Академический Документы
Профессиональный Документы
Культура Документы
Passing on and copying of this document, use and communication of its contents not
permitted without written authorization from Alcatel-Lucent
SAFETY WARNING
Alcatel-Lucent training materials can be for products or refer to products that have both lethal and dangerous voltages present. Always
observe all safety precautions and do not work on the equipment alone. The user is strongly advised not to wear conductive jewelry
while working on the products. Equipment referred to or used during this course may be electrostatic sensitive. Please observe correct
anti-static precautions.
written consent. All characters appearing in this training course are fictitious. Any resemblance to real persons, living or dead, is purely
@@COURSENAME
coincidental.
There may be a number of proprietary logos, marks, trademarks, slogans and product designations found in the Content. Alcatel,
Lucent, Alcatel-Lucent and the Alcatel-Lucent logos are trademarks of Alcatel-Lucent. All other trademarks are the property of their
respective owners. Alcatel-Lucent does not grant you a license to use any of the foregoing logos, marks, trademarks, slogans and
product designations in any fashion. Granting of the right to access and use the Content for training purposes does not confer upon
you any license under any of Alcatel-Lucents or any third party's IP Rights.
DISCLAIMER
ALCATEL-LUCENT DISCLAIMS ALL WARRANTIES REGARDING THE TRAINING COURSES OR THE CONTENT, EXPRESS OR IMPLIED,
INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
THE ALCATEL-LUCENT WILL NOT BE RESPONSIBLE OR LIABLE FOR ANY INJURY, LOSS, CLAIM, DAMAGE, OR ANY SPECIAL,
EXEMPLARY, PUNITIVE, INDIRECT, INCIDENTAL OR CONSEQUENTIAL DAMAGES OF ANY KIND (INCLUDING WITHOUT LIMITATION
LOSS PROFITS OR LOSS SAVINGS), WHETHER BASED IN CONTRACT, TORT, STRICT LIABILITY OR OTHERWISE, THAT ARISES OUT
OF OR IS IN ANY WAY CONNECTED WITH (A) ANY USE OR MISUSE OF THE CONTENT OR THE TRAINING COURSES BY YOU, OR (B)
ANY FAILURE OR DELAY BY ALCATEL-LUCENT, ITS OFFICERS, DIRECTORS, AGENTS OR EMPLOYEES IN CONNECTION WITH THE
CONTENT OR THE TRAINING COURSES (INCLUDING, WITHOUT LIMITATION, THE USE OF OR INABILITY TO USE ANY COMPONENT
OF THE CONTENT OR TRAINING BY YOU). SOME JURISDICTIONS LIMIT OR PROHIBIT SUCH EXCLUSION OF WARRANTIES OR
LIMITATION OF LIABILITIES AND SO THE FOREGOING EXCLUSION OF WARRANTIES OR LIMITATION OF LIABILITY MAY NOT APPLY
TO YOU.
GOVERNING LAW
These Terms of Use and Legal Notice are governed by the laws of France. The operation and use of the training course is governed by
the laws of the country that governs your employment contract, if applicable. If any provision of these Terms of Use and Legal Notice,
or the application thereto to a person or circumstance, is held invalid or unenforceable by law, statute or a court of competent
jurisdiction, for any reason, then such provision shall be modified and/or superseded by a provision that reflects the intent of the
original provision as closely as possible. All other provisions of these Terms of Use and Legal Notice shall remain in full force and
effect. You may not assign these Terms of Use or any permission granted hereunder without Alcatel-Lucents prior written
consent. Nothing herein shall be deemed an employment agreement or an offer of employment or an alteration in any way of a users
terms of employment with or within Alcatel-Lucent.
Copyright 2011 Alcatel-Lucent. All Rights Reserved
Welcome to UTRAN
1. W-CDMA 4. Topic/Section is Positioned Here
R99 Algorithms Description
1. 9300
UA08 UTRAN Parameters
W-CDMA R99and Objects Description
Algorithms
5. Topic/Section is Positioned Here
2. UTRAN Configuration
3. Services
1. W-CDMA R99 Algorithms Description 6. Topic/Section is Positioned Here
1. UTRAN
4. Measurements Parameters and Objects
2. UTRAN Configuration 7. Topic/Section is Positioned Here
5. Call Admission
3. Services
6. Power Management
4. Measurements
7. Call Management
5. Call Admission
8. Mobility in Reselection
6. Power Management
9. Mobility in SHO
7. Call Management
10. 8.Inter-Carrier Mobility in HHO
Mobility in Reselection
11. 9.Inter-Carrier
Mobility in SHO Mobility at RRC Connection
12. 10.
Abbreviations and Acronyms
Inter-Carrier Mobility in HHO
11. Inter-Carrier Mobility at RRC Connection
12. Abbreviations and Acronyms
UTRAN
Upon completion of this course, you should be able to:
Course content
describe the call establishment and the associated parameters: RAB Matching, IRM RAB to RB Mapping, CAC,
1 UTRAN Parameters and Objects
CELL_FACH admission
1 UTRAN Parameters and Objects
describe
2 UTRAN the packet data management principles: Always On, Rb Rate Adaptation, iRM Scheduling, iRM
Configuration
Preemption
3 Services and associated parameters
4 Measurements
describe power management and control with the associated parameters
5 Call Admission
describe
6 Power Handover types and purpose: SHO, Alarm Handovers, iMCTA algorithm
Management
4
7 Call Management
@@PRODUCT
COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
7 Call Management
@@COURSENAME
training.feedback@alcatel-lucent.com
Thank you!
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
ATM
RF
ND IP
Provisioning
Activities WPS for Access
OA&M
Activities
Configuration Data
Main Server
This process describes how to configure UTRAN Network Elements (NEs) during a deployment phase. The main
steps are the following:
Planning Activities:
Check UTRAN CIQs consistency
Provide neighboring XML files for cell planning
Provide the last WPS templates and ATM Profile
Provisioning Activities:
Generate full configuration with WPS
Export XML files from WPS
Operation Administration & Maintenance Activities:
Load configuration data into their respective NEs
Build the database (MIB) of the RNS and make sure all the local pieces of information are up-to-date.
Perform real-time adjustments to the initial network configuration.
Note:
These steps do not necessarily apply to other contexts such as introduction of new features, addition of new NEs,
network optimization, etc.
A
T
RRM Parameters A M Para
T
RRM ParametersrsA M Para meters
e sTM Par meter
D si n
t e
k De sig
RRM Parameters
ram ete
ig n
r ame s
es g
a
or rk De
P P
n
m s ters
I Para eter
tw o rk
P m s
I Para ter
Ne tw o
N etw
I P ra m e e rs
a t
FP me
N
.
.. R Para
e
F
R
Operator Engineering
Teams Teams
CIQ (Core, Local)
The Customer Input Questionnaire is a repository where all parameter values and configuration data required for
the later datafill of the UTRAN subsystem are stored.
As mentioned in the document header: "The CIQ is used by the Wireless Network Engineering team, Regional
Engineering and deployment personnel to better understand the customer requirements.
Each manager of a Local Engineering team (in relation with the other activity groups) is in charge of filling his own
part of the CIQ along with the operator:
RadioFrequency (RF) staff fills RF parameters. The RF team can also provide XML files coming from any cell
planning tool such as iPlanner.
IP engineering staff fills the IP addresses.
ATM engineering staff fills the ATM parameters.
Etc.
The UTRAN CIQ template highlights for each parameter the domain it belongs to (Design, IP, ATM, etc.).
At WPS level, the UTRAN datafill engineer is in charge of checking the consistency and completeness of the UTRAN
CIQs.
Off-
Off-line On-
On-line
WPS Access
Main Server
XML Files
XML
Open BTS
Engineering Tools I-Node C-Node
Interface
RNC NodeB
Provisioning OA&M
For the UMTS Access Network, Wireless-Network Management System provides two complementary sets of
configuration tools:
An off-line configuration tool to support network engineering.
An on-line configuration tool to support network operations.
These two toolkits fully inter-work and provide a consistent user environment for the engineering and operations
staff.
Off-lineconfiguration is designed to support efficient bulk configuration of the UTRAN by the engineering staff.
Users can import, modify and export data, both from the UMTS access network and from 3rd party engineering
tools (such as iPlanner). Off-line Configuration delivers a seamless network-engineering environment from initial
network design through to actual network configuration.
On-lineconfiguration has been designed to change the configuration of the UTRAN in real-time. Not adapted to
bulk configuration, the On-line configuration mainly concerns specific operations such as extending the
network, adding NEs, etc.
WPS
S D
XML snapshots
Import/Export
Import/Export
WPS
At any time during the network building steps, the datafiller can export part of his work towards other platforms.
The following Configuration Management (CM) files can be exported:
Snapshots
Work orders
According to the option selected, the result of the export will be very different:
Exportingthe current state of the network as a snapshot means merging the elementary operations performed
by the work orders with the initial snapshot. As WPS says: the current planning view is the result of the
execution of work orders on the initial snapshot.
Exporting the work order means gathering all the elementary operations performed upon a snapshot into a CM
file for further use (other WPS platforms, W-NMS).
FDDCell /0 BTSCell /0
NodeB BTSEquipment
FDDCell /1 BTSCell /1
RNC
CN IN PP15K -IN
RNCEquipment
MIB
RAN Model
Based
NodeB
Control
Node
W-NMS
WiPS
MIB
Passport
Based
Access Interface
Node Node
Two different types of parameters are designed to configure a UMTS Access Network:
Control Node, Node B and RAN parameters
Interface Node, Access Node and Passport parameters
Changing parameter values may impact the behavior of the live network.
For RAN parameters, the impact triggered by a parameter modification is strongly linked with the parameter classes
(see next slide).
For Passport parameters, it is not always easy to predict the impact of a parameter modification. Possible
consequences are:
nothing
reset of an interface
reset of a module
reset of a node
RNC OMC
Static
Non-Static Customer
Example:
There are two main kinds of parameters in the Alcatel-Lucent system: static and configuration parameters.
The configuration parameters have the following characteristic: they are contained in the Access OAM database.
The customer parameters have been reviewed and tagged with the following rules:
System_restricted:
Parameters which should not be modified in live networks. Proposal also to align the settings for these
parameters on WNE templates at upgrade
Customer_setting:
Parameters which have to be set by the customer, either due to design or to activate optional features
Expert_tuning:
Parameters which can be modified by the customer, but with a specific support from Alcatel-Lucent, because of
the complexity or sensitivity of this parameter with respect to QoS.
Customer_tuning:
Parameters which can be modified by the customer, without specific support from Alcatel-Lucent
MIB
Class 2
MIB build required
(most common case)
lock/unlock required
Class0: value set at object creation. Parameters require a build to be implemented on the NEs large impact on
system
Class2: parameters require a lock of the object(or its parent object) in order to change the parameter value - slight
impact on system
Class3: the parameter can be changed online without impact on the service. Three sub-classes are derived from
Class 3:
Class 3-A1: new value is immediately taken into account.
Class 3-A2: new value is taken into account upon event reception (service establishment, SRLR, LCS, etc.).
Class 3-B: new value is taken into account for next calls.
Customer: the parameter is configurable from the OMC and seen by the operator
Manufacturer: the parameter is configurable from the OMC and only seen by Alcatel-Lucent engineering teams
Example:
Allowed
Lock/unlock required
The object activation class defines the object behavior with respect to the create online and delete online
operations.
Not Allowed: the object cannot be created/deleted online. A build is required.
AllowedWith Parent: the object can be created/deleted online but the operation requires the creation/deletion
of the parent.
Allowed With Lock: the object can be created/deleted online but the operation requires locking the object or
one of its ancestors in the containment tree.
Allowed: the object can be created/deleted online.
RadioAccessService
DedicatedConf
Instance...
Instance...
Instance...
Instance...
Instance3
Instance3
Instance2
Instance2
Instance2
Radio Resource Management is an essential piece of the RNC controlling the radio resources allocated to the users.
The RadioAccessService object is the root of the RRM architecture. It includes a set of parameters that apply to the
whole Radio Network Subsystem.
Some of the parameters of the RRM tree are stored in libraries composed of Configuration Classes and Configuration
Classes Instances.
There are 7 Main Configuration Classes, some of them containing children:
CacConfClass
HoConfClass
MeasurementConfClass
NodeBConfClass
PowerConfClass
PowerPartClass
PowerCtrlConClass
Each of these Configuration Classes can have a maximum of 5 different instances. Each instance then corresponds
to a predefined set of parameters (see example next page).
RNC
RadioAccessService
DedicatedConf
FDDCell
Example
CacConfClass//
CacConfClass 0
firstRlOvsfCodeCacThreshold
maxDlEstablishmentRbRate 85
384
FDDCell
maxUlEstablishmentRbRate 384
64
cacConfId 0
maxUlInterferenceLevel -50.0
powerPartId
powerCtrlConfId
hoConfId
cacConfId
measurementConfId
In the example above, we can see that each instance of CacConfClass includes a set of predefined parameters.
Each parameter belonging to the CacConfClass object can take a different value under each instance.
For example, the maxUlInterferenceLevel can take values from -112 dBm to -50 dBm according to the selected
instance.
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
1 RNC configuration 7
1.1 Identification & capacity 8
2 Node B configuration 9
2.1 FDDCell identifiers 10
2.2 Channel numbers 11
2.3 Scrambling codes 12
2.4 Automatic carrier switch off 13
2.4.1 Switch off description 14
2.4.2 Switch on description 15
2.4.3 RAN model 16
2.5 Node B capacity licensing 17
3 Neighboring cell configuration 19
3.1 FDDCell, RemoteFDDCell, GSMCell 20
3.1.1 Corresponding RAN model 21
3.2 UMTSFddNeighbouringCell 22
3.3 GsmNeighbouringCell 23
3.4 UmtsNeighbouringRelation 24
3.4.1 Example 25
3.5 SIB11/DCH neighboring lists 26
serviceGroupId (NodeB)
NodeB
NodeB
The different RNCs of the network are simply identified by their rncId under the RNC object.
The capacity of the RNC in terms of number of calls, number of supported BTSs and cells depends on two major
factors:
The number of TMUs (hardware configuration)
The software configuration.
The cnodeCapacity parameter is no longer supported from UA06.
From UA06, the service group to which a given Node B is allocated can be specified by the operator
(serviceGroupId parameter of the NodeB object).
It is also possible to know on which TMU an RNC has set a given service group. Thus with these facilities, it is now
possible to modify the affectation of one Node B from a given service group to another one (and consequently from
a PMC-TMU to another one). This possibility may be interesting in order to better balance the load between the
PMC-TMU if needed. Nevertheless, it has to be noticed that the re-affectation of one Node B from a Service Group to
another one implies a loss of service on this Node B.
The number of service groups of an RNC is specified by the numberOfServiceGroups parameter of the RNC
object.
The serviceGroupId parameter of the Iub object specifies which service group this Iub interface is assigned to. All
IubIfs provisioned with the same serviceGroupId will be processed by the same PMC-TMU processor. The
serviceGroupId provisioned on this IubIf must match the serviceGroupId configured on the RNC NodeB managed
object.
Service group id refers to a given TMU board and that numberOfServicesGroup = total number of TMU boards.
Operator (OAM)
RNC
NodeB BTSEquipment
rncId (RNC)
+
cellId (FDDCell)
FDDCell BTSCell
=
ucid (FDDCell)
localCellId (FDDCell)
localCellId (BTSCell)
The standardization of the Iub interface has driven Alcatel-Lucent to define an object model based on a logical part
and a physical part in order to cope with the multi-vendor configurations:
The logical part of the equipment (Node B and RNC) is managed by the OMC-R.
The physical part of the equipment (BTS) is managed by the OMC-B.
The mapping between the two parts is ensured by the localCellId parameter, coded into 28 bits, found under the
FDDCell and BTSCell objects. It is recommended to have a unique localCellId in the whole network for OAM
purposes, to prevent problems during neighboring declaration.
In the UTRAN, the different cells (part of the Node Bs) are identified uniquely by their ucid.
This ucid contains the identifier of the RNC, the rncId, coded into 12 bits, defined under the RNC object and also the
cellId, coded into 16 bits, defined under the FDDCell object.
RNC
NodeB
Operator (OAM)
FDDCell
primaryScramblingCode (FDDCell)
aichScramblingCode(RACH)
sccpchDlScramblingCode(SCCPCH)
Channelizationcode 1
User 1 signal
Channelizationcode 2
User 2 signal
Channelizationcode 3
User 3 signal
A UE is surrounded by Base Transceiver Stations (BTSs), all of which transmit on the same W-CDMA frequency.
It must be able to discriminate between the different cells of different base stations and listen to only one set of
channel codes. Therefore, two types of codes are used:
DLChannelization Code.
The user data are spread synchronously with different channelization codes. The orthogonality properties of
OVSF enables the UE to recover each of its bits without being disturbed by other user channels.
DLScrambling Code.
Scrambling is used for cell identification.
The automatic switch off configurations allowed are STSR 1+1, STSR 2+1 and STSR 2+2.
The automatic carrier shutdown in case of power supply alarm is supported on STSR x+y with AC supply.
The procedure used to unload the cells of a carrier to be shut down is the same as the one used by the Node B Soft
Shutdown feature available from UA06.
The automatic reconfiguration of the remaining carrier for HSDPA support is in restriction.
This means that if R99 is configured on F1 & HSxPA on F2 on a STSR1+1 configuration, when F2 will be shut down,
the customer will lose the HSxPA coverage.
baseAndOverlayCellPowerThres
holdForSwitchOff(autoSwitchO
overlayCellPowerThresholdForSwitchOff ffConfClass)
(autoSwitchOffConfClass)
F1 PA Occupancy
F1 + F2 PA Occupancy
baseAndOverlayCellLoadThres
RL Load on F1 + F2 < holdForSwitchOff(autoSwitchO
F2 ffConfClass)
Feature benefits:
ACSO in case of low traffic allows saving energy during off-peak hours: when all the cells relating to a particular
Power Amplifier of a Node B are switched off, then the Node B will switch off the Power Amplifier (the Power
Amplifier is in standby mode : the radio power module is switched off, while the digital unit remains alive).
ACSO in case of AC mains failures allows saving battery energy in degraded conditions: when the Node B PA
loses AC mains power, backup battery takes over and the node B informs the RNC that this event occurred.
The RNC starts to switch off all the cells relating to that PA. Once all the cells are switched off, the Node B will
switch off the PA.
Switch on when:
baseCellPowerThresholdForSwitchOn
(autoSwitchOffConfClass)
F1 baseCellLoadThreshold
RL Load on Overlay Cell F2 > ForSwitchOn(autoSwitc
hOffConfClass)
F2
RadioAccessService
NodeB
isUplinkRadioLoadEnabled
FDDCell
isAutomaticSwitchOff
cellIdForSwicthOn ShutdownConfClass 132 AutoSwitchOffConfClass 132
shutdownConfClassId
autoSwitchOffConfClassId softShutdownTimer IsAlarmGenerated
isHSxPAToBeReconfigured cellShrinkStep shutdownConfClassId
cellShrinkStepDuration filterCoefficientForSwitchOff
IsEmergencyCallChecked filterCoefficientForSwitchOn
additionalTimerForEmergencyCall startTimeValiditySwitchOff
endTimeValiditySwitchOff
Weighting fators (W1-W18)
baseCellPowerThresholdForSwitchOn
baseAndOverlayCellPowerThresholdForSwitchOff
baseCellLoadThresholdForSwitchOn
overlayCellPowerThresholdForSwitchOff
choiceOfFormulaForSwitchOff
baseAndOverlayCellLoadThresholdForSwitchOff
overlayCellLoadThresholdForSwitchOff
Node B
Per OMC capacity licenses
R99 CE capacity
xTRM
xTRM i or
Node B xCEM/eCEM HSDPA capacity
Node B MCPA RRH
Node B EDCH capacity
Node B
This feature provides the technical base for Pay-as-you-grow commercial schemes. With a licensing scheme in
place, the operator can order HW with a reduced capacity and subsequently purchase licenses for additional
capacity.
