
Index
Index
Value
Remarks
Delay
60 ms
No impact
20.3 Clock Source Selection

Clock source selection is subject to the requirements of operators.
Select a clock source as follows:
In A over IP, the BTS can obtain only the BITS clock because it cannot obtain the line
clock on the A interface.
In Abis over IP, the BTS adopts IEEE 1588 V2, SyncE, or IP Clock.
20.4 Clock Design in Abis over TDM Mode

A clear networking diagram of a clock source and the BSC needs to be drawn.
To use a BITS clock, connect the clock cable from the BITS to the clock interface on the
panel of the GGCU board.
Clock networking in this mode is simple. If an operator puts forward a special clock source,
you need to confirm the principle of the special clock source with the operator and make a
drawing specific to the special clock source.
Figure 1.1 shows how to obtain a clock over the A interface.
Figure 1.1 Clock networking instance 1
Figure 1.2 shows how to obtain a clock on the backbone network.
2015-11-13
Huawei Confidential
Page 207 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.2 Clock networking instance 2
20.5 Clock Design in Abis over IP Mode

Design Principles
Determine a clock in Abis over IP mode according to the BSC, BTS, and bearer
information.
In Abis over MSTP, the BTS obtains the line clock over MSTP to implement clock
synchronization.
In Abis over FE/GE, the BTS adopts IEEE 1588 V2 or SyncE to implement clock
synchronization.
In Abis over microwave, if the microwave device has clock information, the BTS obtains
a line clock through microwave. Otherwise, adopt IEEE 1588 V2 or SyncE.
In Abis over satellite, the BTS adopts GPS to implement clock synchronization.
Design Guidelines
Determine a clock in Abis over IP mode according to the BSC, BTS, and bearer
information. See section "Scheme of Interface Clock Synchronization."
In Abis over MSTP, the BTS obtains the line clock over MSTP to implement clock
synchronization.
In Abis over FE/GE, the BTS adopts Clock over IP to implement clock synchronization.
In Abis over microwave, if the microwave device has clock information, the BTS obtains
a line clock through microwave. Otherwise, the BTS adopts Clock over IP.
In Abis over satellite, the BTS adopts GPS to implement clock synchronization.
Scheme of Interface Clock Synchronization

All radio data services require frequency accuracy. The BTS guarantees stable RF by means
of clock synchronization. Currently, the GSM requires 0.05 ppm frequency accuracy.
After the GSM network is constructed into an IP network, the BTS cannot obtain a clock
through a physical link. This is because the IP network is an asynchronous network.
Therefore, the clock needs to be obtained in a new mode for the BTS to ensure clock
synchronization on the air interface. Furthermore, the BSC does not lock the clock and
therefore the BSC does not require clock synchronization after IP construction. Table 1.1 lists
the schemes recommended for clock synchronization under GSM IP construction.
2015-11-13
Huawei Confidential
Page 208 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Table 1.1 Schemes recommended for clock synchronization under GSM IP construction
Networking
Clock
Synchronization
Scheme
Recommended for
the BTS
Clock
Synchronization
Scheme
Recommended
for the BSC
Abis over
MSTP/PTN
The BTS obtains a line

clock over MSTP/PTN to
implement clock
synchronization.
The BSC does not

require clock
synchronization.
Abis over FE/GE
The BTS adopts IEEE

1588 V2 or SyncE to
implement clock
synchronization.
Abis over
microwave
If the microwave device

has clock information, the
BTS obtains a line clock
through microwave.
Otherwise, the BTS adopts
IEEE 1588 V2 or SyncE.
Abis over satellite
The BTS adopts GPS to

implement clock
synchronization.
GSM BSS A
interface IP
construction
Abis over TDM
The BTS traces the BSC

clock.
The BSC adopts the

BITS clock to provide a
line clock for
downstream BTSs.
GSM BSS Gb
interface IP
construction
A over TDM
The BTS traces the BSC

clock.
The BSC adopts a line

clock and locks the
clock that serves the
SGSN.
GSM BSS all

IP
construction
(IP over FE)
The following section uses the MSTP-based IP networking as an example.

Figure 1.2 shows the MSTP-based Abis IP solution. In this solution, the BSC and BTS
connect to the MSTP device over the FE interface. The MSTP device encapsulates Ethernet
frames into the VC trunk, whose bandwidth is shared by multiple BTSs.
This solution is applicable to a GSM network that an SDH network or MSTP network
operator is constructing. The operator can upgrade the SDH network into an MSTP network,
therefore providing Ethernet access.
If the BTS and MSTP reside in the same site, the BTS connects to the MSTP over an
electrical FE interface and obtains the clock from the MSTP. If the BTS and MSTP reside in
different sites, the BTS connects to the MSTP over an optical FE interface and obtains the
external clock of the MSTP through an E1 link free of services.
2015-11-13
Huawei Confidential
Page 209 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 1.2 MSTP-based GSM IP solution
Currently, the transmission bearer network supports MSTP, PSTN, or L2/L3 networking. In
Abis over MSTP, the BTS obtains the line clock over MSTP to implement clock
synchronization. For details, see Figure 1.2. In Abis over L2/L3 networking, the BTS adopts
Clock over IP (supporting the IEEE 1588 V2) to implement clock synchronization of the BTS.
The following section describes Clock over IP and the design principle.
20.6 Design of the IP Clock Server

Introduction to Clock over IP (Supporting the IEEE 1588 V2)
The Clock over IP (supporting the IEEE 155 V2) is a solution for BTS clock synchronization.
The Clock over IP can be classified into IP Clock Server and IP Clock Client. The IP Clock
Server obtains the reference clock source from other devices, such as the GPS or BITS, and
delivers the clock information to the BTS (namely, the IP Clock Client) by using a clock
packet on the IP network. The BTS performs adaptive processing of the IP clock packet to
obtain the clock information.
Each IP clock packet occupies a certain bandwidth. On a rent network or in satellite
transmission, continuously sending IP clock packets may increase transmission cost and even
affect services in busy hours of the network. In the GBSS system, operators can customize the
time of sending IP clock packets to ensure that IP clock packets are sent during light network
load. The BTS of the GBSS system can retain 0.05 ppm frequency accuracy within 90 days
and periodically send IP clock packets, not only reducing bandwidth use but also preventing
BTS out-of-sync. Figure 1.1 shows the networking of Clock over IP.
2015-11-13
Huawei Confidential
Page 210 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 IP Clock synchronization networking
Table 1.1 describes the support for 2G-based 1588v2 clocks.

Table 1.1 Support for 2G-based 1588v2 clocks
Syste
m
IP
Clock
Type
Support
Clock
Redundanc
y
Backup
or Not
Address
1
Address 2
Synchroniz
ation Mode
2G
1588v2
Layer-3
unicast
Yes
IP address
of the IP
clock
server
IP address of
the IP clock
server
Intermittent
synchronization
(command: SET
BTSIPCLKPA
RA)
1588v2
Layer-2
multicast
Yes
MAC
address of
the IP
clock
server
MAC address
of the IP clock
server
Default
continuous
synchronization
2G networks support 1588v2 Layer-3 unicast and 1588v2 Layer-2 multicast. The clock mode widely
used on the live network is 1588v2 Layer-3 unicast.

You can run the SET BTSCLK command on the maintenance console to set Clock Type to
IP_TIME(IP Clock).
You can run the SET BTSIPCLKPARA command on the maintenance console to set Clock
Protocol Type to HW_DEFINED(Huawei User-defined) or PTP(PTP Protocol). The BTS,

however, supports only the PTP clock mode currently.
Design principle of the IP Clock Server (Huawei's IP Clock Server product is

IPCLK3000)
2015-11-13
Huawei Confidential
Page 211 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
For details about the IPCLK3000 product and the configuration guide, obtain the IPCLK3000
Description and IPCLK3000 User Guide at http://support.huawei.com and
http://3ms.huawei.com.
IPCLK3000 supports the following clock sources:
BITS clock
External 8 kHz clock provided by external devices
Global Positioning System (GPS)/Global Navigation Satellite System (GLONASS)

satellite clock
External 1 pulse per second (1PPS) clock
A built-in satellite card can be installed in IPCLK3000 to obtain GPS clock signals. GPS clock signals
generated by external satellite cards can also be obtained through clock signal interfaces on the panel.
IPCLK3000 clock performance indicators
Table 1.2 lists IPCLK3000 clock performance indicators.