Node B capacity licenses are per OMC; the operator can distribute capacity between controlled BTSs via licensing
parameters.
The following Node B capacity aspects are managed in UA06 via this feature: CEM R99 capacity, CEM HSDPA
capacity, CEM HSUPA capacity, xTRM capacity, RRH capacity, PA power and RRH power.
Additional capacity license is OMC wide and can be distributed between the controlled BTSs (intra-OMC). There is no
exchange of licenses between OMCs.
License file: it is a file describing the total capacity (temporary or permanent) allocated for all BTSs of a given OMC.
This file is protected by a digital signature.
HW Capacities: The customer can purchase HW and capacities independently.
BtsEquipment
r99NumberCECapacityLicensing hsdpaMaxThroughputHbbu*
hSDPANumberUserCapacityLicensing
PaResourceRRH hsdpaMaxThroughputXcem
hSDPAThroughputCapacityLicensing
maxRrhPowerAmplification
edchNumberUserCapacityLicensing edchMaxThroughputXcem
edchThroughputCapacityLicensing edchMaxThroughputEcem
xtrmCarrierCapacityLicensing
rrhCarrierCapacityLicensing
rrhTrduThirdCarrierCapacityLicensing
rrhNumberFrequency
additionalRadioPerSectorCapacityLicensing
paPowerCapacityLicensing
The existing UA06 parameter (rrhCarrierCapacityLicensing) represents from UA07 the number of tokens to
allow the configuration of the second carriers of RRH and TRDU and supplementary carriers (3rd and 4th) if all
second carriers have been configured.
The new parameter (rrhTrduThirdCarrierCapacityLicensing) defines the number of supplementary 3rd or
supplementary (4rth) Rx carriers allowed for the pool of RRHs and TRDUs.
If the 2nd carrier capacity licensing is infinite, the 3rd carrier capacity licensing becomes obsolete and all 2nd and
supplementary carriers will be configured following the pseudo parameter rrhNumberFrequency.
[xCEM][eCEM] The maximal capacity can be further controlled (limited) by the parameter
hsdpaMaxThroughputXcem.
For xCEM and eCEM, the processing resources can be configured accordingly with the license agreement via
edchMaxThroughputXcem and edchMaxThroughputEcem which indicates the maximum throughput in kbps
that can be used for E-DCH traffic (MAC-e layer).
RNC1 RNC2
C must be declared as B must be declared as
RemoteFDDCell in RNC1 RemoteFDDCell in RNC2
controls controls
GSMCell GSMCell
E F
BSC
Adjacency between cells (also called Neighboring relation between cells) represents a couple of cells (A and B)
relation where Cell Reselection and/or Soft Handover procedures can be performed from one cell A called Source cell
or Serving cell towards another cell B called Target Cell.
An adjacency is therefore a directional relation between two cells and should rather be noted (A->B).
Adjacencies managed by an RNC can be of 3 kinds:
3G->3G adjacencies where the 3G Target cell is controlled by the same RNC as the CRNC of the Serving cell.
3G->3G adjacencies where the 3G Target cell is controlled by another RNC than the CRNC of the Serving cell.
In this case, a new RemoteFDDCell object must created in the RNC object tree of the CRNC of the Serving cell.
This RemoteFDDCell object shall represent the Target 3G cell in the CRNC of the Serving cell.
3G->2G adjacencies
In this case, a new GSMCell object must created in the RNC object tree of the CRNC of the Serving cell. This
GSMCell object shall represent the 2G Target cell in the CRNC of the Serving cell.
RNC/1
All RemoteFDDCell objects of an RNC are grouped under the UmtsNeighbouring child object of this RNC.
All GSMCell objects of an RNC are grouped under the GSMNeighbour child object of this RNC.
NodeB/x UmtsNeighbouring/0
UMTSFddNeighbouringCell / A UMTSFddNeighbouringCell / C
B->A adjacency:
FDDcellB and FDDCellA objects being defined under RNC1, the B->A adjacency is configured by creating under
FDDCellB a child object UMTSFddNeighbouringCell instance having its userLabel parameter value set to A.
B->C adjacency:
FDDcellB object being defined under RNC1 and FDDCellC object under RNC2, the B->C adjacency is configured in
two steps:
1. by creating under RNC1 a RemoteFDDCell object instance having its userLabel parameter set to C.
2. by creating under FDDCellB a child object UMTSFddNeighbouringCell instance having its userLabel
parameter value set to C.
GSMCell
E
NodeB/x GSMNeighbour/0
FDDCell/B GSMCell / E
GsmNeighbouringCell / E
B->E adjacency:
FDDcellB object being defined under RNC1, the B->E adjacency is configured in two steps:
1. by creating under RNC1 a GSMCell object instance having its userLabel parameter set to E.
2. by creating under FDDCellB a child object GsmNeighbouringCell instance having its userLabel parameter value
set to E.
Outdoor Outdoor
RNC/1 Cell Cell
UmtsNeighbouringRelation/outdoor->outdoor
UmtsNeighbouringRelation/outdoor->indoor
UmtsNeighbouringRelation/indoor->outdoor
UmtsNeighbouringRelation/indoor->indoor
In the UMTS network, some radio parameters used for Cell Reselection and Soft Handover procedures are defined at
Adjacency level. In order to limit the size of the Parameter Database in the RNC, these parameters cannot be set
per 3G->3G adjacency object (UMTSFddNeighbouringCell) but per Object Class object called
UMTSNeighbouringRelation as one dont expect to have very different values from one adjacency to another. Values
usually differ according to the network design, morphostructure and traffic mix in a certain network area.
RNC/1
Outdoor Outdoor
FDDCell FDDCell
NodeB/x UmtsNeighbouring/0 A B
Indoor
Cell C
FDDCell/A UmtsNeighbouringRelation/outdoor->outdoor
UmtsNeighbouringRelation/outdoor->indoor
UMTSFddNeighbouringCell / B UMTSFddNeighbouringCell / C
umtsNeighRelationId umtsNeighRelationId
= UmtsNeighbouringRelation/outdoor->outdoor = UmtsNeighbouringRelation/outdoor->indoor
A->B adjacency:
A->B Cell Reselection parameters shall be set to the Outdoor->Outdoor type of reselection. Therefore, the
configuration shall be done in two steps:
1. by creating under RNC1 a UmtsNeighbouringRelation object instance having its userLabel parameter set to
Outdoor->Outdoor (for instance).
2. by setting for UMTSFddNeighbouringCellB child object of FDDCellA its umtsNeighRelationId parameter value
set to Outdoor->Outdoor.
A->C adjacency:
A->C Cell Reselection parameters shall be set to the Outdoor->Indoor type of reselection. Therefore, the
configuration shall be done in two steps:
1. by creating under RNC1 a UmtsNeighbouringRelation object instance having its userLabel parameter set to
Outdoor->Indoor (for instance).
2. by setting for UMTSFddNeighbouringCellC child object of FDDCellA its umtsNeighRelationId parameter value
set to Outdoor->Indoor.
OR DCH DCH
UE
Measurement Control DCH DCH DCH
SIB1 SIB1
RRC 1 1
+ +
DCH SIB1 DCH
1
DCH + DCH
SIB1 DCH SIB1
sib11AndDchNeighboringFddCellAlgo 1 1
DCH + + DCH
(FDDCell) SIB1 DCH DCH SIB1
1 FDDCel 1
classic +
DCH SIB1 l SIB1
+
DCH
manual DCH +
1
+
1
DCH
DCH SIB1 DCH
DCH 1+ DCH
sib11OrDchUsage (UMTSFddNeighboringCell) SIB1 DCH SIB1
1 1
sib11OrDchUsage (GSMNeighbouringCell) + +
DCH DCH
DCH DCH DCH
sib11AndDch
sib11Usage DCH DCH
dchUsage
Neighboring:
DCH: max 96 (32 intra-freq + 32 inter-freq + 32 GSM)
Idle, PCH, FACH: max (intra-freq+inter-freq+GSM)
Max 96 if isEnhancedSib11Allowed = True
Max 48 if isEnhancedSib11Allowed = False
isSib11MeasReportingAllowed:
True: the UE will use intra-freq neighboring in SIB11/SIB12 in DCH state (in addition to idle, PCH, FACH)
until it receives the list in an RRC measurement control message.
False: when the UE enters the DCH state, it must wait for the RRC measurement control message to get
the measurement list.
If isEnhancedSib11Allowed is set to FALSE, the SIB11 neighboring list shall usually be a subset of the Cell_DCH
connected mode neighboring list.
The algorithm used to build SIB11 and RRC Measurement Control, for 3G frequency measurements is set by the
value of sib11AndDchNeighbouringFddCellAlgo:
classic (or unset): no distinction between SIB11 and DCH neighboring lists.
automatic: the RNC automatically chooses intra-frequency neighboring cells broadcasted in SIB11 (not
supported in this release).
Manual algorithm is preferred to declare and control correctly the list of neighboring cells, thus allowing to make
differentiation between the configuration of SIB11 neighborhood (i.e. while in idle, PCH and Cell_FACH modes) and
Cell_DCH connected mode neighborhood.
The differentiation is set through the SIB11OrDchUsage parameter on each UmtsFddNeighbouringCell and
GsmNeighbouringCell.
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
1 Radio bearers 7
1.1 Signaling Radio Bearers (SRBs) 8
1.2 Conversational radio bearers 9
1.3 Streaming radio bearers 10
1.4 Interactive/background radio bearers 11
2 Services 13
2.1 Mono and multi-RAB services - examples 14
2.1.1 DCH 16
2.1.2 HSxPA 17
3 Multi-rate AMR 19
3.1 AMR NB configurations 20
3.2 AMR NB TB definition 21
3.3 AMR WB TB definition 22
3.4 UL AMR codec mode adaptation 23
3.5 Multi-rate AMR activation NB and WB 24
3.6 Multi-Rate AMR call setup 25
RNC
RadioAccessService
DlRbSetConf UlRbSetConf
The SRB #5 for AMR rate control (SRB_5_AMR) is not supported in UA07. Thus, the flag isSrb5Allowed must be set
to false.
RadioAccessService
DlRbSetConf UlRbSetConf
The standard voice call consists of two narrowband (300-3400 Hz) sound channels, one in each direction, and that
operate independently.
CS_AMR_NB stands for AMR Narrow Band RB for which AMR NB voice codecs used allow a DL SF of 128 if AMR
RAB is not multiplexed with another RAB.
CS_AMR_LR stands for AMR Low Rate RB for which AMR LB voice codecs used allow a DL SF of 256 if AMR RAB
is not multiplexed with another RAB.
CS_14_4K RB corresponds to the CS data service also provided in GSM networks.
CS_57_6K RB corresponds to the CS data service also provided in HSCSD GSM networks.
CS_64K RB corresponds to the Video Call service not available in GSM networks.
Note: CS_64K_Scudif RB Switching between Video and Speech is optionally supported from UA07.1.2.
RNC
RadioAccessServic
e
DlRbSetConf UlRbSetConf
Streaming PS_384K_STR 10 ms
PS_xxx_STR RB >= 256 kbps are provided for High Quality streaming services which require a higher bandwidth.
PS_8K_IB 40 ms PS_8K_IB 40 ms
PS_8K_IB_MUX 40 ms PS_8K_IB_MUX 40 ms
PS_16K_IB 40 ms PS_8K_IB_MUX3 40 ms
Interactive/
Interactive/ PS_32K_IB 40 ms PS_16K_IB 40 ms
Background
Background PS_64K_IB 20 ms PS_32K_IB 40 ms
PS_64K_IB_MUX 20 ms PS_32K_IB_MUX 40 ms
PS_64K_IB_MUX3 20 ms PS_32K_IB_MUX3 40 ms
PS_128K_IB 20 ms PS_64K_IB 40 ms
PS_128K_IB_MUX 20 ms PS_64K_IB_MUX 20 ms
PS_128K_IB_MUX3 20 ms PS_64K_IB_MUX3 20 ms
PS_256K_IB 10 ms
DlRbSetConf UlRbSetConf
DL 64 DL 64 DL
3,4
DCH DCH
DL 64 DL
SF32 3,4
DPCH
SF32
UL 384+384 Mux (the existing support is only 128+128 Mux for UL)
PS_384K_IB_MUX 10 ms PS_128K_IB_MUX 20 ms
PS_HSDCH_IB 2 ms PS_128K_IB_MUX3 20 ms
Interactive/
PS_HSDCH_IB_MUX 2 ms PS_384K_IB 10 ms
Background
Interactive/
PS_HSDCH_IB_MUX3 2 ms PS_384K_IB_MUX 10 ms
Background
TRB_CellFACH N.A. PS_384K_IB_MUX3 10 ms
TRB_CellFACH N.A.
TRB_CellFACH_MUX N.A.
DlRbSetConf UlRbSetConf
Before UA07, CS AMR NB versions were NB1: [12.2, 7.95, 5.9, 4.75] and NB2: [10.2, 6.7, 5.9, 4.75]
From UA07 onwards, CS NB2 becomes [12.2, 7.4, 5.9, 4.75]
CS_64KxPS_0K_IBxSRB_3_4K CS_AMR_WBxPS_0K_IBxSRB_3_4K
CS_64KxPS_16K_IBxSRB_3_4K CS_AMR_WBxPS_8K_IB_MUXxSRB_3_4K
CS_64KxPS_64K_IB_MUXxSRB_3_4K CS_AMR_WBxPS_16K_IBxSRB_3_4K
CS_64KxPS_128K_IB_MUXxSRB_3_4K CS_AMR_WBxPS_384K_IBxSRB_3_4K
CS_AMR_LRxPS_0K_IBxSRB_3_4K CS_AMR_WBxPS_384K_STRxPS_8K_IBxSRB_3_4K
CS_AMR_LRxPS_8K_IBxSRB_3_4K CS_AMR_WBxPS_384K_STRxPS_HSDSCHxSRB_3_4K
CS_AMR_LRxPS_HSDSCHxSRB_3_4K CS_AMR_WBxPS_384K_STRxSRB_3_4K
CS_AMR_NBxPS_0K_IB_MUXxSRB_3_4K PS_0K_IB_MUXxSRB_3_4K
CS_AMR_NBxPS_0K_IBxSRB_3_4K PS_0K_IBxSRB_3_4K
CS_AMR_NBxPS_8K_IB_MUXxSRB_3_4K PS_8K_IB_MUXxSRB_3_4K
CS_AMR_NBxPS_16K_IBxSRB_3_4K PS_16K_IBxSRB_3_4K
CS_AMR_NBxPS_16K_STRxPS_8K_IB_MUXxSRB_3_4K PS_16K_STRxPS_0K_IBxSRB_3_4K
CS_AMR_NBxPS_64K_STRxPS_8K_IB_MUXxSRB_3_4K PS_16K_STRxPS_8K_IB_MUXxSRB_3_4K
CS_AMR_NBxPS_128K_STRxPS_8K_IB_MUXxSRB_3_4K PS_64K_STRxPS_0K_IBxSRB_3_4K
CS_AMR_NBxPS_384K_IBxSRB_3_4K PS_64K_STRxPS_8K_IB_MUXxSRB_3_4K
CS_AMR_NBxPS_384K_STRxPS_8K_IBxSRB_3_4K PS_128K_STRxPS_0K_IBxSRB_3_4K
CS_AMR_NBxPS_384K_STRxPS_HSDSCHxSRB_3_4K PS_128K_STRxPS_8K_IB_MUXxSRB_3_4K
CS_AMR_NBxPS_384K_STRxSRB_3_4K PS_384K_STRxPS_8K_IBxSRB_3_4K
DCH
Stand-alone
CS Conversational speech: AMR_LR, AMR_NB, AMR_WB
CS Conversational VT: 64/64
CS Streaming: 14.4/14.4, 57.6/57.6
PS Streaming DL: 16,32,64,128 UL: 16,64,128,256,384
PS I/B DL: 8,16,32,64,128,256,384 UL: 8,16,32,64,128,384
Combination
CS Conv. Speech + PS I/B DL: 0,8,16,32,64,128,384 UL: 0,8,16,32,64,128,384
CS Conv. VT + PS I/B DL: 0,8,16,32,64,128 UL: 0,8,16,32,64,128,384
(CS Conv. Speech +) PS I/B MUX DL: 0,64,128,384 UL: 0,64,128
CS Conv. VT + PS I/B MUX DL: 0,64 UL: 0,64,128
(CS Conv. Speech +) (PS I/B +) PS Streaming:
PS Streaming DL: 16,64,128,256,384 UL: 16,32,64,128
PS I/B DL: 8 UL: 8,16,64,128,256,384
UL 384+384 Mux (the existing support is only 128+128 Mux for UL)
CS Voice AMR NB (UL/DL 12.2 kbps or 12.2/7.95/5.9/4.75 kbps or 12.2/7.4/5.9/4.75 kbps, with VAD/DTX)
HSxPA
Stand-Alone
PS I/B HSDPA/DCH DL: f(HSD UE category)
UL: 8,16,32,64,128,384
PS I/B HSDPA/HSUPA DL: f(HSD UE category)
UL: f(HSU UE category, TTI)
PS Streaming HSDPA/DCH DL: (HSD UE category, GBR)
UL: 16,32,64,128
PS Streaming on HSDPA
UL: 16, 32, 64, 128
DL: Max Bit Rate depending on UE Category & GBR
The Multi-rate AMR feature consists of the introduction of a certain number of Multi Mode configurations of the AMR
for the speech service:
A. 12,2 7,95 5,9 4,75
B. 5,9 4,75
C. 4,75
D. 12,2 7,4 5,9 4,75
E. 12,2
All these configurations can be used together with I/B PS services but B and C which are intended to be used with
Spreading Factor 256 in DL, e.g. in capacity limited networks.
Configuration E is intended for legacy purposes. It is the only one which is compatible with the Iu User Plane Frame
protocol V1 (see 3GPP TS 25.415). Other configurations required Iu UP FP V2.
On the radio interface, one dedicated transport channel is established per class of bits, i.e. DCH A for Class A bits,
DCH B for Class B bits and DCH C for Class C bits. Thus, each class can be subject to a different error protection
scheme:
Class
A contains the bits that are most sensitive to errors. Any error in these bits would result in a corrupted
speech frame which would need error correction for proper decoding. It is the only class protected by a CRC.
Classes
B and C contain bits for which increasing error rates gradually reduce the speech quality. However, the
decoding of an erroneous frame can be done without significantly degrading the quality.
The wideband AMR codec consists in 9 sources with bit rates of 23.85k, 23.05, 19.85k, 18.25k, 15.85k, 14.25k,
12.65k, 8.85k and 6.6k. Only 5 modes are used and supported for telephony (3.85k 15.85k 12.65k 8.85k and
6.6k). The other modes are used for other services (e.g., for MMS).
Uplink: 64
At AMR WB Call Setup, the Max Bit rate is initialized to the Max bit rate which is 12.65.
AMR mode
(kbps)
12.2
- 10.2
7.95
UE 7.40
output 6.70
power 5.90
5.15
+ 4.75
The new CS NB2 configuration (12.2, 7.4, 5.9, 4.75) is supported as a replacement of (10.2, 6.7, 5.9, 4.75) kbps.
AMR rate change
For Multi Mode configurations (i.e. A, B and D), the speech rate can change in UL and DL.
DL rate is set according to the rate of the Iu UP Frames received from the CN.
ULrate can change either on decision of the UE according to its TFCS selection function or on request of the CS
CN. This latter case can happen when TFO/TrFO is used in Mobile-to-Mobile calls.