Table 1.2 IPCLK3000 clock performance indicators
Name
Value
Maximum number of
supported clients
Frequency synchronization: 512 NodeBs,

eNodeBs, or BTSs
Time synchronization: 512 WiMAX BTSs
Maximum picket sending

frequency
IEEE 1588 V2: 128 packet per second (pps)
Maximum bandwidth
occupied by each signal
IEEE 1588 V2:

Frequency synchronization: normal: 12 kbit/s;
max: 190 kbit/s
Time synchronization: normal: 14 kbit/s; max:
210 kbit/s
Frequency retention hour

after clock sources are lost
7 days
Frequency retention
precision after clock
sources are lost
(+ 0.016 ppm, 0.016 ppm)
Network topology
Layer-3 networking, layer-2 networking of the private network, and the Internet public
networking are supported. Use the layer-3 networking.
The IP clock server accesses the network through a layer-3 router. If VLAN tags need to be
configured, add VLAN tags to intermediate transmission devices.
The IP clock server accesses the network through a layer-2 switch. If VLAN tags need to be
configured, add VLAN tags to intermediate transmission devices.
2015-11-13
Huawei Confidential
Page 212 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The IP clock server supports VLAN configurations. A maximum of 512 VLANs can be
configured on the IP clock server. On the BTS side, the number of base stations needs to be
less than 50 for a VLAN.
Capacity planning
In frequency synchronization, one IPCLK3000 can support 512 NodeBs, eNodeBs, or
BTSs.
IPCLK3000 is an independent case-shaped device. The actual number of IPCLK3000s to
be deployed depends on the number of Clients and the backup relationship between
IPCLK3000s. Generally, 500 BTSs share one IPCLK3000.
IP address planning
IP addresses of IP clock servers need to be configured on the radio BTS side. You can
plan IP addresses based on the IP address schemes for the layer-3 and layer-2 networking
modes. Generally, two IP addresses are planned.
Run the SET ETHIP command on the IP clock LMT to configure service IP
addresses of IPCLK3000 for matching service ports with Port Type set to
SERVICE.
Run the SET ETHIP command on the IP clock LMT to configure operation and
maintenance IP addresses of IPCLK3000 for matching operation and maintenance
ports with Port Type set to DEBUG.
Reliability
To implement 1+1 backup of reference clock sources, IP addresses of two IPCLK3000
clock servers need to be configured on the BTS side. With enhanced reliability, clock
sources are still available when an IPCLK3000 is faulty.
The two IPCLK3000 clock servers are independent. They are configured to be the primary and
secondary clock server on the BTS side.
Synchronization mode on the BTS

The support for continuous and intermittent synchronizations is introduced to the MBTS
in SRAN5.0. Intermittent synchronization only applies to the 1588v2 Layer-3 unicast
packet mode. Use the intermittent synchronization on the live network.
Intermittent synchronization is used in the following scenarios:
The transmission bandwidth is limited. (Intermittent synchronization helps save the

bandwidth.)
If the network QoS is poor in peak hours and good in off-peak hours, the
synchronization time needs to be set during peak hours and the clock needs to be
locked during off-peak hours.
Matching Version of the IP Clock Server

Table 1.3 describes whether the IP clock server adaptive synchronization protocols are
supported on the hardware platform or the old hardware platform.
Table 1.3 Whether the IP clock server adaptive synchronization protocols are supported on the
hardware platform or the old hardware platform
IP Clock Server IP Clock
(Customized by
Huawei)
V100R001
2015-11-13
1588 V2 (L3
Transparent
Transmission)
Hardware
Platform
Old hardware
Huawei Confidential
Page 213 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
IP Clock Server IP Clock

(Customized by
Huawei)
1588 V2 (L3
Transparent
Transmission)
Hardware
Platform
V100R002
platform
New hardware
platform
V200R002
V100R001 and V100R002 apply only on the old hardware platform and they cannot be upgraded to
V200R002.
V100R002 is a non-productive version and is an upgrade version of V100R001.
V200R002 applies on the new hardware platform. All devices to be delivered have used this version
since August 1, 2009.
Table 1.4 describes whether the GSM products support the IP adaptive synchronization
protocols.
Table 1.4 Whether the GSM products support the IP adaptive synchronization protocols
IP Clock
(Customized by
Huawei)
GBSS15.0
1588 V2 (L3
Transparent
Transmission)
1588 V2 over
MAC
Because the hardware logic resource of the GSM BTS3900 is limited, the GBSS15.0 supports only 1588
V2.
IP clock version under various GSM scenarios

Create an IP clock server.
If an IP clock server is required, V200R002 is delivered.
2015-11-13
Huawei Confidential
Page 214 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
21
Time Synchronization
Design

Select a proper time synchronization source.
Design a networking scheme of time synchronization according to transmission

information and NE position.
Design IP addresses for time synchronization.

Type of time synchronization, and position and IP address of the M2000 server
21.2 Description of Time Synchronization

Time synchronization indicates that the time of communication devices or computer devices
on the communication network is UTC-based and the time offset is small enough, for
example, 100 ms.
A time synchronization network works in client/server mode and adopts a leveled time server
to implement time synchronization. Normally, a time synchronization source is obtained from
the standard time source.
On a BSS network, time synchronization means that the NMS synchronizes with BSS NEs. In
this way, the NMS can record the time of generating alarms and events of each NE,
facilitating fault analysis and performance analysis.
21.3 NTP
Network Time Protocol (NTP) is a complex time synchronization protocol across WAN and
LAN, specific to microsecond. The NTP can be used in two modes: broadcast and
2015-11-13
Huawei Confidential
Page 215 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
client/server. The later mode has higher accuracy than the former one. In client/server mode,
the NTP server needs to exchange NTP packets with NEs requiring tine synchronization, to
obtain the time offset between the NTP server and the NEs. The client/server mode has 1 to
10 ms accuracy. Therefore, this mode is widely used in network time transmission.
NTP uses UDP transmission and adopts the standard port number 123.
Simple Network Time Protocol (SNTP) is a simplified NTP protocol.
21.4 Selection of a Time Synchronization

Source
Currently, the following time synchronization sources are used:
GPS time synchronization source.
Internet-based time synchronization source. The Internet provides many NTP-based time
servers. You can access the servers to implement time synchronization.
Selection of a time synchronization source is subject to the time source provided by an

operator.
21.5 Transmission Mode

Transmission mode is classified into wireless transmission and wired transmission (DCN and
DDN). DCN is a TCP/IP-based network for internal transmission. It features convenient
networking, little investment, high security, and high reliability. Therefore, DCN is used in
mobile networks for time synchronization. DDN, a private transmission network, requires
high cost.
21.6 Typical Networking

Figure 1.1 shows a typical networking for BSS time synchronization.
Figure 1.1 Typical networking for time synchronization
2015-11-13
Huawei Confidential
Page 216 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
21.7 Typical Application

The NTP server of the OMU of the BSC is configured with the IP address of the M2000. The
NTP time is obtained from the M2000.
One BSC can be configured with addresses of multiple NTP servers, with the port number
being 123.
The following table lists the NTP servers and the port number.
BSC Name
NTP Server
Port No
BRBSC10
10.123.0.4
123
BRBSC10
10.123.0.5
123
BRBSC10
10.123.0.6
123
The configuration of daylight saving time (DST) varies with areas. The following tables
describe the DST configuration of a north European country.
2015-11-13
ZONET
DST
SM
SMONTH
GMT+1
YES
Week
March
SDAY
SWSEQ
SWEEK
ST
Last
Sunday
02:00:00
EM
EMONTH
Week
October
EDAY
EWSEQ
Last
EWEEK
ET
TO
Sunday
03:00:00
60
Huawei Confidential
Page 217 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
22
Function Design
22.1 Design of Broadcast Solutions for Cells

22.1.1 Standard Broadcast
The cell broadcast is a broadcast short message (SM) service specific to the GSM system.
Within a same period, the cell broadcast uses radio CBCH channels to send messages to a
specific coverage area in a single direction under certain conditions. In this way, all MSs on a
network can receive messages and do not need to respond to the received messages.
The simple cell broadcast provides the simple cell broadcast service without the CBC system.
The broadcast information includes the cell name, weather information, and social
commonweal information.
Table 1.1 NEs involved in the cell broadcast system
Acronyms and
Abbreviations
Full Name
BTS
Base transceiver station
BSC
Base station controller
CBC
Cell broadcast center
CBS
Cell broadcast server
CBT
Cell broadcast terminal
CBE
Cell broadcast entity
The cell broadcast system comprises the CBE, CBC, BSC, BTS, and MS. Each NE provides
the following functions:
2015-11-13
The CBE is an interface connecting external message sources and the GSM network. It
records cell broadcast information and encodes and formats the cell broadcast
information.
The CBC collects and stores formatted cell broadcast information from the CBE and
then sends the broadcast information to specific BSCs based on scheduling information
Huawei Confidential
Page 218 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
in the broadcast information. In addition, the CBC manages broadcast information on the
BSC, for example, deleting outdated information and querying the status of broadcasting
information in a cell.
The BSC schedules and maintains broadcasting information. It also needs to maintain the
CBCH status.
The BTS distributes cell broadcast information over the Um interface.
The MS receives and displays cell broadcast information.
To enable the cell broadcast, a CBCH channel needs to be configured on a cell. The CBCH
channel is a logical channel and occupies the same physical channel with the SDCCH.
Therefore, the CBCH channel can be configured in the following two ways:
SDCCH/8+CBCH: used for a cell whose channel type is set to the non-combined BCCH
channel
BCCH+CCCH+SDCCH/4+CBCH: used for a cell whose channel type is set to the