AMR configuration at call set-up
TheAMR configuration is selected according to the CS CN request at call setup. If the CN supports Iu UP FP v1
only Configuration E will be used.
If it supports v2 it must indicate one of the A to D configurations.
AMR code mode adaptations occur in both UL and DL for configurations A, B, D for AMR-NB
and for AMR-WB:
In
DL, the AMR rate adaptation occurs in the TFO/TrFO scenario when the distant UE changes its bit rate (also
when the RNC changes the max DL bit rate).
InUL, the UE can select a different AMR rate in case of coverage limit. The UE transmitted power is closed to
the maximum. In this case, the UE can reduce its AMR rate.
isAmrWbAllowed (RadioAccessService)
isAmrMultiModeAllowed (RadioAccessService)
isAmrMultiModeSetupAllowed (FDDCell)
Multi-rate AMR activated in a cell?
isAmrMultiModeSetupAllowed
True False
isAmrMultiModeAllowed
disabled NO NO
enabledPerCell YES NO
The AMR configuration can be specified at call setup through the SRB #5 if present.
The SRB #5 contains the following information:
Signaling RB information to set up
Authorized TFC subset list to be used in UL.
SRB 5 is set up if all of the following conditions are met:
isCnInitiatedRateControlAllowed is True.
isSrb5Allowed is True.
Version 2 of Iu UP is selected.
At least two speech modes are selected (silent mode excluded).
TheUE indicated 3GPP release (UE radio access capability / Access stratum release indicator) is greater than or
equal to the provisioned value of minUeRelForSrb5Amr.
Starting from UA07, SRB#5 is not supported by ALU UTRAN, therefore isSrb5Allowed must be set to False and
SRB#2 is used to transfer AMR adaptation signaling instead of SRB#5.
In UA05.0, the initial AMR codec used at call setup was fixed and equal to the maximum rate allowed among the
ones of the Multi-mode configuration used.
In UA05.1, the initial AMR codec used at call setup could be chosen by the operator thanks to the two following
parameters:
isMaxDlAmrRateConfiguredAllowed is the activation flag for control of the maximum downlink rate for AMR
Narrowband calls based on provisioned cell parameters.
maxDlAmrRateConfigured is the maximum downlink rate for Narrowband AMR calls in the cell.
isCsRabModificationForSpeechAllowed is the activation flag for CS RAB modification between AMR NB and AMR WB
configuration.
Currently, SRB#5 is not used since it is not supported by all UEs.
Copyright 2012 Alcatel-Lucent. All Rights Reserved.
TMO18255_V2.0-SG-UA08-Ed1 Module 1.3 Edition 1
Section 1 Module 3 Page 25
End of module
Services
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
RNC
The Node B has to provide two types of measurements: common measurements and dedicated measurements.
These measurements are also called NBAP measurements because they are reported to the RNC using NBAP
messages.
Beside the NBAP measurements, the BTS also provides measurement results that are sent in-band.
The UE has to be capable of performing 7 different measurement types: intra-frequency, inter-frequency, inter-
system, traffic volume, quality, UE-internal and UE positioning. These measurements are also called RRC
measurements because they are reported to the RNC using RRC messages.
From UA05, the Node B only removes the contribution of HSDPA channels (it does not remove the E-DCH
contribution) to the power measurement. This leads to slightly overestimation of the R99 contribution and impact
DCH call admission control. This effect can be attenuated by increasing DCH admission threshold on power for
HSUPA cells.
Path loss (Inter System Measurements): this measurement figures in the reporting message in the UMTS
specifications but does not correspond to any measurement defined in the GSM specifications.
This measurement is not managed by the RNC.
RRC
Measurements
REPORTING
UE
ESTIMATING
FILTERING
FN = (1 - a).FN-1 + a.MN
filter coefficient
reporting period
acquisition time or
event triggered RNC
NodeB
NBAP
Measurements
Physical Layer provides measurements to the upper layers (Layer 3). For each measurement, a basic measurement
period is defined, which corresponds to the shortest averaging period and also to the shortest reporting period, i.e.
the Node B or UE cannot be required to report a measurement to the RNC in a shorter time period.
Before reporting to the RNC, the Node B or UE Layer 3 performs a filtering operation averaging several
measurements and allowing them to create measurement reports with a period not necessarily equal to the basic
measurement period. The filtering parameter a is defined as a = 1/2k/2, where k is the parameter received in the
Measurement Filter Coefficient IE.
The reporting period for each measurement is configured by the RNC when the UE or the Node B is requested to
perform measurements. The minimum reporting period for each measurement is equal to the basic measurement
period for this measurement. In general, the reporting period is a multiple of the basic measurement period.
For UTRAN measurements reported in-band, the reporting period is the period at which frames are sent from the
Node B to the serving RNC.
isEventTriggeredMeasAllowed isInterFreqMeasActivationAllowed
(FDDCell) (RadioAccessService)
IsInterfreqCModeActivationAllowed
isGsmCModeActivationAllowed
RNC (DlUserService)
ueInternalMeasurementQuantity
ueInternalMeasurementFilterCoeff
(UEIntMeas)
measurementConfId
(FDDCell)
measurementConfClassId
(NeighbouringRNC)
Each FDDCell and NeighbouringRNC must have a pointer to one of the Measurement Configuration Classes stored
under the RNC they depend upon.
The isEventTriggeredMeasAllowed parameter controls the activation of Full Event Triggered RRC measurement
reports per FDDcell.
The isInterFreqMeasActivationtAllowed parameter controls the activation of inter-frequency RRC measurement
reports whether Inter-FDD or Inter-RAT neighboring cells are to be measured.
The IsInterfreqCmodeActivationAllowed parameter controls the activation of the compressed mode for
inter-FDD neighboring cell measurements.
The isGsmCmodeActivationAllowed parameter controls the activation of the compressed mode for inter-
RAT neighboring cell measurements.
When set to true, the UeInternalMeasurementQuantity parameter allows one to choose which measurement
type is selected among the three available types: ueTransmittedPower, utraCarrierRssi, ueRxTxTimeDiff.
isDetectedSetCellsAllowed
(RadioAccessService)
There are 3 different ways to classify the cells that may be involved in handover procedures:
Thecells belonging to the Active Set are the cells involved in the soft handover and that are communicating
with the UE.
Thecells belonging to the Monitored Set, that do not belong to the active set, but that are monitored by the UE
depending on the neighboring list sent by the UTRAN.
The cells belonging to the Detected Set, which are detected by the UE, but that are neither in the Active Set
nor in the Monitored Set.
isDetectedSetCellsAllowed indicates if the cells of the detected set have to be taken into account for RRC Intra-
Frequency measurement management for the calls established in Event-Triggered Reporting Mode.
Intra/Inter-Frequency Intra/Inter-Frequency
-25 -15-10 0
Received Power on
GSM CARRIER RSSI =
the GSM BCCH carrier
CPICH Ec/No
CPICH Ec/No is the received energy per chip divided by the power density in the band, that is, it is identical to
the RSCP measured on the CPICH divided by the RSSI. The UE has to perform this measurement on the
Primary CPICH and the reference point is the antenna connector of the UE. This measurement is used for cell
selection and re-selection and for handover preparation.
CPICH RSCP
CPICH RSCP is the Received Signal Code Power on one channelization code measured on the bits of the Primary
CPICH. The reference point is the antenna connector at the UE. Although the measurement of this quantity
requires that the Primary CPICH is despread, it should be noted that the RSCP is related to a chip energy and
not a bit energy. This measurement is used for cell selection and re-selection and for handover preparation,
open loop power control and pathloss calculation.
GSM carrier RSSI
Thismeasurement is the wide-band received measured on a specified GSM BCCH carrier. This measurement is
used for GSM handover preparation.
DL outer loop
power control DPCCH
n
SF RSCP io
x = SIR at
2 ISCP DPCCH t im
Es
y
lit
ua
Q
nk
Li
=
R
SI
P-CPICH
Path Loss
Path Loss
FDD Cell
The path loss is calculated by the UE and by the RNC using the following formula: Path Loss = Primary CPICH Tx
Power - P-CPICH_RSCP
The RNC does not ask the UE to send the path loss, because the RNC can calculate it directly.
The path loss calculated by the RNC is used for inter-frequency handover criteria evaluation.
The path loss calculated by the UE is used to define the initial PRACH power.
RNC
Physical
Transport Channels Transport Channels
Channel block block
block block
CRC
Indicato
Soft Handover
r UL outer loop PC
qeSelector (Static)
Frame Protocols
DATA CRCtx
IE QE Selector: Quality Estimate:
CRC INDICATOR
TheCRC indicator is attached to the UL frame for each transport block of each transport channel transferred
between the Node B and the RNC. It shows if the transport block has a correct CRC (0=Correct, 1=Not
Correct). This measurement is used for frame selection in case of soft handover.
QUALITY ESTIMATE
Thequality estimate is reported in band in the UL data frames from the Node B to the RNC and it is derived
from the Transport channel BER or Physical channel BER. If the IE QE-Selector is equal to:
Selected, then the Node B calculates the BER of the transport channel measured on DPDCH per TTI (if
no DPDCH is sent, then the Node B will use DPCCH BER).
non-selected, then the Node B calculates the BER of the physical channel measured on DPCCH per TTI.
Incase of soft handover, the quality estimate is needed in order to select a transport block when all CRC
indications are showing bad (or good) frames. The RNC compares the QE value with the qeThreshold (static
parameter) in order to choose the best transport block.
Quality Estimate can also be used to enhance the UL Outer Loop Power Control mechanism.
C-RNC
Node B Common
Measurement Initiation Request
Dedicated
On Demand
Cell
RACH Common Event-Triggered
Dedicated Periodic
Depending on the type of measurement (common or dedicated), measurement requests are initiated by the
controlling RNC by sending a COMMON MEASUREMENT INITIATION REQUEST or DEDICATED MEASUREMENT
INITIATION REQUEST to the Node B.
The common and dedicated measurement messages both contain the following information elements to define the
measurements to be performed:
A measurement id uniquely identifying each measurement.
A measurement object type to indicate the type of object on which the measurement is to be performed, e.g.,
cell, RACH, time slot, etc. It can be common or dedicated according to the message. In the case of a dedicated
measurement, either a radio link is identified on which the measurement has to be performed or the
measurement should be performed on all radio links for the Node B.
A measurement type indicates which measurement is to be performed. It is also common (Received total
wideband power, transmitted carrier power, Acknowledged PRACH preambles, etc.) or dedicated (SIR,
transmitted code power, In-Band (transport channel BER, physical channel BER), etc.).
A measurement filter coefficient gives the parameter for the layer 3 filtering to be performed before the
measurement can be reported.
The report characteristics give the criteria for reporting the measurement. The reporting is on demand, periodic
or event-triggered.
Node B
C-RNC
commonMeasurementReportingPeriod
commonMeasurementFilterCoeff Common Measurement Reports
(NBAP Measurement)
nbapCommonMeasRtwpReportingPeri
od
nbapCommonMeasRtwpFilterCoeff Measurement Report
Measured Quantity
(NbapMeasRtwpParameters) ID Type
The reports are sent in the COMMON MEASUREMENT REPORT, on criteria defined by the report characteristics given
in the measurement request.
For these Common Measurements, the type of measurement report is defined by the commonRepType parameter
[on demand, periodic, event-triggered].
The quantity measured is defined by the measQuantity parameter.
The periodicity is given in the Report_Periodicit yIE of the measurement request message.
The periodicity is given in the Report_Periodicity IE of the measurement request message and corresponds to the
CommonMeasurementReportingPeriod parameter for DL Transmitted Carrier Power and DL Transmitted Carrier
Power of all Codes not used for HS-PDSCH, HS-SCCH, E-AGCH, E-RGCH or E-HICH transmission.
nbapCommonMeasRtwpReportingPeriod is the reporting period to be applied to UL RTWP measurements.
CommonMeasurementFilterCoeff is the filtering coefficient to be applied to DL Transmitted Carrier Power and
DL Transmitted Carrier Power of all codes not used for HS-PDSCH, HS-SCCH, E-AGCH, E-RGCH or E-HICH
transmission measurements.
nbapCommonMeasRtwpFilterCoeff is the filtering coefficient to be applied to UL RTWP measurement
The measurements reporting by the Node B stop upon reception of COMMON MEASUREMENT TERMINATION
REQUESTs sent by the CRNC if any.
transmittedCodePowerRequired
transmittedCodePowerReportPeridodicity
Dedicated Measurement Termination Request
roundTripTimeRequired
roundTripTimeReportPeriodicity
(NbapDedicatedMeasConfigForCallTrace)
In the case of Dedicated Measurements, three different types of measurements reports are supported (SIR, DL
TRANSMITTED CODE POWER and ROUND-TRIP-TIME). For Call Trace purposes, these three types of reports can be
activated separately and can be configured with different periodicities.
The procedure is initiated with a DEDICATED MEASUREMENT INITIATION REQUEST message sent from the RNC to
the Node B. This procedure is used by an RNC to request the initiation of measurements on dedicated resources (all
UE radio links managed by FDDCells belonging to this Node B).
Upon reception, the Node B shall initiate the requested measurement according to the parameters given in the
request and shall periodically send a DEDICATED MEASUREMENT REPORT.
The procedure is operational as long as the RL is established. The RNC does not send a DEDICATED MEASUREMENT
TERMINATION REQUEST message. Instead, even though the trace session is deleted, the NBAP dedicated
measurement reporting, if initiated, will remain until the radio links associated with the call being traced are deleted
or released.
Round trip time (RTT) is defined as: RTT = TRX - TTX, where:
TTX = the time of transmission of the beginning of a DL DPCH frame to a UE.
TRX
= the time of reception of the beginning (the first detected path in time) of the corresponding UL
DPCCH/DPDCH frame from the UE.
Node B
C-RNC
iRM Scheduling Downgraded UE
Primary Cell
RL Monitoring
Event A
Event B1
Event B2
NBAP Dedicated Report Event Bx
The NBAP event triggered report mode is only used in the scope of iRM scheduling downgrade/upgrade procedures
with the RNC perspective to retrieve the transmitted code power by the Node B for a particular radio link (user) and
to order the radio bearer downsizing/upsizing through iRM scheduling towards the more adapted bit rate to
guarantee service continuity.
For the purposes of iRM Scheduling, the RNC configures the Node B with one Event A and two Events B:
Event A indicates that the radio conditions have become bad enough to attempt a downgrading to the fallback
radio bearer in order to maintain a good radio link quality.
Event B1 indicates that the radio conditions have become good enough to attempt an upgrading towards the
original requested RB.
Event B2 indicates that the radio conditions have become good enough to consider an upsizing towards a
relative lower bit rate than the requested RB to maintain a good radio link quality.
Event A
Report
pcpichPower + maxDlTxPowerPerOls
timeToTrigger
(DlIrmSchedDowngradeTxcp)
In order to be able to perform IRM Scheduling downgrade, the RNC configures the NBAP dedicated measurement by
event A report for this UE on the primary cell.
So, each time the primary cell changes, the RNC terminates measurements on the old primary cell and initiates
measurements on the new primary cell.
SI broadcast
P-CCPCH
OR
UE Node B
Measurement Control
RRC
Measurement Measurement
Measurement Measurement Measurement Measurement Measurement Reporting
Reporting Reporting
ID Command Type Object Quantity Mode
Quantity Criteria
Inter-Frequency
Ec/No
Intra-Frequency FDDCell
Setup RSCP
Inter-System Physical Channel Periodical RLC AM
Modify BLER
Traffic Volume RB Event-Triggered RLC UM
Release Traffic Volume
Quality
...
UE internal
In CELL_FACH, CELL_PCH or URA_PCH state, the UE is informed of the measurements to perform via the system
information broadcast on the P-CCPCH.
When the UE is in CELL_DCH state, UTRAN starts a measurement in the UE by sending the MEASUREMENT
CONTROL message, which includes the following information elements to define measurements to perform:
measurement id is a reference number to be used when modifying or releasing measurement.
measurement command indicates the action performed on the measurement (set up a new measurement,
modify the characteristics of a measurement, etc.).
measurement type indicates one of the different types of measurement: inter-frequency, intra-frequency, etc.
measurement object indicates the object on which the measurement shall be performed.
measurement quantity indicates the quantity to be measured (RSCP, SIR, etc.).
measurement reporting quantity indicates quantities that the UE should report together with the measurement
quantity for example, the measurement quantity which triggered the report.
measurement reporting criteria indicates the type of reporting that is, periodical or event-triggered.
reporting
mode specifies whether the UE shall transmit the measurement report using the acknowledged or
unacknowledged RLC mode.
Measurement Report
rrcIntraFreqMeasurementReportingPeriod
rrcIntraFreqMeasurementFilterCoeff
Measurement Report (RRCMeasurement)
The MEASUREMENT REPORT message is sent from the UE to the UTRAN and contains the measurement id, the
measured results and the measurement event result that was required to be reported.
When the rrcIntraFreqMeasurementReportingPeriod time has elapsed, the UE shall send the computed
measurement.
Reporting Quantities
The RNC requests the following quantities to be reported by the mobiles:
Cell Synchronization information: provides the difference between SFN of the reported cell and CFN as
observed by the UE.
CPICH Ec/No: the received energy per chip divided by the power density in the band.
CPICH RSCP: the received power on one code measured on the Primary CPICH.
Other reporting quantities are also supported by UTRAN and are also requested to the UE for tracing purposes:
SFN SFN observed time difference "type 2": the relative timing difference between cell j and cell i
measured on the primary CPICH.
The RepMode parameter indicates that the UE shall report measurements on cells of the active set and
intrafrequency monitored set cells. This STATIC parameter is set to ACTIVE_SET_AND_MONITORED_ON_FREQ.
The maxCellsRepType parameter indicates the number of measured cells belonging to the monitored set. This
STATIC parameter is set to 6.
sib11IntraFreqMeasurementNbrOfCellOnRACH
sib11IntraFreqMeasurementFilterCoeffOnRACH Unused (25.331 8.6.7.2 Filter Coefficient)
(RRCSysInfoMeas)
SIB 11
Reported measurements on RACH
RACH
CPICH Ec/No
or
CPICH RSCP
or
Neighboring Cells Path Loss
measQuantity
(static)
Measurements reported in the RACH message are used by power allocation and RAB assignment algorithms.
The measQuantity static parameter determines the type of reported measurements. Only the value CPICH_Ec/No
is supported by the measQuantity static parameter.
The sib11IntraFreqMeasurementNbrOfCellOnRACH parameter indicates how many cell measurements shall
be reported in the RACH message, including the current cell; for instance the current cell plus the best 4 measured
neighbors.
The number of reported cells on RACH is used by the compound neighbor list feature to create the neighboring list
for the first Measurement Control Message.
As specified by 3GPP 25.212, no filtering shall be performed for the measurements reported in RACH.
Note: For parity reasons, sib11IntraFreqMeasurementNbrOfCellOnRACH is set to the currentCell value in
USA customer templates.
SI broadcast
P-CCPCH
RNC
UE NodeB
RRC Connection Request
RadioAccessService
Event1AHoConfInSIB11
Measurement Control
isSib11MeasReportingAllowed
cpichEcNoReportingRange1A
(FDDCell) hysteresis1A
timeToTrigger1A
maxActiveSetSize
This feature allows UTRAN to provide intra-frequency measurements configuration information to UEs which are in
Idle Mode or in Cell-FACH. Received within the SIB11, information is used by UEs to activate intra-frequency
measurements just after entering the Cell-DCH state, with no need to wait for the first Measurement Control.
If the reporting mode is Event Triggered, only Event 1A is configured in the SIB11 and the UE sends the first
Measurement Report only if the 1A Event has been reached. The rest of the events are configured in the first RRC
Measurement Control message. Event1AHoConfInSIB11 dedicated object has been created under HoConfClass
so that specific 1A setting can be broadcast in SIB11 for faster measurements.