combined BCCH channel
For the SDCCH/8+CBCH configuration mode, select SDCCH+CBCH for Channel Type.
For the BCCH+CCCH+SDCCH/4+CBCH configuration mode, select BCCH+CBCH for
Channel Type.
For the SDCCH+CBCH channel configuration, set CCCH Blocks Reserved for AGCH to 0.
This is because the MS needs to temporarily stop monitoring the PCH channel and to receive
contents over the CBCH channel so that the MS can receive cell broadcast information. In this
case, if a paging message for the MS is reported over the PCH channel, the MS cannot receive
this paging message. When CCCH Blocks Reserved for AGCH is not set to 0, the MS can
receive information over the CBCH channel within the AGCH channel period and can receive
all paging messages. For the BCCH+CBCH channel configuration, CCCH Blocks Reserved
for AGCH needs not to be set to a value other than 0.
The SDCCH+CBCH channel can be configured only on any of timeslots 0 to 3 of the carrier.
The BCCH+CBCH channel can be configured only on time slot 0 of the carrier.
Due to codec constraints, if the CBCH channel is configured on the time slot involved in
frequency hopping, the number of frequencies of frequency hopping configured on the time
slot needs to be less than 32.
Network design: The network design of the cell broadcast is simple. The CBC connects to the
BSC over the CB interface. The XPU board of the BSC connects to the CBC over the
Ethernet network interface. The network design needs to consider the location of the CBC and
BSC and transmission resources. Figure 1.2 shows the network design diagram.
Figure 1.2 Network topology of the cell broadcast
2015-11-13
Huawei Confidential
Page 219 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The BSC directly interwork with the CBC over the IP network. Specifically, the BSC accesses
the CBC network using a network cable from the port on the IP interface board (depending on
the port enabled in Configure boards attributes on the IP interface board. Figure 1.3 shows
the physical cable connection diagram.
Figure 1.3 Cable connection diagram between the interface board and the CBC
Configure cell broadcast data based on the product feature configuration guide without additional data
configuration for the IP interface board.
Table 3.1 Key parameter configurations
2015-11-13
IP address of the
CBC
Negotiated with peer equipment
Subnet mask
MAC address
Huawei Confidential
Page 220 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Table 3.2 Parameter configurations
2015-11-13
Name
Description
Command
Impact NE
Channel
Type
This parameter specifies the

channel type of the timeslot on
the TRX. The channel type of
timeslot 0 must not be set,
because the combined BCCH is
configured by default. The
channel type of other timeslots
can be set to full-rate TCH or
half-rate TCH.
SET
BTSCHNFALLBACK
(Mandatory)
BTS
CCCH
Blocks
Reserved for
AGCH
BS-AG-BLKS-RES, indicating
the number of the CCCH
message blocks reserved for the
AGCH. After the CCCHs are
configured, the value of this
parameter indicates the actual
seizure rates of the AGCHs and
the PCHs over the CCCHs.
SET
GCELLIDLEBASIC
(Optional)
Cell
SMCBC
DRX
This parameter specifies

whether to support the
discontinuous reception
mechanism (DRX). To reduce
the power consumption, the
DRX is introduced into the
GSM Specification. MSs
supporting the DRX can
consume less power to receive
interested broadcast messages.
This prolongs the service time
of MS batteries. BSCs
supporting the DRX must send
scheduling messages to MSs so
that the MSs can use the DRX
function. The period occupied
by broadcast messages that are
contained in a scheduling
message is called a scheduling
period. In the sending sequence,
a scheduling message contains
the description of each short
message to be broadcast and the
position of each broadcast
message in the scheduling
period.
SET GCELLOTHEXT
(Optional)
Cell
Support Cell
Broadcast

whether the BSC6910 supports
the cell broadcast function.
ADD
GCNOPERATOR
(Optional)
BSC6910
Huawei Confidential
Page 221 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
BSC IP
This parameter specifies the IP

address for the communication
between the BSC and the CBC.
It must be consistent with the
configured IP address of the
CPU board.
ADD GCBSADDR
(Mandatory)
BSC6910
BSC Port
This parameter specifies the port

number used for the
communication between the
BSC6910 and the CBC.
ADD GCBSADDR
(Mandatory)
BSC6910
BSC
GateWay IP
When the BSC supports the

standard cell broadcast function,
it sends cell broadcast messages
using the IP address specified by
this parameter.
ADD GCBSADDR
(Mandatory)
BSC6910
CBC ITF
Para
The value of this parameter

must be translated into binary
digits. The following describes
each bit of the value:
ADD GCBSADDR
BSC6910
Bit 0: phase flag
Bit 1: message type flag
Bit 2: cell-list flag
Bit 3: whether to carry a

recovery indication
Bit 4: whether to carry the cell

flag
Bit 5: whether to carry a

recovery indication during
the reset procedure
Support Cell
Broadcast
Name

whether to broadcast the cell
name.
SET GCELLSBC
(Optional)
Cell
CBC IP
This parameter specifies the IP

address of a CBC.
ADD GCBSADDR
BSC6910
This parameter specifies the port

number at the CBC side in the
communication with the
BSC6910.
ADD GCBSADDR

contents of a cell broadcast
message.
SET GCELLSBC
CBC Port
Broadcast
Content
2015-11-13
(Optional)
Huawei Confidential
(Mandatory)
BSC6910
(Mandatory)
Cell
(Optional)
Page 222 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Geography
Scope
Chan ID
Coding
Scheme
Broadcast
Interval

geographic scope of a simple
cell broadcast message. You can
obtain the value of this
parameter using DSP
GSMSCB. Geography Scope,
Code, and Update uniquely
identify a cell broadcast
message.
SET GCELLSBC
This is a key parameter for

adding a simple cell broadcast
message. It specifies the channel
ID of a simple cell broadcast
message. You can obtain the
value of this parameter using
DSP GSMSCB.
SET GCELLSBC
This is a key parameter for

identifying a simple cell
broadcast message. It specifies
the coding scheme of message
contents.
SET GCELLSBC
(Optional)

time interval for a cell broadcast
message.
SET GCELLSBC
(Optional)
Cell
(Optional)
ADD GSMSCB
(Mandatory)
Cell
(Optional)
ADD GSMSCB
(Mandatory)
Cell
ADD GSMSCB
(Mandatory)
Cell
22.1.2 Simple Cell Broadcast

The simple cell broadcast provides the simple cell broadcast service without the CBC system.
The broadcast information includes the cell name, weather information, and social
commonweal information. To enable the simple cell broadcast service, set Support Cell
Broadcast to Support Simple CB.
The simple cell broadcast provides two types of functions: cell name broadcast and cell
broadcast.
The cell name broadcast enables the BSC to send a one-page-long text message to the MSs in
a cell under the BSC. The text message contains the cell name. Therefore, when a roaming
MS enters a new cell where the simple cell broadcast function is enabled, the MS can obtain
and display the cell name. This function equates to the simple location service. The BSC will
keep sending the one-page-long message, but MS users can disable this function on the MS to
stop the receiving of this message. To enable the cell name broadcast function, first set
Support Cell Broadcast Name to Yes and then specify parameters, such as Broadcast
Content (this parameter is set to the cell name by default), Geography Scope, Chan ID,
Coding Scheme, and Broadcast Interval.
The ADD GSMSCB command can be run to send a cell broadcast message to the BSC. If
you send the cell broadcast message in such a way, specify ST and ET of the period during
which the cell broadcast message is broadcast. The maximum size of the cell broadcast
message that can be sent through the MML command is 15 pages. A cell can save a maximum
of 63 pages of cell broadcast messages.
2015-11-13
Huawei Confidential
Page 223 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
In addition, the DSP GSMSCB command can be used to query the cell broadcast messages
saved in a cell; you can also run the RMV GSMSCB command to stop the broadcast of a cell
broadcast message. Figure 1.1 shows the topology of the simple cell broadcast system.
Figure 1.1 Topology of the simple cell broadcast system
The simple cell broadcast cannot be used with the cell broadcast provided by the CBC. They are
mutually exclusive.
The simple cell broadcast provides only simple cell broadcast services. The standard cell broadcast
system is recommended for the dynamic information that is frequently changed.
22.2 Design of Radio Measurement Data

Interface for Navigation (TOM-TOM)
22.2.1 Overview
The radio measurement data interface for navigation (TOM-TOM) is used for real-time
navigation on highways.
The TOM-TOM sends MS-related radio data collected by the BSC to the VNP, and the VNP
reports the data to the third-party collection unit (CU). (The MS-related radio data includes
whether the MS is within the BTS coverage area of the highway, the movement speed of the
MS, and the movement direction of the MS.) The CU processes the reported, collected
information, selects a proper path, and sends the path information to the GPS end user for
providing guidance for the user to bypass the congested road.
The BSC and VNP used during the entire navigation are provided by the device manufacturer
and the CU is from a third party.
You can configure an ENTSWITCH on the BSC. The TOM-TOM cannot be used if the
switch is disabled. You can also configure the IP address of the VNP on the BSC.
For details about the function of the TOM-TOM, see the BSC6910 GSM Product
Documentation.
2015-11-13
Huawei Confidential
Page 224 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
22.2.2 Reference Document

Deployment Guide:
The deployment guide of this feature is available on the deployment guide named BSC6910
GSM Product Documentation.
22.2.3 Limitations on Specifications

The TOM-TOM has the following limitations on specifications:
A BSC can connect to only one VNP.
A VNP can connect to a maximum of five BSCs.
Neither ARP nor BFD is supported.
Neither IP paths nor logical ports can be configured.
The IP PM function is not supported.
Priorities of DSCP values and VLANs cannot be configured.
VLANs cannot be configured based on the service flow. To differentiate the OM service
flow from the TOM-TOM service flow, use different next hops for the OM and the
TOM-TOM. Then configure VLANs based on the next hops.
22.2.4 Software and Hardware Configuration

No new hardware configuration is added to the BSC.
The VNP interface uses the EOMU board of the BSC6910.
This feature is supported in BSC6910 V900R015 and later. Before using the feature, you need
to apply for a license.