If the reporting mode is Periodic, the UE keeps on sending reports at the defined period until reception of the first
RRC Measurement Control.
The first RRC Measurement Control message sent to the UE is of type SETUP instead of MODIFY in order to ensure
that there is no misalignment between the UE and the Network.
The UE starts sending measurements when its state changes:
from Idle mode to Cell-DCH (after the RRC Connection Setup)
from Cell-FACH to Cell-DCH (after the RRC RB Setup).
If isSib11MeasReportingAllowed is set to True, SIB11 (but also SIB12 if Sib12Enable is set to True) will
include the 3GPP TS 25.331 Intra-Frequency Measurement Quantity and Reporting information for state
CELL_DCH Information Elements (IEs) which allow the UE to configure the intra-frequency measurement reporting
mode until reception of the RRC Measurement Control message.
If isSib11MeasReportingAllowed is set to True, the serving cell has to be present in the neighboring intra freq
cells list present in SIB11, thereby reducing the maximum number of neighbor cells by 1.
If isSib11MeasReportingAllowed is set to False, SIB11 and SIB12 do not include these IEs and the UE must wait
for the first RRC Measurement Control to measure neighboring cells when entering Cell_DCH.
isEventTriggeredMeasAllowed
(FDDCell)
Measurement
Measurement Measurement
Reporting
ID Results
Quantity
Active Set cells
+
Node B Best Monitored cells
+
Best Detected cells (call trace
only)
CPICH Ec/No
CPICH RSCP isDetectedSetCellsAllowed
(RadioAccessService)
The MEASUREMENT REPORT message is sent from the UE to the UTRAN and contains the measurement id, the
measured results and the measurement event result that triggered the report.
Reporting Quantities
The RNC requests the following quantities to be reported by the mobiles:
CPICH Ec/No: the received energy per chip divided by the power density in the band.
CPICH RSCP: the received power on one code measured on the Primary CPICH.
In
case or Event Measurements Reported for UE tracing, then up to 3 best detected cells can be reported in
some of the Events.
3GPP specifications define 2 RRC measurements reporting modes; periodical reporting and event-triggered
reporting. For the event triggered reporting mode, RRC standards define a set of events for each type of
measurement:
Events 1X are defined for intra-frequency measurements
Events 2X are defined for inter-frequency measurements
Events 3X are defined for inter-RAT measurements
Etc.
Event-triggered reporting is used in Alcatel-Lucent UTRAN for RRC intra-frequency reporting measurements.
The use of event triggered reporting for intra-freq measurements has a direct impact on the following mechanisms:
primary cell determination
active set management
radio link color determination
6B Measid15 UE Tx Power N.A. Estimated UE Tx power becomes less than an absolute threshold
In ALU UTRAN:
RRC inter-frequency measurements on aused frequency are event-triggered (2D and 2F).
RRC inter-frequency measurements on a non-used frequency are periodical.
RRC inter-RAT measurements are periodical.
RRC UE internal measurements are event-triggered (6A and 6B).
The use of event triggered reporting for inter-freq measurements on used frequency and UE internal measurements
has a direct impact on the following mechanisms:
alarm measurement criteria (Compressed Mode and Hard handover triggering).
inter-frequency or inter-RAT blind handover.
The use of event triggered reporting for UE Traffic Volume measurements has a direct impact on the following
mechanisms:
Always-On upsize from FACH to DCH.
RB Rate Adaptation upsize from a given DCH bit rate to a higher one.
Event1A
Event1A
Event1A
amountRep1A
(FullEventRepCritShoMgtEvent1A)
CPICH_EC/No
Best Cell
New Cell
entering reporting range
timeToTrigger1A repInterval1A
(FullEventHOConfShoMgtEvent1A) (FullEventRepCritShoMgtEvent1A)
NA
i 1
10 LogM New CIO New W 10 Log M i (1 W ) 10 LogM Best ( R1a H 1a / 2)
X
wParam
(static)
cpichEcNoReportingRange1A (FullEventHOConfShoMgtEvent1A)
hysteresis1A (FullEventHOConfShoMgtEvent1A) weight1A
(FullEventHOConfShoMgtEvent1A)
neighbouringCellOffset (UmtsNeighbouringRelation)
1st RACH Transport Block 1st Transport Block of 1st DCH Pad.
1st RACH Transport Block Pad. Last Transport Block of 1st DCH
The propagation delay is reported in the RACH data frames transferred from the Node B to the RNC when a
successful RACH procedure has happened and the RACH has been sent from the UE to the RNC.
The CRC Indicator is attached to the UL frame for each transport block of each transport channel transferred
between the Node B and the RNC. It shows if the transport block has a correct CRC.
The Quality Estimate is reported in band in the UL data frames from the Node B to the RNC. This QE corresponds to
either the transport channel BER or the physical channel BER when no transport channel BER is available, that is,
there is no data transmitted in the UL thus only DPCCH is transmitted.
Single-Frame Mode
power
Double-Frame Mode
frame boundary
power
Inter-frequency measurements
Inter-RAT measurements
The Compressed Mode consists of the creation of regularly spaced short gaps (less than one 10-ms radio frame) in
transmission in the uplink or downlink, or possibly both at the same time, and/or reception without altering the data
to be exchanged on the radio interface.
The Compressed Mode is mandatory in downlink and optional in uplink. It can only be achieved on dedicated
channels. The Transmission Gap Length is 3, 4, 5, 7, 10 or 14 slots.
Three methods are proposed in the standard: Spreading Factor Reduction, Puncturing and Higher Layer Scheduling.
Only the Spreading Factor Reduction method is implemented for both UL and DL (when needed) for either GSM or
FDD inter-frequency measurements:
only the value cmodeDlMethodSfDiv2 is allowed for DlCmodeMethod and UlCmodeMethod. Two
Thus,
methods can be used for time transmission reduction.
The
SF can be reduced by 2 to permit the transmission of the information bits in the remaining time slots of the
compressed frame. In that case, the scrambling code could be different from the normal mode.
isInterFreqCModeActivationAllowed
isGsmCModeActivationAllowed:
(DlUserService)
GSM
UMTS
UE ability
Cap UMTS
GSM450Present (RadioAccessService)
Dual receiver UE? GSM480Present (RadioAccessService)
Mono Receiver UE?
GSM compatible UE? GSM850Present (RadioAccessService)
GSM900PPresent (RadioAccessService)
GSM900EPresent (RadioAccessService)
GSM900RPresent (RadioAccessService)
GSM1800Present (RadioAccessService)
GSM1900Present (RadioAccessService)
The real need for Compressed Mode is provided by the UE itself. Following a network request through the UE
Capability required indicator in the RRC Connection Setup message, the UE indicates in the UE Radio Access
Capability IE (Measurement Capability sub-IE, provided in the RRC Connection Setup Complete message) if the
Compressed Mode is needed in either UL or DL for the FDD and GSM bands.
The network configure and activate the Compressed Mode in 3 possible modes:
Uplink only
Downlink only
Uplink and Downlink
Therefore, regarding CM for GSM, in order not to configure the compressed mode in every case, a set of flags
indicating the frequency bands of the FDD and GSM neighboring cells will be defined and used in the RNC to
determine whether or not the Compressed Mode is needed.
Each flag indicates that there exists at least a GSM cell of the corresponding frequency band in the access network
(that is, not only being part of the GSM neighboring lists seen by the RNC) to which handovers from a 3G cell are
supported by the network. Therefore, if the Compressed Mode is needed by the mobile for that frequency band, it
will be configured accordingly and possibly activated by the network.
False
?
CmodeDl(Ul)MethodSfdiv2 True
Physical Layer
The Spreading Factor Reduction method consists of the creation of gaps in transmission / reception and granting twice the
bandwidth for compressed frames in order to compensate for the loss of bandwidth in not transmitted frames. This method
applies for both uplink and downlink with fixed or flexible position mapping but it requires that the spreading factor be strictly
greater than 4.
The SF is reduced by a factor two for as many slots as used for gaps and the transmitted power of these slots is increased. Thus
the OVSF code needs to be changed. The new one is the parent code of the code used for non-compressed radio frames.
In downlink, the scrambling code management is done through the alternate scrambling code method. This method consists in
applying the new channelization code with SF/2 to the compressed frames, while applying one of the two available alternate
scrambling codes (left or right alternative) depending on the original OVSF.
The figure above gives an example of how this method applies.
In uplink, the compressed mode method by spreading factor reduction is identical to the spreading factor reduction used in the
downlink but with some exceptions.
HLS introduced from 3GPP R99:
Data rate is reduced from higher layers (i.e. MAC) by means of TFC restriction in the TFCS.
SF and scrambling code remain unchanged.
No additional power transmission to keep the same level of protection of the user bit.
Applicable either in UL only, DL only, or both UL/DL.
Prior to UA06, only SF/2 method was supported, whatever RAB, and CM method was uniquely defined using the
dlCModeMethod and ulCModeMethod parameters under the CModePatternSeqInfo object.
In UA06, HLS has been introduced and is supported for some specific UL and/or DL User Services. Therefore, the previous
parameters are not used anymore and are replaced by the isHlsCmMethodPreferred parameter defined per DlUserService
and UlUserService.
Parameters defined under CmodePatternSeqInfo are no longer used. However, they are still present in the RAN Model and
will be removed in the coming releases.
RadioAccessService NeighbouringRnc
isHlsCModeAllowed isHlsCmAllowedOnDrnc
CmodePatternSeqInfo
tgcfnOffset (CModePatternSeqInfo)
tgprc (CModePatternSeqInfo)
#TGCFN
#1 #2 #TGPRC
Sub-pattern 1Sub-pattern 2Sub-pattern 1Sub-pattern 2 Sub-pattern 1Sub-pattern 2
tgsn
tgd tgd
tgpl1 tgpl2
The Compressed Mode is controlled by the UTRAN: it is configured by the RNC on a per UE basis in the form of
Compressed Mode Transmission Gap Pattern Sequences. A CM pattern sequence may be composed of up to two
sub-patterns and is dedicated to one specific measurement purpose.
Each pattern is described by the parameters listed below, those parameters being defined at the cell level:
TGL1 and TGL2: length of transmission gaps 1 and 2 expressed as a number of slots. Possible values are 3, 4,
5, 7, 10 and 14. TGL2 is an optional parameter and if a value is not given by the upper layers, then by default
TGL2 = TGL1,
TGSN: the first gap occurs TGSN slots after the beginning of the pattern,
TGD: the two gaps are separated by a distance of TGD slots,
TGPL1 and TGPL2: length of pattern 1 and 2 expressed in radio frames,
TGCFN: CM start expressed in CFN as CFNx + TgcfnOffset) mod[256],
TGPRC:the number of times the Transmission Gap Pattern is repeated within the Transmission Gap Pattern
Sequence.
A certain number of pattern sequences can be defined in the UTRAN configuration, each of them being associated
with a specific measurement purpose.
The pattern sequence is defined by a set of parameters (transmission GAP and CM patterns parameters), that are
grouped into the CModePatternSeqInfo object:
0 of CmodePatternSeqInfo corresponds to a Compressed Mode measurement purpose GSM RSSI
Instance
Measurements.
1 of CmodePatternSeqInfo corresponds to a Compressed Mode measurement purpose GSM Initial
Instance
BSIC Identification.
2 of CmodePatternSeqInfo corresponds to a Compressed Mode measurement purpose FDD Inter-
Instance
Frequency Measurement.
For FDD inter-frequency measurement, a single pattern of 6 frames repeated 50 times is used, leading to a basic
compressed mode measurement period of 3 s.
The UE is provided with the FDD neighboring cell list, when receiving the RRC Measurement Control message. Using
this list, the UE starts the CPICH_RSCP and CPICH_Ec/No measurement processes that can be seen as a sort of
endless loop, intending to identify the best neighboring cells.
Measurement results are sent to the RNC with periodical measurement reports.
CFN + 48
CFN + 0
In the case of GSM initial BSIC identification, the UE is to take the results of the most recent set of GSM RSSI
samples and attempt to identify the BSICs of the 8 strongest cells, proceeding in single strength order.
It has to be noted that the time required for a measurement report is essentially dictated by the time required to
identify the BSICs of the required number of cells. As a consequence, it is better to choose the compressed mode
patterns for this operation first and then build the patterns for GSM RSSI measurements around this pattern.
Thats the reason why:
a transmission gap shorter that 14 has been chosen in order to allow good performance on BSIC identification.
8 patterns of 6x10ms have been allocated to RSSI measurements since the measurement period for the GSM
carrier RSSI measurement is 480 ms in the CELL_DCH state (as stated in [3GPP_R01]).
3x26 patterns have been allocated to initial BSIC identification in order to allow a minimum of 3 cells to be
reported in the worst case (e.g. when it takes up to nIdentifyAbort to identify each cell).
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
1 Paging 7
1.1 Paging DRX cycle 8
1.2 Paging repetition 9
2 Access preambles & acknowledgment 10
2.1 Preambles transmission 11
2.2 Acknowledgement transmission 12
2.3 Preambles retransmission parameters 13
3 RRC connection establishment 14
3.1 RRC connection setup 15
3.2 UL interference CAC on RACH 16
3.3 RRC connection rejection 17
3.4 RRC speech redirection 18
3.5 FACH power control as the type of DL traffic 22
3.5.1 FACH power adjustment 23
3.5.2 RRC connection setup repetition 24
4 RAB matching principles 25
4.1 RAB request vs. UserServices configuration 26
4.2 Matching main steps 27
5 RAB to RBset matching & TrCH type selection 28
5.1 Candidate RBset selection 29
5.2 Candidate RBset selection algorithm: speech 30
5.3
155
Candidate RBset selection algorithm: streaming 31
5.4
W-CDMACandidate RBset
R99 Algorithms Description selection algorithm: interact./backgr.
COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
Call Admission 32
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
5.5 TrCH type selection 33
6 iRM CAC: target RAB determination 34
6.1 iRM selection 35
6.2 DL IRM target RB selection algorithm 36
6.3 DL iRM on radio link condition 37
6.4 DL iRM on cell color 38
6.5 DL cell color calculation 39
6.6 DL cell color/active set color calculation 40
6.7 DL target RB determination 41
6.8 DL iRM CEM load parameters 42
6.9 DL iRM table: example for PS_384K_IB radio bearer 43
6.10 DL iRM: exercise 44
6.11 UL IRM target RB selection algorithm 47
6.12 UL radio load estimation without RSEPS 48
6.12.1 Self-learning of RTWPref 49
6.12.2 Calculation in Node B 50
6.12.3 Calculation in RNC 51
6.13 UL load estimation with RSEPS: calculation in RNC 52
6.14 UL IRM on cell color 53
6.15 UL cell color calculation 54
6.16 iRM target UL RB rate determination 55
6.17 UL iRM radio load parameters 56
6.18 UL iRM CEM load parameters 57
6.19 UL iRM table parameters 58
6.20 Exercise: iRM UL 59
6.21 iRM CAC for PS streaming RAB 62
7 iRM CAC: admission control & resource reservation 63
7.1 UL radio load token for high data rate 64
7.2 Transport resource reservation Equivalent Bit Rate (EBR) 66
7.3 DL reserved power computation 67
7.4 DL power admission criteria 68
7.5 DL power self tuning 69
7.5.1 Example 70
7.6 OVSF codes reservation & admission 71
8 CELL_FACH admission control 72
8.1 CELL_FACH admission control 73
The number of paging occasions per SFN cycle (4096 TTI) depends on the DRX cycle
coefficient.
Pa
gi
ng
fo
rP DRX
ac
ke FDDCell
tc
all
Pa
gin
g for csDrxCynLngCoef
Cir
c uit
Node B ca
ll
psDrxCynLngCoef
UE
When camping normally on a cell, the UE monitors regularly the paging channel. In order to save some energy, a
discontinuous reception mode (DRX) is used.
The DRX cycle is defined as the individual time interval between monitoring Paging Occasion for a specific UE. The
UE needs only to monitor one Page Indicator (PI) in one Paging Occasion per DRX cycle.
The DRX cycle length is defined as MAX(2k, PBP), where:
PBP is the Paging Block Periodicity and has the fixed value of 1 in UMTS-FDD.
k is an integer and can be specific by the Core Network domain.
The value of k is controlled in the Alcatel-Lucents solution by two parameters, one by the Core Network Domain:
csDrxCynLngCoef and psDrxCynLngCoef.
Since the UE may be attached to two different domains simultaneously, both DRX cycle length values are calculated
and stored in the UE from the values read in the SIB 1 (NAS system information, idle and connected mode timers
and counters). Then the UE should keep only the shortest of both values as the DRX cycle length it will use.
Core
RNC
Network
RANAP Paging (UEx)
RRC Paging Type 1 (UEx, other UEs)
Initial transmission
RRC Paging Type 1 (UEx, other UEs)
1st retransmission tpageVal
RRC Paging Type 1 (other UEs)
Tpaging
RRC Paging Type 1 (other UEs)
nth retransmission
X
RRC Paging Type 1 (UEx, other UEs)
Call Forwarded to Vmail
RN
nrOfPagingRepetition C
isPagingRepetitionAllowed RadioAccessService
(RNC)
PCH SCCPCH FDDCell tpageVal
In areas of poor radio coverage, it can happen that UE miss paging request what translates into the subscriber missing
terminating calls. In order to cope with radio transmission losses, the UTRAN can repeat the paging request so as to increase the
probability for the UE to hear it.
Paging repetition is applicable to mobile in idle, CELL_PCH or URA_PCH states.
Alcatel-Lucent implements two algorithms:
Paging Record priority: When several paging records have to be sent at the same paging occasion, the records are sent in
the same RRC PAGING message. Nevertheless, if the number of records exceeds the message size (i.e. more than 8
records), then the following priority will apply:
Priority 1: fresh paging record for packet or circuit services (i.e. no difference between the two domains). A fresh
paging record is a record which has not been previously sent.
Priority 2: repeated paging record for packet or circuit services (no difference between the two domains).
Limitation of repetitions: nrOfPagingRepetition is the number of times a paging record is repeated. This is a customer
configurable parameter.
nrOfPagingRepetition parameter: indicates the number of retransmissions of the paging by UTRAN. Specific value 0 means that
the paging will not be repeated (only the fresh paging is sent).
New UA07.1.2:
When searching for the next free Paging Occasion (PO) for a paging type 1, the RNC shall consider only POs within the time
given by the new parameter tpageVal. If no free PO is found within this time then the RNC shall discard the paging.
The RNC timer tpageVal should be aligned with the core network (MSC and SGSN) paging repetition timers.
Notes:
A PO is considered as free if the paging message has room left to add the new paging.
If the feature paging repetition is enabled and a PO contains RNC repeated pagings then these are removed if required to add
the 'fresh' paging. This is existing UA05.0 functionality from feature 24075.
If paging repetition is enabled then the RNC shall schedule repeated paging only within time tpageVal.
If the value of parameter tpageVal is changed then the RNC shall keep already scheduled pagings unchanged. The new value
applies for new pagings only.
no 3
detection RACH
RACH preambleThreshold
preambleThreshold Preamble
detection
2 rachsubChannel
AS 1
0 AS 1 AS 2 AS15
Ack.
PRACH PRACH
4
4 6
1
1 preambleSignature
prachScramblingCode
Preamble
aichTransmissionTiming
UE PRACH use is composed of two parts: the preamble part and the message part.
Before transmitting the message part of the preamble, the UE waits for an acknowledgement from the network (on
the AICH), confirming that the network has detected the UE.
The transmission of the preamble part consists of the repetition of a preamble composed of a 16-chip signature
repeated 256 times for a total of 4096 chips.
Basically, the UE is assigned one of the 16 possible preambles signature and transmits it at increasing power until it
gets a response from the network. The parameter preambleSignature of the RACH object, defines the set of allowed
signatures of the PRACH preamble part.
The preambleThreshold parameter is defined as the threshold (in dB) over the interference level used for preamble
detection in the CEM card. (The real value in dB of the preamble threshold is given by the formula: -
36+0.5*preambleThreshold.)
The rachSubChannels parameter defines the set of access slots on which the mobiles are authorized to transmit
their access on the PRACH. It defines a sub-set of the total set of uplink access slots.
The aichTransmissionTiming parameter of the RACH object defines the timing relation between PRACH and AICH
channels.
The scrambling code of the PRACH preamble part is defined by the prachScramblingCode parameter of the RACH
object.
Note: The RACH preambleSignature is limited to 1 signature for iCEM128. The allowed signature will be
0000000000000001 or 1000000000000000 or 0000000100000000
aichTransmissionTiming = 1
5 TS
Downlink AICH
AICH
The aichTransmissionTiming parameter of the RACH object defines the timing relation between PRACH and AICH
channels.
For example, when aichTransmissionTiming is set to 1:
The minimum inter-preamble distance tp-p,min = 20480 chips (4 access slots)
The preamble-to-AI distance tpa = 12800 chips (5 time slots)
The preamble-to-message distance tp-m = 20480 chips (4 access slots)
Note: only aichTransmissionTiming = 1 is supported for iBTS (Global Market) and aichTransmissionTiming =
0 is supported for OneBTS (US Market).
PREAMBLE #1
Access Cycle #1
preambleRetransMax (RACH)
PREAMBLE #2
NB01min (RachTxParameters)
Access Cycle #2
PREAMBLE #1
NB01max (RachTxParameters)
PREAMBLE #N
When a negative answer is received by the UE from the network after a given period, the UE re-sends a preamble at
a higher transmission power, so that the Node B can detect it better among the other information received. This
ramping up process is thus characterized by:
of the preamble retransmission: 3GPP (cf. 25.321) has defined two parameters: NB01min and
Periodicity
NB01max, setting respectively the lower and the upper bounds of the retransmission periodicity (unit is
expressed in tens of ms).
Maximum number of preambles transmitted: this limitation is defined through preambleRetransMax and Mmax
parameters.
preambleRetransMax gives the maximum number of PRACH time slots allowed within an access cycle.
Mmax gives the maximum number of access cycles. An access cycle is defined by a number of radio
frames on which the PRACH access (and therefore a preamble ramping cycle) is allowed on specific slot
numbers.
The ramping process stops when the number of preambles transmitted has reached the maximum allowed number
of PRACH retransmissions, either within an access cycle, or when the maximum number of access cycles is reached.
SRB = ???
RadioAccessService
UserServices
When the UE attempt to establish an RRC Connection is accepted, the corresponding Signaling Radio Bearers can be
supported on two different RRC states and with two different throughputs:
CELL_FACH
CELL_DCH
The parameters which allow selection of the RRC state to support the Signaling Radio Bearers are UlUserviceId for
the UL direction, and DlUserserviceId for the DL direction.
The selection of the SRB xxServiceId to accommodate the RRC connection is distinguished by RRC establishment
Cause (UserServices instance):
IMSI Detach, Registration, Originating Low Priority Signaling, Originating Low Priority Signaling: SRB_CellFACH
Emergency Call: SRB_3_4K_DCH
Others (normal Originating and Terminating calls): SRB_13_6K_DCH
Eb/No
required Call is accepted Call is rejected
Interference
level
Yes No
UL RTWP
Received Power
RTWP < Maximum UL Interference
Level
RNC
H
P- RAC
The overall interference level received in a cell is measured with the UL RTWP measurement (Received Total
Wideband Power measured at the Node B and forwarded to the RNC).
On RACH reception, before the allocation of the standalone signaling radio bearer, and during the resource
reservation phase, the RNC compares the measured RTWP with a fixed value, the maxULInterferenceLevel
parameter.
If the RTWP is below this threshold, the criterion is met.
If the RTWP is over the threshold, the call is rejected.
The RTWP is measured by the Node B and sent towards the RNC by sending a NBAP common measurement
report.
RNC
If RRC connection fails, the UE will re-attempt a 3G call establishment up to N300 times. However, the UE is
required to wait (at least) a predetermined time before the subsequent attempt on the 3G network. This wait time is
sent by the RNC to the UE in the Wait Time IE in the RRC Connection Reject message.
Subsequent call attempts may or may not be redirected to the 2G network, depending on whether the initial cause
for RRC Redirection still persists on the 3G UTRAN.
The Wait Time parameter will be set to the value associated with one of the following parameters:
timeReject. If the admission failure which causes the redirection is RNC overload
waitTimeOnRrcConnectionRejection. If the admission failure which causes the redirection is congestion
waitTime3Gto2GRedirectFailure. In the case of a 3G-2G Emergency Redirection or 3G-2G Redirection based
on cell load
The waitTimeOnRrcConnectionRejection parameter is in seconds and the value 0 indicates that the repetition
is not allowed.
Notes:
SRB CAC: The RNC will check that the resources required to support the SRB (e.g. power and codes) are
available.
Max UE contexts: The RNC checks that the maximum number of concurrent UE contexts will not be exceeded
IF
RRC Establishment Cause = MO Conversational AND
is3Gto2GConversationalCallRedirectOnRrcEstabFailAllowed =
TRUE
OR
RRC Establishment Cause = Emergency AND
is3Gto2GRedirectForEmergencyAllowed = TRUE
AND
2G neighbor configured
THEN
include Redirection IE in RRC Connection Rejection message
RRC Connection Rejected (cause)
Redirection IE (Inter-RAT info = GSM)
is3Gto2GConversationalCallRedirectOnRrcEstabFailAllowed (RadioAccessService)
WaitTime3GTo2GRedirectFailure
(FDDCell) is3Gto2GRedirectForEmergencyAllowed (FDDCell)
Upon reception of the RRC Connection Request message, the RNC executes the usual RRC Connection Admission
Controls.
If failure occurs for SRB assignment, the RNC verifies that some pre-conditions for redirection to GSM are fulfilled.
Then the RRC Connection Reject message contains the Redirection IE with Inter-RAT info set to GSM. Note that
the RNC is unaware of the UE capabilities at RRC Connection Request time. Therefore, the RNC attempts an RRC
Redirection independently of whether the UE supports GSM or the Redirection IE. If the UE supports GSM and the
Redirection IE, it will perform inter-system cell reselection and will re-originate the speech call on the 2G network.
All types of MO Conversational calls are redirected to 2G upon admission failure independently of the service type or
domain. This includes non-speech calls such as Video Telephony. This is a consequence of the fact that the RRC
establishment cause is not able to uniquely identify a CS speech at this early stage of the call progression.
Redirection is not triggered if the UE already has an established RRC connection prior to invoking the MO call
request (for example when a PS call is already established).
(FDDCell)
FDD Originating Cell Load
is3G2GRedirectOnCellLoadAllowedForEmergency(
RadioAccessService)
Code Color
UL RTWP Load
Color
Worst
UL Cell Color
redirection3G2GOnCellLoadThreshold
CEM Color
(FDDCell)
From UA07 onwards, this feature enables 3G/2G RRC redirection of CS Speech calls when the cell load on the
originating UMTS cell reaches a configurable threshold.
Mobile then selects a GSM cell based on previously measured neighbouring cell list and retries a call establishment.
Compared to Load based Handover, this procedure does not consume any 3G resources.
IF
RRC Establishment Cause = MO Conversational AND
is3G2GRedirectOnCellLoadAllowed = TRUE)
OR
RRC Establishment Cause = Emergency AND
is3G2GRedirectOnCellLoadAllowed = TRUE AND
is3G2GRedirectOnCellLoadAllowedForEmergency =
TRUE
AND
other conditions are fulfilled
THEN
RRC/FACH (RRC Connection Reject)
include Redirection IE in RRC Connection Rejection
Wait time
message
Inter-RAT info=GSM
GSM Target cell info list (1 to 32 Cells)
BCCH ARFCN redirection3G2GOnCellLoadThreshold
Band Indicator (FDDCell)
Rejection Cause: congestion
is3G2GRedirectOnCellLoadAllowedForEmergenc
WaitTime3GTo2GRedirectFailure
y
(FDDCell)
(RadioAccessService)
1 5 20 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Admission
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
The RRC Connection Request initiated by the UE contains the establishment cause.
The other conditions that must be fulfilled are explained in the next slide.
If the call is eligible to the 3G-2G redirection criterion, an RRC Connection Reject is sent to the UE with redirection
info included and may include the GSM target cell info list IE.
The RNC always sends the GSM target cell Info List whatever UE release (R99, R5, R6, R7).
For R5 & R99 Mobiles, no differentiation in the RRC Connection IE between video & voice thus: risk to
move video calls in 2G is to be considered.
Upon receiving the RRC Connection Reject message from UTRAN, the UE will process the GSM cell selection process
using or not the GSM target cell info and will attempt an RACH on 2G if it finds an eligible GSM target cell.
If no GSM cell fulfills the selection criteria, the UE will re-attempt a new RACH towards the UTRAN after the wait
time timer (waitTime3Gto2GRedirectFailure) has elapsed. The UE may camp on the same Fdd cell or another
Fdd cell (the cell reselection process may change the Fdd cell).
When the re-attempt occurs in the same FDDCell within a certain period of time (RrcCnxRepeatTime), the RNC
doesnt redirect the call to the 2g and attempts to establish the call on the FDDCell thanks to a mechanism (already
used for all features dealing with RRC Redirection) at cell level which registers the UE identity before launching the
3G2G redirection. If the re-attempt occurs after the timer elapses or in a different FDDCell, the RACH is managed as
a first RRC establishment request.
2G load is not taken into account to take the decision to trigger the redirection.
isFachPowerDifferentForSrbtraffic
= True
(RadioAccessService)
UE in Cell_FACH
A) FACH
D AT
ck s ( PWR
MA C blo fachTrbPowerOffset
RLC/ (FACH)
PCH
/ S-CC
F AC H
RLC/MAC blocks (SIGNALING)
LING
) FACH
s (SIGNA
c k
C blo
fachSrbPowerOffset
MA PWR
RLC/ (FACH)
PCH
/ S-CC
F AC H
The feature differentiates the power level used for FACH channel depending on whether data or signaling is
transmitted.
From UA07, a Boolean switch allows choosing between the two alternative approaches.
It is possible for operator to choose one of the following options:
Power ramping for RRC Connection Setup message based on UE feedback and fixed (same) power for the rest
of signaling and traffic on FACH (UA05 onwards mechanism called FACH Power Adjustment and Quick Repeat).
Configurable fixed (different) power level based on whether the FACH frame contains Signaling Radio Bearer
data or Traffic Radio Bearer data (new option from UA07).
isFACHpowerdifferentForSRBtraffic allows one to enable/disable the use of the configured fixed power offset
based on whether the Radio Bearer is Signalling or Traffic.
If isFACHpowerdifferentForSRBtraffic is TRUE, regardless of whether isFachPowerAdjustmentActivated
(flag to activate FACH power adjustment) is turned on or off, the configured FACH power offset values based for
SRB fachSrbPowerOffset or TRB fachTrbPowerOffset are used.
Maximum FACHpower
AbsoluteMaxFachPower
FachTransmitPowerLevelStep
FachTransmitPowerLevelStep
InitialFachPowerAdjustment
sccpchPowerRelativetoPcpich
third RRC
PCPICHPower
second Connectio
first RRC RRC n Setup
all other Connection Connectio
P-CPICH Setup n Setup
FACH
messages
sccpchPowerRelativeToPcpich (SCCPCH)
It is proposed to adjust the FACH power while sending the RRC Connection Setup message based on the CPICH
Ec/No measurement received from the RRC Connection Request message. The preferred power setting change is
only applied to the FACH frames which carry the RRC Connection Setup message. For other messages, the RNC
should set the power setting level to the nominal value.
Once the FACH power adjustment is required for the first RRC Connection Setup, at every next subsequent
repetitions (that is, T351 expiration), the FACH power is further stepped up.
The feature is activated both at the RNC (isFachPowerAdjustmentEnabled) and FddCell levels
(isFachPowerAdjustmentActivated).
If the quality measurements of either Ec/No (by default) or RSCP is below the threshold, the FACH power
adjustment will be performed.
UE RNC-IN
RNC-
CN Control of RRC Connection Setup
RRC Connection Request (TM) (CPICH Ec/No) t300
isQuickPepeatActivated (FDDCell)
isQuickPepeatAllowed
RRC Connection Setup (UM) RRC Connection
numberOfQuickRepeat
Setup
deltaCpichEcNoUsedQuickRepeat
The RRC Connection Setup message is sent over CCCH/FACH in RLC UM mode. Without its retransmission, the
message could be lost over the air due to bad RF conditions. The objective of this feature is to provide the RRC
Connection Setup message retransmission functionality if the RRC Connection Setup Complete message from the UE
has not been received within the duration of T351 timer.
The retransmission of the RRC Connection Setup message based on a quicker timer T351 than T300 reduces the call
setup duration. By reducing the need for the UE to submit another RRC Connection Request message as a result of
the expiry of timer T300, this feature has a positive impact to the RACH capacity.
This feature provides quick repetition functionality of the RRC Connection Setup message without waiting for the
acknowledgement from the UE (RRC Connection Setup Complete message).
The quick repetition of the RRC Connection Setup is activated based on the P-CPICH Ec/No measurement received
from the RRC Connection Request message reported by the UE. If quality measurements are below a certain
threshold, that is; CPICH EC/No < fachPowerAdjustmentCpichEcNoThreshold +
deltaCpichEcNoUsedQuickRepeat. The likelihood of high BLER on the FACH channel is increased, thus reducing
the probability of RRC Connection Setup being successfully received by the UE, since it is sent on RLC UM. In order
to increase the probability of successful RRC Connection Setup transmission, the message is quick-repeated, that is
to say, without waiting for an acknowledgement.
T352 is the Alcatel-Lucent Timer to control the release of UE call context by the RNC.
ice
Serv est
u RNC Core Network
Req
Service
Request
RadioAccessService
UserServices
The objective of the RAB matching algorithm is to translate the RAB parameters specified in the RAB ASSIGNMENT
REQUEST received from the Core Network into a pre-defined RAB supported in the RNC.
The requested RAB is matched to the closest RAB provisioned in the RNC, using the RAB matching algorithm.
UE Capability Check
RAB Matching is done at call establishment. For soft handover, only resource reservation and Call Access Control are
performed.
The above diagram describes the main steps of the RAB Matching algorithm used.
Step 1: RAB to RBset Matching:
UL & DL RBs are selected according to the RAB Request and stored in an RB set.
this RB list is filtered according to UE capabilities.
Step 2: Transport Channel Type Selection:
DCH or HDSCH in DL, DCH or E-DCH in UL is selected according to the UE and cell capabilities
Step 3: iRM RAB to RB Mapping (DL only):
a DL Target RB is elected among all the selected RBs of the RBset.
Step 4: RBset to User Services Matching:
Target User Services are extracted and explicitly defined from the RBsets.
Reference User Services are extracted and explicitly defined from the RBsets.
enabledForRABMatching (DlRbSetConf)
enabledForRABMatching (UlRbSetConf)
DL Reference RB UL Reference RB
The Purpose of this algorithm is to get as output a radio bearer table containing all the acceptable Radio Bearers (DL
Candidate RBset and UL Candidate RBset), among which one is marked as the Reference RB in UL & DL. These
RBsets also include the RBs to be used for Always On and iRM Scheduling (when applicable).
From all Radio Bearers defined in the RNC, the RNC selects the Radio Bearers (DlRbSetConf and UlRbSetConf)
having the following properties:
It is eligible for RbSet Matching (enabledForRabMatching).
The Traffic Class corresponds to the requested Traffic Class.
The Bit Rate is compliant with the RBset selection criteria (see next slide).
For PS Calls, the rule is to sort all eligible radio bearers in decreasing bit rate order, and to select the reference radio
bearer as being the first element in the top of the list.
Other radio bearers are kept as fallback radio bearers.
From UA06.0, RAB negotiation may be supported at establishment time. If the enableRabNegotiation flag is set
to True, the presence of the Alternative RAB Parameter Values IE is checked in the RANAP RAB ASSIGNMENT
[3GPP_R18] or RELOCATION REQUEST message and if present with either Alternative MBR or Alternative GBR then
negotiating the Maximum Bit Rate or Guaranteed Bit Rate (Streaming class) respectively is allowed.
RAB matching and call admission is performed as normal and if the requested rate is not achievable a lower data
rate may be selected. This is applicable to both Interactive/Background and Streaming RABs.
Note: The enableRabNegotiation activation flag shall be set to True only if the SGSNs in the CN support the RAB
negotiation facility with the Unspecified Type of Alternative Bit Rate.
DL Candidate UL Candidate
RbSet RbSet
The allocation of bearer for voice call depends if the multi-mode AMR is activated at RNC level:
If Traffic Class = Conversational and Source = Speech (Speech case)
Bit Rate (RbSetConf) = MaxBitRate (rabParam)
Bit Rate (RbSetConf) Guaranteed Bit Rate (rabParam)
The RNC shall determine the speech bearer according to the AMR activated modes:
CS_AMR_WB: CS_AMR_NB: CS_AMR_LR
o {12.65k, 8.85k, 6.6k} o {12.2k, 7.95k, 5.9k, 4.75k} o {5.9k, 4.75k}
o {12.2k, 7.4k, 5.9k, 4.75k} o {4.75k}
DL Candidate UL Candidate
RbSet RbSet
DCH HSDPA DCH HSUPA
PS_16K_STR PS_HSDSCH_STR PS_16K_STR PS_EDCH_STR
PS_64K_STR PS_32K_STR
PS_128K_STR PS_64K_STR
PS_256K_STR PS_128K_STR
PS_384K_STR
The following table shows an example of the Streaming Radio Bearer Table.
Note: Streaming over EDCH is an optional feature/service supported from release UA07.1.2.
DL Candidate
RbSet
Candidate RBset Selection
Example
Candidate RBset Selection
DL Candidate RbSet
CS MaxBitRate = 12.2k DCH
PS MaxBitRate DL = 2048K UE Capability Check* UL Candidate RbSet PS 384K (Ref.)
PS MaxBitRate UL = 384K PS 256K
CS Speech + PS I/B
DCH
PS 384K PS 128K
A/R Priority Level = 2 PS 64K
... PS 128K (Ref.)
PS 64K PS 32K
PS 32K PS 16K
* If isUeCapabilitiesInRabMatchingAllowed = True PS 16K PS 8K
PS 8K CS 12.2k (Ref.)
(RadioAccessService)
CS 12.2k (Ref.)
From all Radio Bearers defined in the RNC, the RNC selects the Radio Bearers (DlRbSetConf and UlRbSetConf)
having the following properties:
It is eligible to RbSet Matching (EnabledForRabMatching parameter)
The Traffic Class corresponds to the requested Traffic Class and:
If Traffic Class = Conversational and Source = Speech (Speech case)
Bit Rate (RbSetConf) = MaxBitRate (rabParam)
Bit Rate (RbSetConf) Guaranteed Bit Rate (rabParam)
If Traffic Class = Streaming
Bit Rate (RbSetConf) Guaranteed Bit Rate (rabParam)
If Traffic Class = Interactive or Background
Bit Rate (RbSetConf) MaxBitRate (rabParam)
The Candidate RBset Selection produces two Radio Bearers lists (one list for UL and one list for DL) that are further
filtered according to UE capability.