22.2.5.1 Logical Networking
Figure 1.1 shows the logical networking for the TOM-TOM.
2015-11-13
Huawei Confidential
Page 225 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Logical networking for the TOM-TOM
The logical networking shows only the logical connection, but not the actual physical connection
between network elements.
The BSC interworks with the VNP using the IP bearer. Subscriber events generated on the
BSC are reported to the VNP over the TCP/IP (the BSC provides TCP 6200 port). Then, the
VNP sends the data to a third-party server for calculation. A BSC can connect and report
events to only one VNP. A VNP, however, can connect to multiple BSCs simultaneously.
The M2000, CU, and Traffic server obtain the synchronization time from the NTP server, and
the BSC and VNP obtain the synchronization time from the M2000.
22.2.5.2 VNP Interface Networking Design

The VNP interface uses the EOMU board of the BSC6910. Figure 1.1 shows the physical
networking over the VNP interface.
The EOMUs are configured in active/standby mode. The Ethernet interfaces 0 and 1 of each
EOMU are connected to different LAN switch ports. Two VRRP routers are configured.
Figure 1.1 Physical networking on the VNP interface
2015-11-13
Huawei Confidential
Page 226 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The use of the TOM-TOM feature does not cause a change to the original OMU interface
networking generally, because the interface of the OMU adopts the IP transmission
networking. A new logical IP address of the OMU, however, is required as a communication
address of the VPN interface (see VPN logical IP address in Figure 1.1). In addition, the IP
address of the peer VPN needs to be configured as the destination IP address.
The local IP address of the VNP must be different from the virtual IP address of the external
network of the OMU or the fixed IP address of the external network. If the local IP address of
the VNP is a logical IP address, it can be configured as a 32-bit mask, and in the same
network segment or different network segments with the IP address of the OM.
Figure 1.2 shows the networking of the active/standby OMUs with a single port and directly
connected routers. This networking is used only by some operators, because the networking
has lower reliability, although it saves two LAN switches as compared with the networking in
Figure 1.1.
Figure 1.2 Networking of the active/standby OMUs with a single port and directly connected
routers
22.2.5.3 VLAN Planning

You can configure different VLANs to distinguish the OM service flow from the TOM-TOM
service flow. (This configuration is optional.)
22.2.5.4 QoS Planning

Priorities of DSCP values and VLANs cannot be configured.
22.2.6 Bandwidth Design

None.
2015-11-13
Huawei Confidential
Page 227 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
22.2.7 Time Synchronization

The M2000, CU, and Traffic server obtain the synchronization time from the NTP server, and
the BSC and VNP obtain the synchronization time from the M2000. Figure 1.1 shows the
networking for time synchronization.
Figure 1.1 Networking for time synchronization
22.3 MOCN II Design

22.3.1 Overview
The principles, specifications, networking, parameter settings, and feature activation have
been described in the MOCN II Feature Parameter Description. This section only describes
the planning and design related to this feature.

For details, see section "Network Topology" in MOCN II Feature Parameter Description.
MOCN II achieves RAN equipment sharing, including BSC, BTS, Abis transmission
resources. The core network resources cannot be shared.
The A and Gb interfaces can share interface boards and transmission on the BSC but use
different logical resources, such as IP addresses and routes.
2015-11-13
Huawei Confidential
Page 228 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
22.3.3 Capacity Planning

The MOCN II networking does not affect capacity planning of RAN resources. RAN
resources are planned in the same way as in traditional networking mode. During the
planning, obtain the total amount of resources required by different operators.
MOCN II does not allow operator-based configuration of Abis transmission resources. Abis
transmission resources are shared and therefore the method for planning the resources is the
same as the traditional one.
The A and Gb interfaces must be connected to CNs of different operators. Therefore, the
bandwidths over the A and Gb interfaces are planned according to the need of each operator.
22.3.4 Interface Design

The interface networking does not change during the implementation of MOCN II. Therefore,
the traditional networking modes over each interface still apply in MOCN II-enabled
scenarios. For details, see chapter 19 "Transmission Interface Design."
Since the A and Gb interfaces are connected to the CNs of different operators in MOCNenabled scenarios, the interfaces use the following configurations:
Operators use different interface boards, but the interface networking modes are the same
as the traditional ones.
Operators share interface boards. The physical networking in this situation is the same as
the traditional one, but multiple device IP addresses are used to distinguish operators.
Transmission resource pool is a network networking mode. The following table lists the
designed specifications in the transmission resource pool mode.
The MOCN networking can be classified into operator-based independent configuration and
configuration sharing among multiple operators for A interface boards.
Mode
Operator-based Independent
Configuration of A interface
Boards (Recommended)
Configuration Sharing of A
Interface Boards Among
Multiple Operators
Pooled with operators

distinguished
Pooled according to the number of

operators or pooled with the boards
belonging to the same operator.
Pooled with all boards. Configure certain

number of device IP addresses for all
boards and form multiple pools where
physical ports can be shared or
independently used and port IP addresses
can be shared or independently used (in
this situation, configure multiple port IP
addresses). Then, transfer services to
CNs of different operators according to
the destination IP addresses.
2015-11-13
Huawei Confidential
Page 229 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Pooled without
distinguishing operators
N/A
Boards in active/standby mode, two

ports in load sharing mode, or an
independent board can form a pool.
Services are transferred to CNs of
different operators according to the
destination IP addresses.
22.4 Design of BSC Node Redundancy

22.4.1 Overview
The BSC Node Redundancy feature allows two BSCs to form a redundancy group. Two BSCs
in a redundancy group work in 1+1 backup mode, and a BTS is connected to both BSCs.
Under normal circumstances, the BTSs controlled by each BSC operate properly. If one BSC
fails or all the signaling links on the A interface of one BSC are disconnected, the other BSC
takes over services from the failed BSC. The following figure shows the networking diagram
of two BSCs working in a redundancy group.
2015-11-13
Huawei Confidential
Page 230 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Unlike cold standby, the BSC Node Redundancy feature recovers services without adjusting
the transmission data over interfaces or reconfiguring data. However, if a BSC switchover is
triggered, all ongoing services will be interrupted because no backup data is available, but
new services will not be affected. In this sense, this feature is neither hot standby nor warm
standby, but a redundancy between the warm standby and cold standby.
The principles, specifications, networking, parameter configurations, and feature activation
have been described in BSC Node Redundancy Feature Parameter Description. This section
only describes the planning and design related to this feature.
22.4.2 Constraints
The design of BSC node redundancy is subject to the following constraints:
2015-11-13
Huawei Confidential
Page 231 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
The settings of the parameters that do not support configuration synchronization must be
consistent between two BSCs. These parameters include equipment parameters, BSC
radio parameter, and transmission-related parameters (excluding IP address-related
parameters).
To enable successful configuration synchronization, the parameters that do not support

synchronization must be configured first.
For example, the radio parameter settings for a dual-homed BTS can be synchronized
only after Abis transmission data has been configured on the secondary-homed BSC.
If the settings of parameters that do not support configuration synchronization on the

secondary-homed BSC conflict with those on the primary-homed BSC, the
synchronization of parameter settings may fail. If the configuration synchronization fails,
a report is displayed on the CME, reminding you to take the following measures to
rectify the situation:
Manually modify the parameters that do not support configuration synchronization on
the secondary-homed BSC so that the parameter settings are consistent with those on the
primary-homed BSC. Then, manually trigger a configuration synchronization task.
Example 1:
Problem description: A BTS under BSC 1 uses IP transmission, and the BSC interface
board connecting BSC 1 to the BTS supports IP transmission, but the interface board in
the slot with the same slot No. on BSC 2 is not an IP interface board. As a result, a
configuration synchronization task fails.
Rectification measures:
Install an IP interface board in the slot with the same No. on BSC 2.
On BSC 2 LMT, run the ADD BRD command to add data configurations.
Then, execute a configuration synchronization task.
Example 2:
Problem description: For BSC 1, Support RAN Sharing is set to NO(No) in the SET
BSCBASIC command, and Sharing Allow is set to NO(NO) in the SET
BTSSHARING command. However, for BSC 2, Support RAN Sharing is set to
YES(Yes) in the SET BSCBASIC command. As a result, a configuration
synchronization task fails.
Rectification measures:
On BSC 2 LMT, run the SET BSCBASIC command with Support RAN Sharing set to
NO(No). Then, perform a configuration synchronization task.
The cascaded BTSs working in IP over E1 mode must have the same homing attributes
under one BSC.
Two BSCs support the configuration synchronization function only when they run the
same software version (VxxxRxxxCxx). To enable two BSCs running different software
versions to support this function, upgrade them to the same software version.
When you reconstruct single-homed BTSs to dual-homed BTSs under an existing BSC,
perform the following steps:
2015-11-13
Configure Abis-interface transmission data for multiple BTSs ( 50).
Synchronize in batches the BTS-level radio parameter settings on the primary-homed

BSC to the secondary-homed BSC.
When you first perform data synchronization for the BSC Node Redundancy feature
during networking configuration, perform an immediate synchronization task and then
perform periodic synchronization tasks.
Huawei Confidential
Page 232 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
If you configure data for primary-homed BTSs and then use the configuration
synchronization function to generate the configuration data for secondary-homed BTSs,
ALM-21829 BSC Node Redundancy Configuration Exception is reported because a
periodic synchronization task is executed on the CME on a daily basis. It is
recommended that you configure data for primary-homed BTSs and then use immediate
synchronization to generate the configuration data for secondary-homed BTSs.
If the configuration synchronization parameters include BSCIP, configuration

synchronization cannot be implemented by running the ADD BTSEXTOPIP command
in the Abis Independent Transmission feature. In this situation, manually configure
parameters on the peer BSC.
The ADD PTPBVC command used to configure the Gb interface includes the NSEI
parameter. Therefore, configuration synchronization cannot be implemented by running
this command. In this situation, manually configure parameters on the peer BSC.
The MML commands listed in the following table support configuration synchronization,
but the listed parameters do not support synchronization. To change the values of these
parameters, manually configure the parameters on the peer BSC.
MML Command
Parameter ID
ADD BTS
BTSTYPE
SET BTSALM
BTSTYPE
SET BTSAUTOPLANCFG
BTSTYPE
ADD BTS
SEPERATEMODE
ADD BTS
RFUCFGBYSLOT
ADD BTS
SRANMODE
ADD BTSCABINET
SRANMODE
All boards on the backup BSC must be normal. The XPU/SPU that accommodates the
main control AICP module of the backup BSC must be normal. Otherwise, node
redundancy is unavailable because links cannot be synchronized.
In node redundancy scenarios, the main control AICP module is used for managing heartbeat links of the
primary and secondary BSCs.
To query the number of the XPU/SPU where the main control AICP module resides, run
the DSP FAMDATA command with Data Table set to CCENTRALTRAFFICCPU.
In the command output, TrafficSubrackNo, TrafficSlotNo, and TrafficCpuNo
corresponding to TrafficType equal to 3 indicate the XPU/SPU's CPU where the main
control AICP module resides. As shown in the following figure, the main control AICP
module is located on CPU 2 of the board in slot 0 of subrack 0.
2015-11-13
Huawei Confidential
Page 233 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Fault detection is implemented between two BSCs in a redundancy group by checking

the heartbeat messages periodically transmitted from the peer end over the inter-BSC
interface. The BEATSENDINGDISparameter specifies the interval for sending heartbeat
messages between two BSCs.Heartbeat messages are transmitted over SCTP links on the
inter-BSC interface. To check the status of an SCTP link, run the DSP SCTPLNK
command.

The BSC Node Redundancy feature is used in two typical scenarios: load sharing mode and
active/standby mode.
The two scenarios have been described in the BSC Node Redundancy Feature Parameter
Description.
For details, see section "Network Topologies" in the BSC Node Redundancy Feature
Parameter Description in GBSS17.0.
22.4.4 Capacity Planning

Capacity planning can be performed in terms of control plane, user plane, and transmission.
The control-plane capacity planning includes BHCA, number of cells, number of activated
subscribers, and number of subscribers who are processing services.
The user-plane capacity planning includes CS traffic volume (Erlang), PS throughput, number
of cells, and number of activated subscribers.
The transmission-plane capacity planning includes CS traffic volume (Erlang), number of
BTSs, number of subscribers who are processing services, and number of call connections.
The capacity planning for the BSCs in a redundancy group is the same as that for an
independent BSC. Each BSC in a redundancy group is configured with a specified number of
boards that can meet the total specifications of two BSCs.
As shown in the following figure, one BSC reserves certain capacity for the other BSC in a
redundancy group. The number of purchased boards for each BSC in a redundancy group can
meet the total specifications of two BSCs. For example, if the BSCa supports 3000 TRXs and
the BSCb supports 1000 TRXs, the XPU of each BSC supports 4000 TRXs. This rule also
applies to other boards.
2015-11-13
Huawei Confidential
Page 234 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
As shown in the following figure, the BSCs support dual-homed BTSs. Therefore, some BTSs
serving VIP subscribers can be configured as dual-homed BTSs and required service
processing boards are purchased for the BSCs. The remaining BTSs are configured as singlehomed BTSs. After a BSC is faulty, services of the BTSs serving VIP subscribers can recover,
but services of the remaining BTSs are interrupted. For example, the BSCa supports 3000
TRXs of which 1000 TRXs are configured for dual-homed BTSs; the BSCb supports 1000
TRXs of which 500 TRXs are configured for dual-homed BTSs. In this situation, the XPU of
the BSCa can support 3500 TRXs and that of the BSCb can support 2000 TRXs. This rule
also applies to other boards.
If the active BSC has been configured with the SAU, NIU, NASP, or GCG, the standby BSC
must also be configured with the same board to ensure that related functions and features can
recover after services on the standby BSCs are recovered.
The inter-BSC detection link interface used in the BSC Node Redundancy feature is a
Huawei-customized interface which carries necessary equipment information for inter-BSC
interaction, such as heartbeat messages. When the inter-BSC SCTP detection link is
configured, the inter-BSC interface can only use IP transmission. Therefore, an IP interface
board is required or the existing IP interface board is used.
The GOUc or GOUe on the BSC6900 can carry the inter-BSC detection link interface.
The traffic on the inter-BSC detection link can be neglected. Therefore, capacity planning is
not required for the traffic.
22.4.5 Interface Design

The following table describes the requirements for transmission modes used by different
interfaces during the implementation of BSC Node Redundancy.
2015-11-13
Huawei Confidential
Page 235 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
A Interface
C
o
n
t
r
o
l
P
l
a
n
e
A Interface
User Plane
Abis Interface
Support BSC Node

Redundancy
IP over ETH
IP over ETH
IP over ETH
Manual
switchover/automatic
switchover
IP over ETH
IP over ETH
(E1/T1 transmission
terminated at the router)
Manual
switchover/automatic
switchover
IP over ETH
IP over ETH
(the BTS directly

connected to the BSC)
Manual switchover
TDM
TDM
TDM
Not support
For details about transmission modes used by different interfaces, see section "Network
Topologies."
2015-11-13
Huawei Confidential
Page 236 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
According to the GSM 03.71, Figure 1.1 shows the logical structure of the LCS system on the
GSM network.
Figure 1.1 Logical structure of the LCS system on the GSM network
LMU
Type A
CBC
BTS
(LMU
Type B) Abis
Abis
SMLC
SMLC
HLR
CBC-BSC
Um
MS
Lp
CBC-SMLC
Ls
Lb
Lh
A
BSC
MSC/VLR
Lg
Le
GMLC
Gb
Gs
External
LCS client
Lc
Lg
SGSN
LMU
Type B
gsmSCF
GMLC
Other PLMN
Table 1.1 NEs involved in the LCS service
2015-11-13
NE
Function Description
LCS Client
The LCS client is a logical function entity that requests location

information of one or multiple MSs from the LCS server. The
location request message contains the QoS parameter. The LCS
client can reside in entities on the PLMN (including MSs) or
entities out of the PLMN.
MS
The LCS server can provide location information for an MS. For
the network-based LCS, a destination MS does not need to support
the LCS. For an MS-assistant or MS-based LCS, an MS needs to
support the LCS. For all LCSs, the MS privacy can be controlled
through registration in each location request. On the LCS client, a
destination MS can be identified using MSISDN. On the PLMN
network, a destination MS can be identified using MSISDN, IMSI,
or an internal flag of the PLMN network. In emergency calls, a
destination MS can be identified using MSISDN, IMSI, or NAESRK+IMEI.
Huawei Confidential
Page 237 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
NE
SMLC
The serving mobile location center (SMLC) coordinates and

schedules resources required for the LCS and calculates location
estimation results and precision.
Two SMLCs are available: NSS-based SMLC and BSS-based
SMLC. The NSS-based SMLC interworks with one or multiple
MSC servers over the Ls interface to support the LCS and manage
the LMU. The BSS-based SMLC interworks with one or multiple
BSCs over the Lb interface to support the LCS and manage the
LMU. Both NSS-based SMLC and BSS-based SMLC can obtain
resources and information of other SMLCs over the Lp interface.
The SMLC and the gateway mobile location center (GMLC) can be
integrated into one physical node or can reside in different physical
nodes. When the CBC connects to the BSC, the SMLC needs to
interwork with the CBC to perform assistance data broadcast using
the cell broadcast function of an existing cell.
GMLC
One PLMN network can have multiple GMLCs. The GMLC is the
first node through which an external LCS client accesses the GSM
network. After the GMLC receives an LCS request from an LCS
subscriber, it queries route information of a destination MS from the
HLR over the Lh interface. After the GLMC authenticates an LCS
subscriber, it sends the LCS request to the VMSC over the Lg
interface. After the LCS flow ends, the GMLC obtains the location
estimation result from the VMSC.
LMU
The LMU is a logical network entity. Its LCS measurement function

can support one or multiple LCS methods. The LMU measurement
is classified into the following two measurements:
LCS measurement for an MS: is used to calculate the location

estimation result of an MS.
Assistance measurement for all MSs in a specific geographic

area: is used to perform periodic measurement over radio
interfaces, such as Absolute Time Differences (ATD) and Real
Time Differences (RTD).
Each LMU is controlled and managed by an SMLC on the network.