HS-DSCH HS-DSCH
STR DCH STR STR
R6 DCH E-DCH
HS-DSCH HS-DSCH
I/B DCH I/B I/B
DCH E-DCH
HSxPA,
FACH.
Note: Streaming over EDCH is an optional feature/service supported starting release UA07.1.2
UL iRM DL iRM
iRM Tables
According to the cell load (DL and UL) and radio conditions of the UE (DL only), from a Reference RB bit rate
deduced from CN QoS requirements, the RNC will determine, in UL and in DL, a Target RB bit rate in order to avoid
congestion in the cell.
iRM UL and iRM DL Tables are used for Target RB determination according to some criteria (see next slides).
Besides, RB adaptation based on traffic is a feature introducing PS I/B RB bit rate downsizing/upsizing based on user
estimated average throughput.
DL and UL rate adaptation are performed independently.
In UL and/or DL, an initial RB Rate Adaptation can be performed at RAB establishment to admit a user at a
configurable low bit rate.
Consequently, the allocated UL PS RB bit rate and/or UL PS RB bit rate is limited at RAB Establishment, even if the
user is requesting more.
Once the RAB is established, it may be possible to upsize the RB to UL PS 384 kbps if needed thanks to RB
Adaptation:
"Max UL RB bit rate" ("Max DL RB bit rate") specifies the maximum UL rate (DL rate), which may be allocated
at service establishment time (RANAP RAB Assignment Request) or after relocation (RANAP Relocation
Request).
This parameter is significant when isUlRbRateAdaptationAllowed = True (isDlRbRateAdaptationAllowed =
True).
It depends on the activation of the "Initial Rate Capping during RB reconfiguration" feature:
2 DL RL Quality
DL Cell Color
olo r
Link C
Radio
Gold
Bronze Silver
(Ref.) (Ref.)
1 (Target)
5 DL IRM Tables
No
isIrmOnRlConditionAllowed
(RadioAccessService) ?
Good RL
Yes Condition
If
- CPICH_Ec/No < irmDlPowerThreshold (IrmOnRlConditionParameters) dlRbSetConfId
And (IrmOnRlConditionParameters)
Then Else
fallbackRbRate (DlRbSetConf)
Good RL Bad RL
Condition Condition
iRM Target RB selection shall be limited to fallBackRbRate in case of bad UE radio conditions in order to reduce RLC
re-transmissions and guarantee a minimum level of throughput.
Radio Link conditions are assessed from RRC measurements reported by UE.
Link color calculation is based on the following algorithm:
if (-CPICH Ec/N0 primary < IRMDlPowerThreshold) and (CPICH RSCP > IRMDlCoverageThreshold)
then link color = GREEN
else link color = RED
DL iRM Target RB bit rate shall be limited to fallBackRbRate if the radio link color is RED, otherwise no limitation is
requested.
Note:
During transition from cell FACH state to Cell DCH state, CPICH RSCP is not reported by the UE, therefore:
the coverage criterion (IrmDlCoverageThreshold) is ignored.
theRadio link color evaluation is then based only on CPICH Ec/No measurements as reported on the last RACH
message.
FDDCell Color
DL Cell Color
CEM load is not only used in the iRM CAC algorithm. Therefore if CEM load criterion is not to be used in iRM CAC
although CEM load is being computed for iMCTA feature, then:
The isCEMColourCalculationEnabled parameter has to be set to TRUE
The isCEMModelValidForDlColour parameter has to be set to FALSE
In this case, the CEM Color used in iRM CAC will be equal to the dlCEMColourDefaultValue parameter
value.
Copyright 2012 Alcatel-Lucent. All Rights Reserved.
TMO18255_V2.0-SG-UA08-Ed1 Module 1.5 Edition 1
Section 1 Module 5 Page 38
6 iRM CAC: target RAB determination
6.5 DL cell color calculation
Code Load
yellow2RedCLCThreshold 70 % red2YellowCLCThreshold
(IrmOnCellColourParameters) (IrmOnCellColourParameters)
60 %
50 %
green2YellowCLCThreshold yellow2GreenCLCThreshold
(IrmOnCellColourParameters) 40 % (IrmOnCellColourParameters)
Power Load
yellow2RedPLCThreshold 70 % red2YellowPLCThreshold
(IrmOnCellColourParameters) (IrmOnCellColourParameters)
60 %
50 %
green2YellowPLCThreshold yellow2GreenPLCThreshold
(IrmOnCellColourParameters) 40 % (IrmOnCellColourParameters)
Iub Load
Worst DL Cell Color
yellow2RedDlTLCThreshold 90 % red2YellowDlTLCThreshold
(IrmIubTransportLoadParameters) (IrmIubTransportLoadParameters)
80 %
70 %
green2YellowDlTLCThreshold yellow2GreenDlTLCThreshold
(IrmIubTransportLoadParameters) 60 % (IrmIubTransportLoadParameters)
CEM Load
yellow2RedDlCEMThreshold 90 % red2YellowDlCEMThreshold
(DlIrmCEMParameters) (DlIrmCEMParameters)
80 %
70 %
green2YellowDlCEMThreshold yellow2GreenDlCEMThreshold
(DlIrmCEMParameters) 60 % (DlIrmCEMParameters)
1 5 39 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Admission
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
NOTE: The values provided here for the different Power load, Code load, Iub load and CEM load are just examples.
They are neither Alcatel-Lucent default values nor recommended values as those ones are driven by the
configuration of Node B and cell and by the operator strategy as a trade-off between capacity (number of
simultaneous users) and quality (throughput for PS service).
Indeed:
Codeload thresholds setting is driven by the code capacity of the cell for DCH traffic which depends on the
number of S-CCPCH channels configured, on the fact that the cell might also carry HSDPA traffic, and in that
case on the Dynamic Code Tree Management feature activation.
Powerload thresholds setting is driven by the power capacity of the cell for DCH traffic which depends on the
usage of OCNS, and on the power reserved for HSDPA.
Iub load thresholds setting is driven by the Iub bandwidth capacity of the BTS which depends on the number of
E1 links equipped, the IMA activation and the CAC method used.
CEM load thresholds setting is driven by the CEM capacity of the BTS which depends on the type and number
of CEM boards equipped, on the number of Local Cell Group configured and on the DBBU Frequency Pooling
activation (dbbuPoolMode parameter).
CEM color in DL is calculated by the iRM mechanism comparing the DL CEM load estimation (expressed in percent)
with the different CEM DL load thresholds configured at OAM
Once computed, the CEM color is applied to all the cells of a BTS, cells belonging to the same Local Cell Group.
Worst
Worst + =
+ =
Code Color
Worst Iub Color + =
Green Yellow
Red
Cell N
Active Set Color
Green Yellow
Red
Cell Color
SRNC
PS RAB Assignment Request
Core Reference RB Bit Rate
Traffic class
DL Maximum Bit Rate
Allocation/Retention Priority Level
olo r
Link C
Radio fallBackRbRate MIN Target RB Bit Rate
+ Gold
iRMRbRate
Bronze Silver
DL Cell Color
OLS
1 5 41 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Admission
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
The RAB to RB mapping function consists in defining the target RB Set that will replace the Reference RB.
For each triplet {DlRbSetConf, OLS, CellColor}, an iRMRbRate parameter is defined in a DL iRM Table:
DlRbSetConf
PS_384K_IB
PS_384K_IB_MUX
PS_256K_IB DlIrmTable
PS_ 256K_STR OLS Cell Colour Cell Colour Cell Colour
PS_128k_IB = Green = Yellow = Red
PS_128k_IB_MUX Gold 384 128 64
PS_128K_STR Silver 384 128 64
Bronze 384 128 64
NeighbouringRnc
defaultDlIrmCellColour
NodeB RadioAccessService
isCEMColourCalculationEnabled
1..15
nodeBConfClassId
3
fallBackRbRate
IrmRbRateList
1..15
NodeBConfClass 3
isCEMModelValidForDlColour
IrmRbRateEntry
dlCEMColourDefaultValue
iRMRbRate
DlIrmCEMParameters
green2YellowDlCEMThreshold
yellow2RedDlCEMThreshold
red2YellowDlCEMThreshold
yellow2GreenDlCEMThreshold
DlRbSetConf
PS_384K_IB
PS_384K_IB_MUX
PS_256K_IB DlIrmTable
PS_ 256K_STR OLS Cell Color Cell Color Cell Color
PS_128k_IB = Green = Yellow = Red
PS_128k_IB_MUX Gold 384 128 64
PS_128K_STR Silver 384 128 64
Bronz
384 128 64
e
1 5 45 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Admission
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
SRNC
PS RAB Assignment Request
Core Traffic class = Background
DL Maximum Bit Rate = 2048000
UL Maximum Bit Rate = 128000
Allocation/Retention Priority Level = 1
Iub
NodeB
io Li n k
Rad
Load
Radio
4 Olympic Service Level
UL Candidate
UL Candidate RbSet RbSet
(Ref.) (Ref.)
1 (Target)
5 UL IRM Tables
The aim of UL iRM CAC is to provide the operator means to manage efficiently I/B and streaming RABs on R99
resources as a function of:
traffic
conditions (through radio load color evaluated by the RNC thanks to Noise Rise estimated by the NodeB
and reported to the RNC)
CEM load (through CEM color)
OLS (Olympic Service Level is either Gold or Silver or Bronze arrording to the Allocation/Retention Priority IE
provided in the RAB ASSIGNMENT REQUEST message).
Further reconfiguration can be triggered either by DL related events or by the UL RAB adapt feature.
This feature provides the operator with the best trade-off in terms of offered QoS and NodeB/Cell available
resources.
This minimizes blocking and call drops.
18
E-DCH load 16
14
RoTma 10
x
8
PS64 RoTnon E-DCH
6
RTWPnon E-DCH
CS12.2 4
2
CS12.2
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
NodeB NF UL load (%)
RTWPref
Thermal Noise
Acceptable Max load
The way to control the Uplink traffic QoS is to maintain the UL load under a fixed level.
The current absolute UL RTWP (i.e. Received Total Wideband Power) as defined in the 3GPP cannot be measured
with enough accuracy (+/- 4 dB). Indeed, it depends on the temperature and site conditions. It is therefore varying
in time.
Due to these constraints, UL load cannot be controlled based on direct UL RTWP measurements => Needs for
enhanced estimation.
Therefore in order to improve the accuracy of the R99 UL CAC algorithm, the Node B provides the RNC with the
Rise over the varying Thermal noise (RoT) corresponding to the Noise Rise induced by the UL R99 traffic.
As the measurement provided by the Node B in the Common Measurement Report should be the RTWP expressed in
dBm, the Node B adds a fixed reference value equal to 106.1dBm to the actual RoT.
The UL load should be monitored in order not to overload the system. It should be kept lower than a fixed threshold
to keep the system stable.
The UL load estimation is required for correct E-DCH scheduling and efficient UL iRM CAC.
The thermal noise should be well estimated in order to compute the UL load.
Reference_RTWP
Yes
rtwpReference
(BTSCell)
RTWPref,0 = -106.1dBm
RTWP when no traffic
+++ +
+ ++ + +
++
++
++ +
+ + +
+++
+
+ +
+ +
+ + + +++
++ + +
++
+ +++
+
+ +++ + +
+ +
+ + +++
++ +
+ +++++
+
+
+ ++ + + +
+
+
++ ++++
+++ + +
+
+ ++
+
-105.4dBm ++
time
Day n-1 Day n Day n+1
The RTWP reference value (called RTWPref) should correspond to the minimum value of RTWP values received in
the cell when there are no connections in the Node B.
During the learning time (24hours), the Node B keeps the RTWPcur values measured (filtered by L3 filtering-
param sent by the RNC) if no traffic in the Node B.
Please note that the first learning cycle is faster than 24 hours.
The feature Self learning of RTWP ref is used in the two cases: UL Radio Load Estimation Without RSEPS & UL Radio
Load Estimation With RSEPS.
CS64
RTWPnon E-DCH = RTWPcur RTWPE-DCH
RTWPcur
PS64 RoTnon E-DCH RoTnon E-DCH = RTWPnon E-DCH - RTWPref
RTWPnon E-DCH
CS12.2
CS12.2
NodeB NF
RTWPref
Thermal Noise
The non E-DCH load is obtained by substracting this computed E-DCH load from the total RTWP.
For each E-DCH connection, the SIR will be estimated in function of the SIR on UL DPCCH and the gain factors.
These SIR are cumulated and then the contribution of E-DCH to the current total RTWP is estimated.
Note that the E-DCH traffic that belongs to other cells is included in the non E-DCH RTWP measurement.
The total RTWPcur is the average between the two RX diversity branches if any.
CRNC
nbapCommonMeasRtwpReportingPeriod
(NbapMeasRtwpParameters)
UL Radio Load = 1 10-(RoTnon E-DCH/10)
nbapCommonMeasRtwpFilterCoeff
(NbapMeasRtwpParameters)
C-RNC
Common Measurement Initiation Request (Total RTWP, Reference RTWP)
With this measurement : no more need to adjust the RTWP Measurement at BTS to
report the Non-Edch RoT
The RNC computes the Non-Edch load for iRM UL load
WARNING: if RSEPS are activated, the #303 UL_RSSI counter gives the actual total RTWP whereas in UA05 it
corresponded to -106.1dBm + RoT_non_Edch.
No No
Yes
isUlIrmOnCellColourAllowed
(RadioAccessService) ? isCEMColourCalculationEnabled
(RadioAccessService) ?
Yes Yes
isUplinkRadioLoadEnabled
(RadioAccessService) CEM Color
isUlRadioLoadColourEnabled
(NodeBConfClass)
Yes
Radio Load
? UL Cell Color
Color
No
UL IRM Tables
As any other, the CEM load criteria can be used or not thanks to the isCEMColourCalculationEnabled parameter.
But CEM load is not only used in iRM CAC algorithm. Therefore if the CEM load criterion is not to be used in iRM CAC
although CEM load is being computed for iMCTA feature, then:
The isCEMColourCalculationEnabled parameter has to be set to TRUE .
The isCEMModelValidForUlColour parameter has to be set to FALSE.
In this case, the CEM Color used in iRM CAC will be equal to ulCEMColourDefaultValue parameter value.
Radio Load
yellow2RedUlRadioLoadThreshold 70 % red2yYellowUlRadioLoadThreshold
(UlIrmRadioLoadParameters) (UlIrmRadioLoadParameters)
60 %
50 %
green2yYellowUlRadioLoadThreshold yellow2GreenUlRadioLoadThreshold
(UlIrmRadioLoadParameters) 40 % (UlIrmRadioLoadParameters)
CEM Load
yellow2RedUlCEMThreshold 90 % red2yYellowUlCEMThreshold
(UlIrmCEMParameters) (UlIrmCEMParameters)
80 %
70 %
green2yYellowUlCEMThreshold yellow2GreenUlCEMThreshold
(UlIrmCEMParameters) 60 % (UlIrmCEMParameters)
CEM color in UL is calculated by the iRM mechanism comparing the UL CEM load estimation (expressed in percent)
with the different CEM UL load thresholds configured at OAM
Once computed, the CEM color is applied to all the cells of a BTS, cells belonging to the same Local Cell Group.
NOTE: The values provided here for the different Radio load and CEM load are just examples. They are neither
Alcatel-Lucent default values nor recommended values as those ones are driven by the configuration of NodeB and
cell and by the operator strategy as a trade-off between capacity (number of simultaneous users) and quality
(throughput for PS service).
Indeed:
Radio load thresholds setting is driven by the code capacity of the cell for DCH traffic which depends on the
fact that the cell might also carry HSUPA traffic. Attention should be paid to the fact that
yellow2RedUlRadioLoadThreshold should be lower or equal to the UL CAC threshold
rtwpMaxCellLoadNonEdch.
CEM load thresholds setting is driven by the CEM capacity of the BTS which depends on the type and number
of CEM boards equipped, on the number of Local Cell Group configured and on the DBBU Frequency Pooling
activation (dbbuPoolMode parameter).
SRNC
PS RAB Assignment Request
Core Reference RB Bit Rate
Traffic class
UL Maximum Bit Rate
Allocation/Retention Priority Level
+ Gold
iRMRbRate
Bronze Silver
Cell Color
OLS
1 5 55 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Admission
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
NeighbouringRnc
defaultUlIrmCellColour
BTSEquipment NodeB
isRtwpReferenceSelfLearning isNbapCommonMeasRsepsAllowed
isRtwpAdjustmentForRnc
RadioAccessService
nodeBConfClassId isUplinkRadioLoadEnabled
cacConfClassId
UlIrmRadioLoadParameters NbapMeasRtwpParameters
green2yellowUlRadioLoadThreshold nbapCommonMeasRtwpReportingPeriod
yellow2redUlRadioLoadThreshold nbapCommonMeasRtwpFilterCoeff
red2yellowUlRadioLoadThreshold
yellow2greenUlRadioLoadThreshold
RadioAccessService
isCEMColourCalculationEnabled
NodeB
DedicatedConf
nodeBConfClassId
1..15
NodeBConfClass
isCEMModelValidForUlColour
ulCEMColourDefaultValue
UlIrmCEMParameters
green2YellowUlCEMThreshold
yellow2RedUlCEMThreshold
red2YellowUlCEMThreshold
yellow2GreenUlCEMThreshold
RadioAccessService
1..15
Example
3 UlRbSetConfId = PS_384K_IB
IrmRbRateList IrmRbRateEntry
IrmRbRateList Instance
Instance iRMRbRate
3
0 OLS = Gold 384
/0 Cell Color = Green 1 OLS = Silver 384
IrmRbRateEntry
2 OLS = Bronze 384
0 OLS = Gold 128
iRMRbRate /1 Cell Color = Yellow 1 OLS = Silver 128
2 OLS = Bronze 128
0 OLS = Gold 64
/2 Cell Color = Red 1 OLS = Silver 64
2 OLS = Bronze 64
Assumptions
UlRbSetConf
PS_384K_IB
PS_384K_IB_MUX
PS_256K_IB
PS_128k_IB
PS_128k_IB_MUX
PS_128K_STR
UlIrmTable
OLS Cell Color Cell Color Cell Color
= Green = Yellow = Red
Gold 384 128 64
RTWP = -103.7dBm
Allocation/Retention Priority Level = 1
Iub
d NodeB
Loa
adio
UL R
PS Streaming RB
MIB DlRbSetConf
PS Streaming
RAB request PS_256K_STR
Candidate RB
selection PS_128K_STR
(MBR, GBR) OLS RL Cell PS_64K_STR
color color PS_16K_STR
Yes Target
RB bit rate < GBR?
RB bit rate No
>= GBR
CAC
Alcate-Lucent implements PS streaming Radio Bearers (RBs) since UA04.1. Support of Streaming RB allows
operators to differentiate streaming traffic from best effort traffic (i.e. Interactive and Background traffic) at the
transport level (e.g. Iub) or at RRM level, therefore providing streaming service of a superior quality compared to
when I/B RB are used.
When speaking about streaming quality, another important parameter is the rate at which the streaming content
has been encoded. For example, it is generally acknowledged that high quality video streaming on mobile device
requires data rate of around 100kbps, and potentially more. As a matter of fact, high quality streaming content
requires to introduce higher streaming RB bit rate such as 128 kbps or even 256 kbps. PS128kbps was introduced in
UA04.2 and PS256kbps was introduced in UA05.
Since high bit rate RBs are radio resources consuming, enhanced RRM is required to optimize radio resources usage.
iRM CAC: same as for PS I/B RAB except that the allocated RB must be of a Bit Rate greater or equal to the
Guaranteed Bit Rate required by the SGSN for the PS Streaming RAB.