Measurement parameters and relevant commands of the LMU can
be provided by this SMLC or preset in the LMU. All measurement
results of the LMU are reported to the SMLC through an LCS
request.
The LMU is classified into the A-type and B-type LMUs:
2015-11-13
The A-type LMU is identified with the IMSI. It adopts the same
frequency with an MS and accesses the BTS over the Um
interface. It does not connect to any NE. It has an independent
subscription profile in the HLR and supports the mobility
management function of all radio resources and interfaces. The
HLR differentiates the A-type LMU and an MS based on
settings in the subscription profile.
The B-type LMU accesses the BSC over the Abis interface. It can
be deployed independently or be integrated into the BTS.
Huawei Confidential
Page 238 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
NE
MSC/VLR
The MSC/VLR registers and authenticates an MS and manages

LCS requests relevant or irrelevant to GSM calls. The MSC server
accesses the GMLC over the Lg interface and the SMLC over the
Ls interface. If the MSC server connects to the SGSN over the Gs
interface, the MSC server checks whether an MS is in the GPRS
attach status to determine whether it pages the MS over the A
interface or the Gs interface.
SGSN
The SGSN transfers paging requests in the CS domain received

over the Gs interface to the BSS.
BSC
The BSC connects to the SMLC over the Lb interface. It provides

system operation capability and LCS assistance function in the LCS
flow.
HLR
The HLR stores LCS subscription data and route information of an

MS. It connects to the GMLC over the Lh interface. For a roaming
MS, the HLR serving the MS and the SMLC may reside in different
PLMN networks.
CBC
The CBC connects to or is embedded the broadcast entity of the

BSC. It broadcasts LCS assistance information specified by the
SMLC to cells managed by the BSC using the signaling between
the CBC and the SMLC.
gsmSCF
The gsmSCF connects to the GMLC over the Lc interface and can
visit the LCS using the CAMEL III.
Huawei BSC supports the LCS service only in the CELL+TA mode. It supports NSS-based
SMLC and BSS-based SMLC and does not support the LMU. 22.5.1.1.1Step 1Figure 1.2
shows the logical structure of the NSS-based SMLC. 22.5.1.1.1Step 1Figure 1.3 shows the
logical structure of the BSS-based SMLC. In this scenario, Huawei BSC and the SMLC are
integrated.
Figure 1.2 Logical structure of the NSS-based SMLC
2015-11-13
Huawei Confidential
Page 239 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.3 Logical structure of the BSS-based SMLC
Figure 1.4 shows the LCS flow initiated by an external LCS client.
Figure 1.4 LCS flow initiated by an external LCS client
For the A interface in the LCS design, configure the LCS function data on the BSC and
longitude and latitude information of each cell.
2015-11-13
Huawei Confidential
Page 240 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
22.5 LCS Function Design

The location service (LCS) is a special service specific to the GSM network. The precision of
the LCS service implemented on the GSM network is low, but can meet the requirements of
those services requiring low precision. The LCS service provides MSs with the weather
report, tour arrangement, emergency aid, and traffic conditions based on the location of MSs.
According to the GSM 03.71, Figure 1.1 shows the logical structure of the LCS system on the
GSM network.
Figure 1.1 Logical structure of the LCS system on the GSM network
CBC
BTS
(LMU
Type B) Abis
Abis
SMLC
SMLC
HLR
CBC-BSC
Um
MS
Lp
CBC-SMLC
LMU
Type A
Ls
Lb
Lh
A
BSC
MSC/VLR
Lg
Le
GMLC
Gb
Gs
External
LCS client
Lc
Lg
SGSN
LMU
Type B
gsmSCF
GMLC
Other PLMN
Table 1.1 NEs involved in the LCS service
2015-11-13
NE
LCS Client
The LCS client is a logical function entity that requests location

information of one or multiple MSs from the LCS server. The location
request message contains the QoS parameter. The LCS client can reside in
entities on the PLMN (including MSs) or entities out of the PLMN.
MS
The LCS server can provide location information for an MS. For the
network-based LCS, a destination MS does not need to support the LCS.
For an MS-assistant or MS-based LCS, an MS needs to support the LCS.
For all LCSs, the MS privacy can be controlled through registration in each
location request. On the LCS client, a destination MS can be identified
using MSISDN. On the PLMN network, a destination MS can be identified
using MSISDN, IMSI, or an internal flag of the PLMN network. In
emergency calls, a destination MS can be identified using MSISDN, IMSI,
or NA-ESRK+IMEI.
Huawei Confidential
Page 241 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
SMLC
The serving mobile location center (SMLC) coordinates and schedules

resources required for the LCS and calculates location estimation results
and precision.
Two SMLCs are available: NSS-based SMLC and BSS-based SMLC. The
NSS-based SMLC interworks with one or multiple MSC servers over the
Ls interface to support the LCS and manage the LMU. The BSS-based
SMLC interworks with one or multiple BSCs over the Lb interface to
support the LCS and manage the LMU. Both NSS-based SMLC and BSSbased SMLC can obtain resources and information of other SMLCs over
the Lp interface.
The SMLC and the gateway mobile location center (GMLC) can be
integrated into one physical node or can reside in different physical nodes.
When the CBC connects to the BSC, the SMLC needs to interwork with
the CBC to perform assistance data broadcast using the cell broadcast
function of an existing cell.
GMLC
One PLMN network can have multiple GMLCs. The GMLC is the first
node through which an external LCS client accesses the GSM network.
After the GMLC receives an LCS request from an LCS subscriber, it
queries route information of a destination MS from the HLR over the Lh
interface. After the GLMC authenticates an LCS subscriber, it sends the
LCS request to the VMSC over the Lg interface. After the LCS flow ends,
the GMLC obtains the location estimation result from the VMSC.
LMU
The LMU is a logical network entity. Its LCS measurement function can
support one or multiple LCS methods. The LMU measurement is classified
into the following two measurements:
LCS measurement for an MS: is used to calculate the location estimation

result of an MS.
Assistance measurement for all MSs in a specific geographic area: is

used to perform periodic measurement over radio interfaces, such as
Absolute Time Differences (ATD) and Real Time Differences (RTD).
Each LMU is controlled and managed by an SMLC on the network.

Measurement parameters and relevant commands of the LMU can be
provided by this SMLC or preset in the LMU. All measurement results of
the LMU are reported to the SMLC through an LCS request.
The LMU is classified into the A-type and B-type LMUs:
MSC/VLR
2015-11-13
The A-type LMU is identified with the IMSI. It adopts the same
frequency with an MS and accesses the BTS over the Um interface. It
does not connect to any NE. It has an independent subscription profile
in the HLR and supports the mobility management function of all radio
resources and interfaces. The HLR differentiates the A-type LMU and
an MS based on settings in the subscription profile.
The B-type LMU accesses the BSC over the Abis interface. It can be
deployed independently or be integrated into the BTS.
The MSC/VLR registers and authenticates an MS and manages LCS

requests relevant or irrelevant to GSM calls. The MSC server accesses the
GMLC over the Lg interface and the SMLC over the Ls interface. If the
MSC server connects to the SGSN over the Gs interface, the MSC server
checks whether an MS is in the GPRS attach status to determine whether it
pages the MS over the A interface or the Gs interface.
Huawei Confidential
Page 242 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
SGSN
The SGSN transfers paging requests in the CS domain received over the Gs
interface to the BSS.
BSC
The BSC connects to the SMLC over the Lb interface. It provides system
operation capability and LCS assistance function in the LCS flow.
HLR
The HLR stores LCS subscription data and route information of an MS. It
connects to the GMLC over the Lh interface. For a roaming MS, the HLR
serving the MS and the SMLC may reside in different PLMN networks.
CBC
The CBC connects to or is embedded the broadcast entity of the BSC. It

broadcasts LCS assistance information specified by the SMLC to cells
managed by the BSC using the signaling between the CBC and the SMLC.
gsmSCF
The gsmSCF connects to the GMLC over the Lc interface and can visit the
LCS using the CAMEL III.
Huawei BSC supports the LCS service only in the CELL+TA mode. It supports NSS-based
SMLC and BSS-based SMLC and does not support the LMU. Figure 1.2 shows the logical
structure of the NSS-based SMLC. Figure 1.3 shows the logical structure of the BSS-based
SMLC. In this scenario, Huawei BSC and the SMLC are integrated.
Figure 1.2 Logical structure of the NSS-based SMLC
Figure 1.3 Logical structure of the BSS-based SMLC
Figure 1.4 shows the LCS flow initiated by an external LCS client.
2015-11-13
Huawei Confidential
Page 243 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.4 LCS flow initiated by an external LCS client
For the A interface in the LCS design, configure the LCS function data on the BSC and
longitude and latitude information of each cell.
2015-11-13
Huawei Confidential
Page 244 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
23
BTS Design
23.1 BTS Cable Design

Design the internal cable diagram of the BTS in different BSC types and configurations to
provide guidance for engineers in construction and improve work efficiency.