New RL
UL Cost Yes No
PS 128 UEK
UL Cost
PS 384 UEI
(FDDCell)
Established RLs
RNC
UL Cost
isUlTokenCacAtRncAllowed (RadioAccessService)
PS 128 UEJ
ulCostForUlTokenCac (UlUserService)
PS 128 UEK
The costs associated to each UlUserService have been evaluated using UL RSSI measurement and depend on the
data rate or spreading factor SF:
Inorder to control the admission for the high data rate (upon SF 8) and not the low data rate and to
guarantee the old network capacity for these services, the cost of the low data rate (SF greater than 8) is set
to zero.
For high data rate bearer (SF 8), the cost depends on the throughput and on the SF.
The uplink radio admission control for high data rate calls has been introduced together with the UL PS384 RAB in
order to enable an uplink call admission control mechanism and thus avoid UL congestion.
Labtests show that in ideal radio conditions three PS I/B 384 generate a noise rise higher than 3 dB
(corresponding to 50% of UL Load). Beyond 75% load, the system is no longer stable and it could lead to
significant neighboring cell interference, cell coverage reduction and mobiles dropping calls.
The solution is to define a cost per UL RAB and a total UL capacity threshold. This cost can be tuned per UL PS RB
bit rate thanks to the ulCostForUlTokenCac parameter.
Ateach allocation, release or reconfiguration of a UL resource, the UL load is incremented, decremented or
adjusted in function of the source and target UL RAB costs.
This
UL capacity pool is compared to a configurable threshold: if below this target, the call is accepted,
otherwise it is refused.
If a high bit rate UL PS RB is limited at RAB establishment because of this feature, it can be upgraded thanks to UL
RB Rate Adaptation feature if possible (see Packet Data Management section).
rejected
rtwpMaxCellLoadNonEdch
(BTSCell) accepted accepted
accepted accepted accepted
green2yellow
UL384 UL384 UL384 UL384
Cell is Green
UL384 UL384 UL384 UL384
Node B hardware resources are usually properly dimensioned to process the achievable cell rate. However, there are
some scenarios where the bottleneck is not the Node B available resources but the UL radio interference radio
induced by the traffic.
Inthis latter case, admitting new calls or reconfiguring some of the ongoing calls with higher rates will create
too much Multi Access Interference (MAI) and consequently decrease the Radio Links quality and Cell
Breathing.
It is better in this case to reject such an RL establishment.
The improvement of the CAC is achieved by taking into account the current UL Load. If it has reached a certain
value, no new RL is admitted.
Two thresholds are defined:
Max RTWP for total UL traffic (R99+E-DCH): totalRotMax
Max RTWP for non E-DCH traffic only used for R99 CAC: rtwpMaxCellLoadNonEdch
The Node B performs a very basic CAC without considering the cost of the link to be
established/reconfigured/released:
It
compares the current UL load for non E-DCH calls to the rtwpMaxCellLoadNonEdch configurable threshold
parameter.
In case this UL load is lower or equal, it is admitted, otherwise it is rejected.
The non E-DCH UL load CAC threshold is configured in %.
As non E-DCH traffic is lower or equal to the total UL traffic (R99+E-DCH), the non E-DCH maxload should be lower
or equal to the total max load and the following parameter rule should be fulfilled:
rtwpMaxCellLoadNonEdch <= 1 1/10(totalRotMax/10)
rtwpMaxCellLoadCacActivation is used to activate the UL CAC on non-EDCH traffic at BTS based on the RTWP.
rtwpMaxCellLoadNonEdch is used only if rtwpMaxCellLoadCacActivation is set to TRUE.
AAL2 CAC
EBR
Call is accepted Call is rejected
Yes No
New RL
EBR Cost
PS 128 kbps
Common channels
* Bandwidth threshold depends on CacMethod value
Each Radio Bearer has a cost called Equivalent Bit Rate (EBR) which represents the amount of bandwidth to reserve
for the bearer, to guarantee acceptable QoS (delays, errors, etc.).
When the RNC performs resource reservation, it simultaneously covers resource allocation and admission control
(check if necessary resources are available).
First step of resource allocation aims at picking up an AAL2 connection and a CID according to the load balancing
and cid selection methods chosen.
A call admission control at ATM level is then performed by the RNC in order to prevent admission of AAL2
connections in excess of the available transport bandwidth.
This CAC applies to all AAL2 based interfaces: Iub, Iur and IuCs.
Each time a new CID needs to be allocated, the AAL2 link CAC checks, according to the CAC method configured
(cacmethod parameter), whether the cost of the call (EBR) to be allocated fits in the remaining available
bandwidth. If it doesnt fit, the CID cannot be allocated.
Algorithm Selection
dlAlgoSelector (PowerConfClass)
powerConfId (FDDCell)
PowerConfClass
After the RAB Matching and RAB Mapping algorithms have been processed, the RNC estimates the necessary power to initially
support the call.
This power estimation (Pres) corresponds to the power that will be reserved by the RNC if the admission criterion is passed.
Pres is calculated differently depending on which algorithm is used to perform the downlink power allocation:
algorithm 1: Pres = pcpichPower + maxDlTxPowerPerOls - algo1DeltaTargetPower
Where:
Pini = pcpichPower + initialDlEcnoTarget CPICH_EC/NO
The choice between these two algorithms is done through the dlAlgoSelector parameter of the PowerConfClass object:
With the dlAlgoSelector, the operator can decide which algorithm will be used in the different power control configuration
classes.
Each FDD cell points to a specific PowerConfClass, identified by powerConfId.
On RNC side, depending on the value of the parameter isCellReconfSupported (NodeB), the RNC knows if the NodeB supports
the Cell Reconfiguration procedure or not.
if it does not, the Class 2 parameters are applied.
if the NodeB supports the Cell reconfiguration, the RNC takes the Class 3 parameters. When they are changed online, the
RNC sends the Cell Reconfiguration procedure.
On OMC side:
If isCellReconfSupported is False, then the OMC maintains Class 2 and Class 3 parameters aligned: every change on
Class 3 parameters implies an update of Class 2 parameters and then a Cell lock/unlock.
If isCellReconfSupported is True, the Class 3 parameters are no more linked to Class 2 parameters and may be changed
on-line (without Cell lock/unlock)
P traffic Traffic
Power
P traffic admission Pres
(Dedicated
Channels)
Pused
Overhead
Power
(CCC+OCNS
+HSDPA+E-
DCH) Pres + Pused <=Ptraffic_admission
No Yes
Once the downlink power Pres is assessed for the call, some admission criteria are checked by the RNC.
The admission criterion is the following:
Primary link admission (call establishment): Pres + Pused Ptraffic admission
Soft handover link addition: Pres + Pused Ptraffic
Note: Pused is the sum of the Pres of all calls being actually supported.
If this criterion is fulfilled, the power Pres is reserved by the RNC. Otherwise, the call is rejected.
From UA5 release (E-DCH introduction):
Ptraffic = PMaxCell - PCCC * ActivityFactorCch - POCNS - Pedch - PminHsdpa
Where:
PMaxCell is the maximum total allowed DL power in the cell
PCCC is the total power allocated for all Common Control Channels in the cell
ActivityFactorCch is hard coded to 66%
POCNS is the optional power allocated to OCNS if needed (can be pre-empted for R99 traffic).
OCNS=Orthogonal Code Noise Simulator
Pedch is the power reserved for DL transmission of E-AGCH and E-RGCH/E-HICH channels (can be
pre-empted for R99 traffic)
PminHsdpa is the power reserved for a minimum HSDPA traffic in the cell
isBtsPowerSelfTuningActivated
(PowerConfClass)
New Allocated
Power Allocated Power
(new Pused at RNC) powerMargin
(PowerConfClass)
Measured Power
(Pused at NodeB) overEstimate
(PowerConfClass)
Allocated Power New Allocated
Measured
(Pused at RNC) Power
Power
(new Pused at RNC)
Case 1
Case 2
Power consumption
underestimated at the RNC Power consumption
overestimated at the RNC
Power consumption underestimated at the RNC: In this case, it is proposed to update the allocated power (power
consumed as seen by the RNC) based on the measured power (as measured by the Node B) plus a powerMargin.
Power consumption overestimated at the RNC: The power consumption is confirmed as overestimated if the
difference between the measured and allocated is above an overEstimate threshold. In that case, the new allocated
power (power consumed as seen by the RNC) is made equal to the measured power (as reported by the Node B).
C 256,1 P-CCPC
P-CCPCH 3GPP
H Alcatel-
C 256,2 AICH Alcatel-
Lucent Channelization Code (SIB 5)
Alcatel-
C 256,3 PICH Alcatel-
Lucent
Alcatel- Spreading Factor (SIB 5)
C 64,1 S-CCPCH Lucent
Lucent Code Number (SIB 5)
SF4
SF8
SF16
SF32
SF64
OVSF Codes Allocation
Common
SF128 Channels
SF256
For each allocation, the OVSF tree will be run from up to down (filling the gaps when any), which avoids to block too
many branches.
If a free code is found, the resource is granted to the call and the OVSF code CAC is successful, otherwise the call is
rejected and the CAC on OVSF code is declared failed.
The Dynamic DL Code Tree Management feature was introduced in UA05 in order to avoid R99 code blocking.
IDLE
RAB Assignment
Bucket Occupancy
trbEstThreshold
AO Upsize (CacOnFachParam)
AO Downsize
Each cell can only accept a limited number of simultaneous UEs in the CELL_FACH state:
Each mobile on CELL_FACH is allocated a token.
Each time a CELL_FACH admission is tried in a given cell, the current number of used tokens is compared to a
specific threshold. If below the threshold, the admission is successful and a token is allocated.
There are 2 thresholds used according to the reason for CELL_FACH admission. In the Alcatel-Lucent
implementation, they are defined as:
MaxNumberofUsersPerMacC (signaling dealing with Cell_FACH state as RRC Connection Request, Cell Update
with at least one SRB allocated-) is used to limit the number of simultaneous user connections being supported
by a given Mac-C instance.
trbEstThreshold
(transition from Cell_DCH state to Cell_FACH due to Always-On feature) defines the maximum
number of users that can have TRB configuration in CELL_FACH.
These parameters are set at the OAM in order to give a higher precedence to a new incoming call (RRC connection
request) than to a mobile already in call and aiming to transition from Cell_DCH to Cell_FACH.
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
paRatio (BTSCell)
F3
F2
F1 paRatio (BTSCell)
paRatio (BTSCell)
Cell Setup
At cell setup, the RNC calculates the max Tx Power, which is the maximum power that will be used to configure the cell:
Max Tx Power (FDDCell) = min (Max Tx Power Required, Max DL Power capability)
At the Node B level, the power is owned by Power Amplifiers which can be shared by multiple cells.
In Alcatel-Lucent configurations, cells on the same sector but on different carriers may share or not the same Power Amplifier.
This capability should allow optimization of the use of the PA. The sharing of power between different cells associated with the
same PA is static.
A configuration parameter at the OMC-B (called PA_Ratio) allows sharing of a PA power between 2 cells. From an RNC
perspective, the sharing is transparent.
Possible values of maxPowerAmplification are = {fullMode, max30W , max45W , max60W , max85W}
The Max PA Power, in dBm unit, represents the maximum output power of an MCPA board. If maxPowerAmplification is set
to the fullMode value then Max PA Power value depends on the HW type of the MCPA board. For instance, it is equal to 46.5
dBm for a 45-W MCPA and to 47.8 dBm for a 60-W MCPA?.
As the names indicate:
the class2CellReconfParams object contains Class 2 parameters.
the class3CellReconfParams object contains Class 3 parameters.
On RNC side, depending on the value of the isCellReconfSupported parameter (NodeB), the RNC knows if the Node B
supports the Cell Reconfiguration procedure or not:
If it does not, the Class 2 parameters are applied.
If the Node B supports the Cell reconfiguration, the RNC takes the Class 3 parameters. When they are changed online, the
RNC sends the Cell Reconfiguration procedure.
On OMC side:
If isCellReconfSupported is False, then the OMC maintains Class 2 and Class 3 parameters aligned: every change on
Class 3 parameters implies an update of Class 2 parameters and then a Cell lock/unlock.
If isCellReconfSupported is True, the Class 3 parameters are no more linked to Class 2 parameters and may be changed
on-line (without Cell lock/unlock).
externalAttenuationXXXDl = 0
BTS
Reference point
externalAttenuationXXXDl 0
BTS
Reference point
externalAttenuationXXXDl = 0
BTS
Reference point
Global Losses = Internal Losses +Feeder Losses +TMA Insertion Losses + Jumper Losses
externalAttenuationXXXDl 0
BTS
Reference point
Global Losses = Internal Losses +externalAttenuationXXXDl +TMA Insertion Losses + Jumper Losses
When a TMA is specified (tmaAccessType = tmaUmtsOnly or tmaMix), the transmission power reference point
moves to the antenna port of the TMA. Additional losses are taken into account:
TMA insertion losses are equal to 0.3 dB in the transmission path.
jumper losses are set to 2*0.6 dB (0.6 dB for each jumper).
If externalAttenuationXXXDl is set to 0, the feeder losses are equal to 2 dB. Otherwise, the feeder losses are
equal to the datafilled value of externalAttenuationXXXDl.
activityFactorCcch
(FDDCell)
Node B
Traffic
Power Channel parameter Object
(SHO
reserved) PCPICH pcpichPower FDDCell*
P-SCH pschPowerRelativeToPcpich FDDCell*
S-SCH sschPowerRelativeToPcpich FDDCell*
P-CCPCH bchPowerRelativeToPcpich FDDCell*
Traffic
Power S-CCPCH sccpchPowerRelativeToPcpich SCCPCH
AICH aichPowerRelativeToPcpich RACH
PICH pichPowerRelativeToPcpich PCH
Overhead
Power
(Common
FDDCell* FDDCell-> Class2CellReconfParams
Channels) or
FDDCell-> Class3CellReconfParams
1 6 11 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Power Management
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
In the slide, the Pilot power, that is, the P-CPICH power is defined by the pcpichPower parameter of the FDDCell
object as an absolute value in dBm, referenced at the BTS antenna connector.
All the other common channel powers are given relative to the P-CPICH level.
Because of the check in the BTS (CCM) at call setup, this relationship must be true for maxTxPower and
PcpichPower: PcpichPower > MaxTxPower - 15 dB.
A sensor at the output of the MCPA allows measurement of the effective output power of the amplifier. The range of
sensitivity of this sensor is [25 dBm..46.5 dBm]. So as to be sure to detect power, it is recommended that the Pcpich
Power (at PA Output) be higher than the minimum sensibility of this sensor).
PcpichPower > 25 dBm-total_losses_between_PA_output_and_reference_point
P-CPICH power is recommended to be set at:
35 dBm in case of one channel.
the half (32 dBm) if two carriers are supported by the same PA.
From UA05, when evaluating the power used by common channels, the RNC considers a certain activity factor
ActivityFactorCcch by which is multiplied the amount of power used by common channels as calculated prior to
UA05. It better reflects the actual common channel usage when calculating the Power Load Color of a cell and the
total power available for R99 and HSDPA GBR traffic Ptraffic.
RadioAccessService
f)
Co n
w er
sP o
( Ul U
wer
xP o DedicatedConf
Ul T
w ed
Allo
max
FACH
maxTxPower (FDDCell*)
powerConfId (FDDCell) PowerConfClass
minDlTxPower (DlUsPowerConf)
max UE Tx Power (UE Power Class) maxDlTxPowerPerOls (DlUsPowerConf)
DlUsPowerConf UlUsPowerConf
SIB 55
SIB
sibMaxAllowedUlTxPowerOnRach (PowerConfClass)
PRACH
Data part SIB 5
SIB 55
SIB betaC
PRACH betaD
powerOffsetPpmx (RACH)
Control part (RACHSignalled
SIB
SIB 55 GainFactors)
powerOffsetPO(RACH)
Pini
SIB 55 SIB 5
Pini =pcpichPower (FDDCell*) + RTWP + constantValue (RACH) - P-CPICH_RSCP
SIB 7
powerOffsetPpM0 is the power offset between the last transmitted preamble and the control part of the message
for PRACH CTFC0.
powerOffsetPpM1 is the power offset between the last transmitted preamble and the control part of the message
for PRACH CTFC1.
UL
DPC
C HT
xa
t Pi
ni
U p S
link -CCPC
RadioAccessService pow H or
er c F
ont ACH
rol
i n fo
IE
DedicatedConf
PowerCtrlConfClass
UlInnerLoopConf
dpcchPowerOffset
When establishing the first DPCCH, the initial power used by the UE to start the UL DPCCH transmission is:
DPCCH_Initial_power = dpcchPowerOffset CPICH RSCP
It is provided by the RNC to the UE via RRC signaling (FACH / S-CCPCH), in the Uplink power control info IE or in
the Uplink power control info short IE.
These IEs are included (one or the other) in the RRC messages of the radio bearer setup, reconfiguration and
release, transport channel and physical channel reconfiguration, RRC connection setup and re-establishment and in
the handover to UTRAN command.
OVSF1 bd
UL DPDCH I
UE
Scrambling code Modulation
OVSF256,0 bc
UL DPCCH Q
The above figure illustrates the principle of the uplink spreading of DPDCH and DPCCH. The first step, the NRZ
modulation, consists in associating a real signal to each bit of these channels. The binary value 0 is mapped to
the real value +1 and the binary value 1 is mapped to the real value -1. Then, each channel is spread by an OVSF
code. As it was mentioned before, channelization codes are only used to spread the information in uplink (not for
channel multiplexing) because synchronization between UEs is too complex to achieve:
The channelization code used for DPCCH is always Cch,256,0 (all ones).
If
only one DPDCH is used, it is spread by code Cch,SF,k , where k is linked to SF by k=SF/4. When more than
one DPDCH is used, they will all have an SF equal to 4.
After channelization, the spread signals are weighted by a gain factor (c for DPCCH and d for DPDCH). These
gain factors are quantized into 4 bits, giving values between 0 and 1. There is at least one of the values c and d
that is equal to 1. These gain factors may vary for each TFC, and are either signaled or computed.
Then, the streams of chips are summed up giving a multilevel signal. After this addition, the real-valued chips on
the I and Q branches are summed up and treated like a complex-valued stream of chips. This stream is scrambled
by a complex-valued scrambling code. For DPDCH and DPCCH, a unique scrambling code of 38,400 chips
(corresponding to one radio frame) is used. That code can be either of long or short type.
Finally, the complex chips are I and Q multiplexed and sent over the air interface. The result of all this is a BPSK
modulation, which gives us 1 bit per symbol.
UlUserService
RNC CSDTCH12_2Kx4SRBDCCH3_4K
2PSDTCH64Kx4SRBDCCH3_4K
CSDTCH12_2Kx2PSDTCH64Kx4SRBDCCH3_
4K
initialSirTarget (UlUsPowerConf) CSDTCH64Kx4SRBDCCH3_4K referenceUlRbSetConfId
PSDTCH64Kx4SRBDCCH3_4K (ReferenceUlRbSetList)
(NBAP) PSDTCH384Kx4SRBDCCH3_4K
StandaloneSRBDCCH3_4K
...
Partial blerTarget
OLPC (BlerQualityList)
CRCI SRBDCCH3_4K
CSDTCH12_2K
Partial blerTarget CSDTCH64K
OLPC (BlerQualityList) PSDTCH64K
CRCI PSDTCH128K
PSDTCH384K
2PSDTCH64K
...
OLPC UlRbSetConf
Master
The initial SIR target is sent by the RNC to the Node B through the initialSirTarget parameter. This parameter is
instantiated per RAB. Consequently, once the RNC has matched an RB onto the RAB requested by the Core
Network, it points to the initial SIR target value corresponding to this RB in the initialSirTarget parameter. This
value is transmitted to the Node B using NBAP signaling at each RADIO LINK SETUP or reconfiguration.
For each UlUserService, the list of radio bearers (UlRbSetConf) used in the multiple reference OLPC is given
through the referenceUlRbSetConfId parameter.