Network plan data of a cell (BTS type, carrier quantity, SXXX, and frequency)
23.2 Design Tool of the BTS Cable Diagram

Currently, tools are available for the cable connection design for major BTS types. Onsite
TSD personnel can obtain the tool at http://support.huawei.com.
23.3 BTS Transmission Design

For the Huawei's third and fourth generation of BTSs with the most installed base in the
market, design BTS transmission networking methods under different transmission conditions
to provide guidance for engineers in construction and improve work efficiency.
23.3.2 BTS Transmission

Huawei's third and fourth generation of BTSs support the TDM and IP over FE/GE. The
carrier quantity supported by different transmission protocols varies. The TDM supports a
maximum of 126 carriers and the IP over FE supports a maximum of 60 carriers.
For details about networking, see section 19.5"Abis Interface Design."
2015-11-13
Huawei Confidential
Page 245 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Constraints on usage of different transmission media (E1 and FE): Currently, transmission
interface boards of GSM BTSs are DTMU and GTMU. The DTMU does not support hybrid
usage of E1 and FE interfaces.
BTS Type
Transmission
Board
Supported
Transmission Mode
Supported
Transmission
Interface
Hybrid Usage
of
Transmission
Interface
Supported?
BTS3012
DTMU, DPTU
TDM, IP over FE
E1/T1, FE
No
BTS3012AE
DTMU, DPTU
TDM, IP over FE
E1/T1, FE
No
BTS3006C
DMCM
TDM
E1/T1, STM-1
N/A
BTS3900
GTMU, UIEB,
UTRPC
TDM, IP over FE/GE
E1/T1, FE/GE
Hybrid usage of
TDM and IP is not
supported.
BTS3900E
MICRO
TDM, IP over FE
E1/T1, FE
No
BTS3900L
GTMU, UIEB,
UTRPC
TDM, IP over FE/GE
E1/T1, FE/GE
Hybrid usage of
TDM and IP is not
supported.
BTS3900B
PICO
IP over FE
FE
N/A
BTS3900A
GTMU, UIEB,
UTRPC
TDM, IP over FE/GE
E1/T1, FE/GE
Hybrid usage of
TDM and IP is not
supported.
BTS3900AL
GTMU, UIEB,
UTRPC
TDM, IP over FE/GE
E1/T1, FE/GE
Hybrid usage of
TDM and IP is not
supported.
2015-11-13
Huawei Confidential
Page 246 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
DBS3900
GTMU, UIEB,
UTRPC
TDM, IP over FE/GE
E1/T1, FE/GE
Hybrid usage of
TDM and IP is not
supported.
The BTS communicates with the BSC either in the port IP communication mode or in the
logical IP communication mode. The logical IP communication in GBSS14.0 is the same as
that in GBSS9.0. If the BTS adopts the port IP communication mode, only configurations of
port 0 are supported in GBSS9.0 and configurations of the FE optical interface are added to
support the GU transmission backup scenario in GBSS12.0 and later versions. The following
table describes comparison of two versions.
Comparison of GBSS9.0 and GBSS12.0 and later versions when the BTS adopts the port
IP communication mode
Configuration of the Port IP

Communication Mode
GBSS9.0
GBSS12.0 and Later

Version
FE0 configured
FE1 not
configured
Support
Support
FE0 configured
FE1 configured
Not support
Support
FE0 not
configured
FE1 configured
Not support
Support
If BTSIP configured on the BSC side is the same as the device IP address of any port configured on the
BTS, the BTS adopts the port IP communication mode. If BTSIP differs from the device IP configured
on the BTS, the BTS adopts the logical IP communication mode.
When the BTS adopts the port IP communication mode and only the electrical interface is used, the IP
address needs to be configured on port 0. That is, BTSIP needs to be the same as the device IP address of
Ethernet port 0.
When the BTS adopts the port IP communication mode and only the optical interface is used, the IP
address needs to be configured on port 1. That is, BTSIP needs to be the same as the device IP address of
Ethernet port 1.
In an upgrade of an installed site, VLAN tags can be added based on the service type. For new sites,
except for the GTMUa board, add VLAN tags based on the IP address of the next hop.
The BSC communicates with the BTS either in the port IP communication mode or in the
logical IP communication mode. The following table shows advantages and disadvantages of
these two communication modes. For details about configuration methods, see the
deployment guide of the BSC6910.
2015-11-13
Advantages and disadvantages of the port IP communication mode and the logical IP
communication mode
Huawei Confidential
Page 247 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Advantage/Disadva
ntage
Port IP
Communication
Mode
Advantage
The configuration is
simple.
Port addresses of
intermediate
transmission networks
are visible.
Disadvantage
When physical links are

faulty, cooperation of
links is difficult to be
implemented to ensure
that services are not
affected.
Port IP addresses are
used as service IP
addresses, which
requires more IP
addresses.
Logical IP
Communication
Mode
Logical IP addresses are

used as service IP
addresses, which requires
less IP addresses.
When physical links are

faulty, cooperation of
links is easy to be
implemented to ensure
that services are not
affected.
Static routes to logical IP

addresses need to be
configured.
When intermediate layer-3

transmission devices are
available, routes of this
logical IP address need to
be identified, which has
high requirements on
transmission networks.
23.3.3 eGBTS Networking

The eGBTS is introduced in the GBSS15.0. Compared with the transmission GBTS, 3900
series base stations support the eGBTS and matching BTS types are DBS3900 eGSM,
BTS3900 eGSM, BTS3900A eGSM, BTS3900L eGSM, and BTS3900AL eGSM.
The eGBTS has the following characteristics:
2015-11-13
The BTS provides the southbound interface and OM channels. The M2000/CME can
directly manage the eGBTS.
The BTS LMT maintains and manages local and remote eGBTSs.
The original Abis interface is adjusted to the Abis 2.0 interface. The original OML
changes to CSL and the message flow also changes.
The layer-2 LAPD over the original Abis interface changes to the SCTP, which is the
same as that of the UMTS.
The BTS license file is added and the file is directly loaded on the BTS.
The BSC supports the hybrid networking of non-eGBTS and GBTS, but does not
support hybrid cascading.
Huawei Confidential
Page 248 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 1.1 Networking topology change of the eGBTS
Figure 1.2 Change of northbound and southbound interfaces of the eGBTS
2015-11-13
Huawei Confidential
Page 249 of 258
GBSS17.0 & BSC6910)
24
CONFIDENTIAL
OM Networking Design

Physical position of the MSC, BSC, and PCU and the topology
Number of BTSs
24.1.2 Design Content
Networking of the BSS and the NMS
Calculation of NMS bandwidth
24.1.3 Reference
M2000 Commissioning Guide
24.2 Introduction to OMU

24.2.1 Standalone OMU
The ETH5-SCU6 and ETH4-SCU7 are network cards for internal communication. The OMU
connects to the SCU using the cards to obtain the performance and alarm information of the
BSC board.
The ETH0 and ETH1 are network cards for external communication, connecting to the LMT
through the LAN switch or hub.
Download the BSC GOMU Management Guide at http://support.huawei.com.
2015-11-13
Huawei Confidential
Page 250 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Standalone OMU
24.2.2 Dual OMU

The ETH5-SCU6 and ETH4-SCU7 are network cards for internal communication. The OMU
connects to the SCU using the cards to obtain the performance and alarm information of the
BSC board.
The ETH3-UPDATE is a network card used to connect two OMUs working in active/standby
mode, to implement data synchronization and software update.
The ETH0 and ETH1 are network cards for external communication, connecting to the LMT
through the LAN switch or hub.
The active and standby OMUs are configured with the same IP address.
The active and standby OMUs connect four LAN cables to the hub or LAN switch.
Download the BSC GOMU Management Guide at http://support.huawei.com.
2015-11-13
Huawei Confidential
Page 251 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.1 Dual OMUs
24.3 OM Networking Design

If the network of a customer is IP-based, use the IP networking.
If the existing network is TDM-based, entire E1 networking is preferred.
If the existing network is TDM-based and the customer has high requirements on cost,
networking for part of E1 timeslots is recommended. That is, some of timeslots of the E1 link
are used as operation and maintenance channels.
24.3.1 Networking for Part of E1/T1 Timeslots

In this mode, the following devices are required: router, MSC, and Mercury3600.
The core of networking for part of E1 timeslots is as follows: Cross-connect the NMS
information to the idle timeslots of the existing E1 link by using the Mercury3600 for
transmitting the information to the peer end. The peer end cross-connects the NMS
information and provides the information to the NMS by using a router.
The timeslot extraction principle (specific to Mercury3600) is as follows:
2015-11-13
Huawei Confidential
Page 252 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
In the networking mode of part of E1 timeslots, timeslot cross-connect devices, such as

Mercury3600 are used. The four universal slots of a Mercury3600 can be inserted with 4-E1
interface cards and 2-V35 interface cards. A maximum of 16 E1 channels can be provided.
Cross connection of any timeslot can be performed on any two ports. Any port can use the
near loop or remote loop to implement performance monitor and system maintenance.
In the current OM networking, Mercury3600 uses a 4-E1 interface card (with 25-pin D
model).
Figure 1.1 shows the pins.
Figure 1.1 25-pin D model interface
For details, see the Mercury 3600 Manual.

Figure 1.2 shows a typical OM networking.
2015-11-13
Huawei Confidential
Page 253 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Figure 1.2 Networking for part of E1/T1 timeslots
24.3.2 Entire E1/T1 Networking

Compared with networking for part of E1 timeslots, entire E1 networking does not require
cross-connect devices for timeslot extraction and exchange. An entire 2 Mbit/s E1 link can be
used for information transmission, applicable to a network requiring abundant transmission
resources and large data volume.
In this mode, routers and MSCs are required.
Figure 1.1 shows a typical OM networking.
Figure 1.1 Entire E1/T1 Networking
2015-11-13
Huawei Confidential
Page 254 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
24.3.3 IP Networking
Usually, IP networking is used in the private network of an operator. With high reliability and
high transmission efficiency, this networking mode is preferred in the area where the operator
has constructed a private IP network. In an IP network, the NMS device only needs to provide
a network interface because the routes of the IP network and remote end are completed. The
bandwidth for IP networking is allocated by the operator as required. A private IP network
requires private transmission resource. Therefore, cost is high.
Figure 1.1 and Figure 1.2 show typical OM networking.
Figure 1.1 OM network topology
Figure 1.2 IP networking in dual OMU mode
In dual OMU mode, three IP addresses need to be planned. Each OMU is configured with a
physical IP address. Two OMUs share a logical IP address. The three IP addresses are in the
same network segment.
2015-11-13
Huawei Confidential
Page 255 of 258
GBSS17.0 & BSC6910)
CONFIDENTIAL
Preferentially, switch is performed on ports of the same board.
24.3.4 Networking Instances

Figure 1.1 and Figure 1.2 show an office adopting E1 networking.
Figure 1.1 OM E1 networking instance 1
To facilitate M2000 maintenance, three M2000s are placed in an equipment room. The BSCs
in other cities are remotely connected to this equipment room. E1 transmission is used.
Mercury3600s in this networking provide port conversion and timeslot adjustment functions.
2015-11-13
Huawei Confidential
Page 256 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
Figure 1.2 OM E1 networking instance 2
24.4 OM IP Address Planning

Negotiate with operators how to plan IP addresses.
When planning IP addresses, pay attention to the following:
Ensure the IP addresses of the LANs connected to a WAN in different network segments.
Ensure the LAN port and the WAN port of a router in an office in different network
segments.
Ensure the WAN ports of two routers connecting and communicating with each other in
different network segments.
Ensure the WAN ports of routers with different office directions in different network
segments.
Divide a large network segment into small subnets based on the subnet masks of IP
addresses and allocate the IP addresses of the subnets to LANs, to save IP network
segment.
Assess the number of IP addresses that can be allocated and the extension space in future
when allocating IP addresses of subnets.
24.5 Route Planning

The network between the network card of extranet of the GOMU and the LMT/M2000 is
defined as the extranet. The GOMU can connect to the LMT/M2000 directly or through
multiple routers (gateways).
2015-11-13
Huawei Confidential
Page 257 of 258
CONFIDENTIAL
GBSS17.0 & BSC6910)
When the OMU communicates with the M2000 through routers (gateways), the OMU must
connect to the master server and the slave server of the M2000. In this case, routs between the
OMU and M2000 need to be added. Set the destination IP address of the OMU route to the
network segment address of M2000 by running ADD OMUIPRT, rather than the IP address
of the master server and the slave server of the M2000. If multiple the M2000 has multiple
network segments, set the destination IP address to multiple network segment addresses.
Ensure that when a slave server of the M2000 is added, a route to the newly added slave
server of the M2000 is added on the OMU.
24.6 Impact of eGBTS on the O&M

The eGBTS is introduced in the GBSS15.0. Figure 1.1 shows the change of the OM structure.
Figure 1.1 Change of the OM structure
2015-11-13
The OM model and function of the non-eGBTS are deployed on the BSC side. BTSs are
managed by the BSC and access the OSS. BTSs do not have independent southbound
interfaces. In the eGBTS, relevant models and functions of physical devices and
transmissions of the BTS are adjusted from the BSC to the BTS. The BSC implements
only the logical model and service processing. The OSS manages the OM of physical
devices, transmission, and local logical objects of the BTS. The OM management
channel and the southbound interface are added between the OSS and the eGBTS and are
managed by the M2000 as the new NE type.
The BTS LMT is added for the eGBTS. Operation and maintenance can be performed on
the local eGBTS or by remotely connecting to the eGBTS through the communication
network.
The BSC6910 LMT is removed with the OM management function of public physical
devices and transmission of the eGBTS. This OM management function is implemented
by the M2000 and the BTS LMT currently.
For the eGBTS, a new NE type is added on the OSS northbound interface. The
northbound model of this NE type differs from that of the non-eGBTS. On a hybrid
network, a set of northbound interfaces are available and can be differentiated through
the NE type.
The local maintenance tool SMT of the non-eGBTS is removed in the eGBTS and its
function is migrated to the BTS LMT.
Huawei Confidential
Page 258 of 258

Technical Guide To SRAN Network Design (GO Applicable To SRAN10.0 & GBSS17.0 & BSC6910) - 20140902-A-1.0

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Technical Guide To SRAN Network Design (GO Applicable To SRAN10.0 & GBSS17.0 & BSC6910) - 20140902-A-1.0

Загружено:

Авторское право:

Доступные форматы

Technical Guide to SRAN Network Design (GO Applicable to the SRAN10.

Total 258 pages

Technical Guide to SRAN Network Design

fuqiang (employee ID:00283077)

lishuanghua (employee ID: 00101863)

Li Yongqing (employee ID: 00141602)

hepeng (employee ID: 00110002)

Huawei Technologies Co., Ltd.

Add a note about the

Deleted eGSM and

Add 16.4 source IP

2 Overview of Network Design........................................................21

8 BOQ Review Guide......................................................................34

9 Parameters for Capacity Calculation.............................................42

12 Naming Rules Design.................................................................58

13 BSC6910 Networking Principles..................................................63

14 Optical Interface MSP................................................................70

14.2 MSP Technical Description........................................................................................................................................74

17 Network Topology Design...........................................................91

19 Transmission Interface Design..................................................118

19.2.9 QoS Design (A over IP).........................................................................................................................................139

19.8 BTS Homing Allocation...........................................................................................................................................199

20 Clock Synchronization Design...................................................201

21 Time Synchronization Design...................................................214

22.3.4 Interface Design.....................................................................................................................................................228

23 BTS Design.............................................................................. 243

Figure 18-11 Dual EOMUs.................................................................................................................................114

Figure 19-27 BTS networking diagram..............................................................................................................178

Figure 24-7 IP networking in dual OMU mode..................................................................................................250

Table 17-7 Gb over IP interworking parameters.................................................................................................151

Technical Guide to SRAN Network Design (GO

Keywords: network design

The BSC6910 does not support Abis over HDLC.

The A interface does not support IP over E1/T1.

Not supports the EXOUa in 10

Supports the EXOUa in 10 GE

Supports the calculation of the CS

Supports the calculation of the

Supports the bandwidth

Supports bandwidth calculation

Supports the bandwidth

Supports bandwidth calculation

Supports the bandwidth

Not supports this interface.

Supports the bandwidth

Not supports this interface.

Supports the bandwidth

Not supports Gb over FR over

The value of both BTS ID and

The value of both BTS ID and

Supports the IP design on the

Supports the IP design on the

Supports the TDM networking,

Not supports the TDM

Supports the reliability design of

Supports the reliability design of

Supports A over TDM, A over IP

Not supports A over IP over E1.

Supports Abis over TDM over

Supports Abis over TDM over

Supports Gb over FR and Gb over

Supports Gb over IP. Not

Supports line clock, GPS, BITS