The outer loop power control algorithm takes into account all transport channels. For each transport channel, a
separate outer loop machine is run. Each outer loop machine updates its partial SIR target according to its transport
channel quality target (UlBlerTarget) as soon as it receives at least one transport block CRC. The partial SIR target
is then sent to the outer loop power control master.
The OLPC master determines the new SIR target as:
The maximum partial SIR target if at least one OLPC machine increases its partial SIR target.
The minimum partial SIR target if all OLPC machines reduce their partial SIR target.
Whenever the new SIR target is different from the old one, it is sent to the Node B.
ulUpSirStep
ulUpSirStep SIR Target 1
-1
BlerTarget
maxSirTarget (UlUsPowerConf)
transmitTimeInterval (static)
minSirTarget (UlUsPowerConf)
TTI
enablePeriodicSirTargetUpdate
(UlOuterLoopPowerCtrl)
Triggered Update if SIR Variation updateThreshold (DynamicParameterPerDch)
The RNC computes actualized partial SIR targets every TTI. The TTI value in milliseconds is given by the
transmitTimeInterval static parameter relative to the TrCHs used as references for the outer loop power control.
The reference TrCHs depends on the service type.
Each outer loop machine updates its partial UL SIR target according to its transport channel UL quality target
(BlerTarget) as soon as it receives at least one transport block CRC. An update from each OLPC machine to the
OLPC master is sent every update period or if the SIR target variation exceeds an upper limit (updateThreshold).
The update period is defined by ulUpdatePeriod, and is provided in a number of TTIs.
When updating the SIR target at the Node B, the RNC sends on the user plane a specific control frame, called Outer
Loop Power Control, to the Node B. Consequently, for a given DPCCH, the period between two uplink SIR target
updates cannot be shorter than the shortest TTI of the DL associated transport channels.
In UA07.1, a new parameter (enablePeriodicSirTargetUpdate) was introduced to enable or disable the periodic
sending of Sir target when computed Sir Target is equal to the last value. This occurs when there is no traffic or at
the Sir target boundaries.
QE QeThresholdForUlOLPC
Partial_UL_SIR_Target is increased
Power UP! even if the BLER is good!
The most important change is that the Bit Error Rate (Quality Estimator - QE) is taken into account.
This algorithm takes place only if the QE is bad.
1 QE QeThresholdForUlOLPC ulUpSirStep
1
Partial_UL_SIR_Target += [Nerr*UL_UPSTEP] [( N Nerr)*UL_DOWNSTEP -1
BlerTarget
2 1
UlRbSetConf
1 1
SIR target
DynamicParameterPerDch
QEThresholdForUlOlpc qeThresholdForUlOlpc
ulUpSirStep
Quality Estimate
BlerQualityList
CRC error
blerTarget
From UA07 onwards, an option is provided to also use the QE information to calculate the UL SIR target value.
3GPP defines 2 kinds of Quality Estimates which can be provided by the Node B to the RNC:
QE vs TrCh BER is specified in 3GPP 25.133
QE vs PhCh BER is specified in 3GPP 25.133
Quality Estimate is provided by the Node B in DCH FP UL frames. The type of QE depends on qeSelector IE in NBAP
Radio Link Setup/Reconfiguration messages:
qeSelector = 'Selected' means QE = transport channel BER.
qeSelector = Non-Selected means QE = physical channel BER
In ALU UTRAN, QeSelector is set by the RNC as follows:
non-selected: AMR 2nd DCH subflow and AMR 3rd DCH subflow
selected: all other DCHs
Advantages:
QE is more accurate in average.
Provides a QE threshold above which we prevent the Sir to decrease even if CRC passed.
Potential drawbacks: QE can be very variant instantaneously making difficult to set a unique threshold for a given
service.
THEN
Initial Sir Target
Partial_Ul_Sir_Target -= acceleratedSirDownStep
SirTarget
QE AcceleratedSirDownQEThreshold DynamicParameterPerDch
2
acceleratedSirDownQeThreshold
CRC error acceleratedSirDownCountThresh
old
AcceleratedSirDownStep
AcceleratedSirDownCountThreshold
1 1 1
1 6 23 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Power Management
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
The accelerated SIR Target convergence mechanism is based on a consecutive number of good frames and is
used to enable faster convergence when a certain number of consecutive good frames are received.
This mechanism aims to complement the existing Validity Condition algorithm also called Initial Convergence
algorithm which applies at call setup only. When one mechanism is active, the other is not.
Advantages:
Potential drawbacks: underestimation of Sir can lead to BLER and SIR spikes and ping-pong in OLPC adjusments
New T
UL PL Data1 P TF Data2 PL
DPC CI
C
CH
T xp
ow DL
er DP C
CH
Tx
Po w
er C
om
ma
nds
The Uplink Power Control is controlled by the Node B which orders Power Control Commands (increase or decrease)
through TPC bits in DL DPCCH channels.
The UE then applies the PC command at the next UL DPCCH transmission.
DPDCH power is then adapted thanks to gain factors as seen previously.
Rake
Pilot
DPCCH SIRestimate
TPC
TPC > or <
0 or 1
SIRTarget
Down Command:
if SIRestimate > SIRtarget then TPC = 0
Up Command:
if SIRestimate < SIRtarget then TPC = 1
The uplink inner loop power control algorithm is located in the Node B physical layer. It is a fast procedure (up to
1500 Hz power change rate) used to derive power control commands (to be applied by the UE) from the SIR target
(set by the RNC) and UL measurements.
The Node B estimates the instantaneous SIR on the pilot bits received on the UL DPCCH and compares it to the SIR
target signaled by the RNC. In case the instantaneous SIR is lower (respectively higher) than the target SIR, an up
(respectively down) command is sent to the UE in the downlink DPCCH TPC field of each DPCCH radio time slot:
up command: TPC = 1
down command: TPC = 0
In every slot, there is either an up or a down power control command: this process does not provide good stability
of the transmission power.
TPC commands are computed in each Node B independently from the others, so if the UE is in Soft Handover with
several Node Bs, the TPC commands received from the different Node Bs may be conflicting. In the case of a softer
handover, the unique Node B involved sends the same TPC command on all the radio links of the same radio link
set.
Softer handover
identical TPC commands
for the radio link set
powerCtrlAlgo (UlInnerLoopConf)
TPC1
SIR target
algo1 algo2
Node B 1
Soft handover
1 RL
possibly different TPC commands
One TPC_cmd
In the case of soft handover (where TPC commands come from different Node Bs), the UE has to combine different
TPCs in order to derive one single internal TPC_cmd (internal power control command applied to adjust the UL
transmission power).
There are 2 standardized algorithms (named: algorithm 1 and algorithm 2 in the 3GPP Specifications) for the UE to
process TPC commands.
The choice between these two algorithms is under the control of the RNC 1000, and is managed through the
powerCtrlAlgo parameter (manufacturer parameter).
It is UE specific in the sense that a specific message is sent to each UE in order to indicate which algorithm to use,
but Alcatel-Lucent sets the same value for all cells managed by an RNC.
x ulTpcStepSize x ulTpcStepSize
ulTpcStepSize (UlInnerLoopConf)
This algorithm is well adapted for average speed UEs in urban or suburban environments. The principle of algorithm
1 is that the UE adjusts its DPCCH transmission power every slot (frequency = 1500 Hz), according to TPC_cmd
(internal power control command applied to adjust the UE transmission power) derived from the TPC commands
received from all Node Bs involved in the communication.
We can distinguish three cases of TPC_cmd generation:
Nomacrodiversity: the UE receives a single TPC command in each slot (on the single radio link established for
the communication), from which it derives a TPC_cmd as follows:
if TPC command = 0, then TPC_cmd = -1
if TPC command = 1, then TPC_cmd = 1
Softer
handover: in this case, the UE is aware (from TPC combination index parameter transmitted through
RRC protocol) that it will receive identical TPC commands in the downlink. The UE is then able to combine these
commands into a single TPC command, for example (UE implementation is proprietary) using maximum ratio
combining with all TPC commands received in order to optimize the TPC command decoding.
Softhandover: in this case, the TPC commands may be different. This case may even involve a softer handover
(from which a single TPC is derived, using for example MRC). The UE has first to use soft decision in order to
decode the different TPC commands transmitted. Then it has to combine them in order to deduce a single
TPC_cmd value. This TPC_cmd is equal to 1 only if all TPC received from other Node Bs are equal to 1,
otherwise, TPC_cmd is equal to -1.
Then, after deriving a unique TPC_cmd, the UE implements a power change based on the ulTpcStepSize
parameter of the UlInnerLoopConf object.
Algo2 output +1 -1 0
x ulTpcStepSize
UE Tx output power
ulTpcStepSize (UlInnerLoopConf)
This algorithm is adapted to high- or low-speed environments (typically: dense urban or rural). With this algorithm,
the UE concatenates N TPC commands received on consecutive radio slots to derive a TPC_cmd to be applied after
the Nth slot. N can be different according to the handover situation, but it does always divide 15 (the combining
window of the TPC commands does not extend outside the frame boundary). Allowing a decision every N = 5 radio
slots instead of every slot, algorithm 2 is a way of emulating step sizes smaller than 1 dB (typically: 0.2 dB or 0.4
dB, corresponding to the step sizes of 1 and 2 dB respectively, but applied every 5 TSs).
Note: In the TPC combination algorithm 2, the TPC_cmd is either 1, 1 or 0.
Algorithm 2 works in the following way:
Nomacrodiversity: the UE concatenates commands received from 5 consecutive TSs to derive a TPC_cmd
value:
For the first 4 slots of a set, TPC_cmd = 0
For the fifth slot of a set, the UE uses hard decisions on each of the 5 received TPC commands as
follows:
if all 5 hard decisions within a set are 1, then TPC_cmd = 1
if all 5 hard decisions within a set are 0, then TPC_cmd = -1
otherwise, TPC_cmd = 0
Softer
handover: similarly to what happens for algorithm 1, for each slot, the UE soft-combines the TPC
commands known to be the same (received from the same Node B).
+1 0 +1
0 -1 0
+1 -1 -1
Sum of TPC
N
> 0.5 < - 0.5 Otherwise
Algo2 output +1 -1 0
x ulTpcStepSize
UE Tx ouput power
Soft handover: the derivation of the TPC_cmd from the TPC commands of the different radio links is done in the
following way:
First,
the UE determines 1 temporary TPC command called TPC_tempi for each of the N sets of 5 TPC
commands. It is done as follows:
If all 5 hard decisions within a set = 1, TPC_tempi = 1.
If all 5 hard decisions within a set = 0, TPC_tempi = -1.
Otherwise, TPC_tempi = 0
Then the UE derives the combined TPC_cmd for the 5th slot as a function of all the N TPC_tempi:
TPC_cmd = 1 if (Sum of TPC_tempi)/N > 0.5
TPC_cmd = -1 if (Sum of TPC_tempi)/N < -0.5
Otherwise TPC_cmd = 0
Finally, after deriving a unique TPC_cmd, the UE implements a power change:
Uplink Power Change = TPC_cmd x ulTpcStepSize.
maxDlTxPowerPerOls (DlUsPowerConf)
minDlTxPower (DlUsPowerConf)
cell = N
S
P-CPICH_Ec /No Equivalent = P-CPICH_Ec /No(cell)
cell =1
When a traffic (dedicated) channel is set up, it is done at a certain downlink power called Pini defined by the
following equation:
Pini = pcpichPower + initialDlEcnoTarget CPICH_Ec/No
Where pcpichPower is the downlink P-CPICH power, initialDlEcnoTarget depends on the service allocated to the UE
(access stratum configuration) and CPICH_EC/N0 is the EC/N0 of the Pilot received by the UE.
The Pini is used in the Call Admission Control downlink power reservation algorithm.
The downlink transmission power is limited by an upper and lower limit for each radio link. This limitation is set
through the maxDlTxPowerPerOls and minDlTxPower parameters (DlUsPowerConf object). Both parameters
actually provide a value for each access stratum configuration, so they correspond to a set of values rather than to a
single value. The value (in dB) of these parameters is provided with respect to P-CPICH power defined by the
pcpichPower parameter.
For SHO Leg Addition, the initial power is calculated once for all the new links to be added. Pini depends not only on
the CPICH Ec/No of the selected cell to be added, but also on all the CPICH Ec/No of the cells of the old active set.
An equivalent CPICH Ec/N0 is calculated:
N CPICH Ec / N 0( celli )
CPICH _ Ec / N 0 (dB ) 10 log 10 10
equiv
i 1
po3ForPilotBits (DlUserService)
po2ForTpcBits (DlUserService)
po1ForTfciBits (DlUserService)
The RNC can also configure static downlink physical channel parameters in the Node B. In the downlink, it is
possible to give power offsets to the pilot, TPC and TFCI fields of the DPCCH relative to the DPDCH.
They are given at radio link setup in the Power Offset information IE:
PO1: TFCI bits
PO2: TPC bits
PO3: pilot bits
In the Alcatel-Lucent implementation, the power offsets used to determine the transmission power of the TFCI, TPC,
and PILOT bits are defined by the po1ForTfciBits, po2ForTpcBits and po3ForPilotBits parameters respectively.
These parameters of the DlUserService object are transmitted in the Power_Offset_Information IE of the RADIO
LINK SETUP, ADDITION or RECONFIGURATION (NBAP signaling). They are identical for all TFCs in the TFCS.
isDynamicPO2Enabled
(DlUserService)
Nb of RL Power Offset 2
1 po2ForOneRlSet
2 po2ForTwoRlSet
3 or more po2ForThreeRlSetOrMore
The DL Power Control Management provides an option to dynamically configure different PO2 values depending on
the number of radio link sets (RLS) involved in the call.
The PO2 value is variable depending on the number of RLs involved in the call.
The initial PO2 is provided to the Node B in NBAP messages.
Then,
at each new RL Setup/Addition/Deletion, the RNC shall support the transmission of the RADIO INTERACE
PARAMETER UPDATE message over the Iub and Iur user plane interfaces, as per TS 25.427, for the signaling of
TPC PO updates.
RNC DlUserService
CSDTCH12_2Kx4SRBDCCH3_4K
2PSDTCH64Kx4SRBDCCH3_4K
CSDTCH12_2Kx2PSDTCH64Kx4SRBDCCH3_4K
Initial Quality target CSDTCH64Kx4SRBDCCH3_4K
PSDTCH64Kx4SRBDCCH3_4K
RB Setup
PSDTCH384Kx4SRBDCCH3_4K
TPC StandaloneSRBDCCH3_4K
Quality ...
Measurements
TX
SRB
TRB
The DL outer loop power control algorithm is mobile-manufacturer specific, and DL power control outer loop is not
necessarily based on SIR (as UL outer loop is). The only information signaled to the UE by the RNC is a quality
target for each radio bearer, expressed as a BLER. This quality target is sent to the UE through RRC signaling (DL
Outer Loop Control procedure) for each transport channel of the connection. This quality target information is
mandatory for handover to UTRAN, radio bearer setup and transport channel reconfiguration messages. It is
optional for radio bearer reconfiguration and release, RRC connection setup, and re-establishment messages.
The DL outer loop power control algorithm is located in the UE, but the RNC may further use the downlink Outer
Loop control procedure to control the DL outer loop algorithm in the UE. To prevent the UE from increasing its DL
BLER target value above its current value (the initial one, transmitted by the RNC via RRC signaling), the RNC sets
the Downlink Outer Loop Control IE to increase not allowed. This allows reducing the impact of the UE
proprietary outer loop algorithm on the system.
isDlReferenceTransportChannelAllowed indicates that the first Transport Channel of the RbSetConf is allowed
to be used as an Outer Loop Power Control Reference Transport Channel.
BlerTarget is used if isDlReferenceTransportChannelAllowed is TRUE. BlerTarget is the BLER DL quality
target which must be met during Outer Loop Power Control. It is the Log10 of the BLER.
DL TX_PWR optimization
is UE constructor dependent
(not necessarily PL TFCI TPC
based on SIR
measurements)
TPC
TPC =1=> TPC_
cmd = +1
TPC_cmd TPC =0 => TPC_
cmd = -1
BLER est
New
BT
appl S outpu DL power
ied t
o DP t power
UE Algo change
DCH
PTPC +PBalancing
DL Power Change =
SHO
SPBAL
TPC_cmd x dlTpcStepSize(DlInnerLoopConf)
The DL inner loop power control algorithm is a fast procedure (1500 Hz) used to optimize DL transmission power by
sending power control commands to the Node B in the TPC field of UL DPCCH time slots.
At each TPC (Transmit Power Command = 0 or 1) field decoded (on UL DPCCH), the BTS estimates the TPC_cmd
(TPC command = -1 or 1) based on TPC and Limited_Power_Increase values, and implements a DL power change
as shown in the above slide.
As the Limited_Power_Increase functionality is not implemented, TPC_cmd values are directly deduced from TPC
values as following:
TPC = 0 => TPC_cmd = -1
TPC = 1 => TPC_cmd = 1
So TPC_cmd never has the value 0 (either decrease or increase command for the transmission power), as with
combination algorithm 2 for UL power control.
The downlink power adjustment (increment or decrement according to the power control command) step size is
tuned through the dlTpcStepSize parameter. This parameter is transmitted by the RNC to the Node Bs using the
TPC_DL_Step_Size IE contained in the RADIO LINK SETUP REQUEST message (NBAP). It cannot be reconfigured
during the connection. 3GPP TS allowed values are: 0.5 dB, 1 dB (mandatory), 1.5 dB and 2 dB. Alcatel-Lucent
implementation proposes only the two mandatory values: 0.5 dB and 1 dB.
P2
time
RL Addition
isPowerBalancingAllowed (RadioAccessService)
powerBalancingRequired (DlUserService)
adjustmentPeriod (PowerBalancingAdjustmentParameters)
Radio Frame
The objective of the downlink power balancing function is to equalize powers on the different radio links, eliminating
power drifting effects.
This function is triggered by the SRNC, which provides balancing parameters to the Node Bs and executed by the
Node Bs.
The power balancing function brings a corrective factor Pbal which is added to the power as calculated by the DL
inner loop power control.
This Pbal is such that:
SPbal = (1 R).(Pref + Ppcpich Pini)
where:
SPbal is the sum of these corrective factors over an adjustment period corresponding to a number
of frames
Pbal = 0 or -0.5 or 0.5 dB (in first implementation)
R is the adjustment ratio
Pref is the value of the DL Reference Power
Ppcpich is the power used on the primary CPICH
Pinit is the power of the last slot of the previous adjustment period
Instead of specifying which maximum correction should be applied to one slot, a period is specified, as a number of
time slots, where the accumulated power adjustment should not be greater than 1 dB.
The above slide shows an example with SPbal = - 4 dB, adjustment period = 2 Radio Frames, max adjustment step
= 5 Time Slots.
1 0 1 0 0 1 PL TFCI TPC
U L DP
CCH
500 TPC commands / s
1 1 1 0 0 0
DPC_Mode = 1 (tpcTripletInSoft)
dpcMode (DlInnerLoopConf)
The RNC may activate a rate reduction algorithm. If rate reduction algorithm is applied, then the UE issues one new
command every 3 slots and repeats it over 3 slots, so the DL inner loop TPC commands frequency is divided by 3
(1500 Hz down to 500 Hz).
This algorithm is controlled by the dpcMode parameter (DlInnerLoopConf object), which is signaled to the UE in
the Downlink DPCH Power Control Information IE using RRC signaling:
IfdpcMode = singleTpc (0 on ASN1 interface), then the UE sends a specific TPC command in each DPCCH time
slot (starts in the first available slot).
If
dpcMode = tpcTripletInSoft (1 on ASN1 interface), then the UE repeats the same TPC command over 3
successive DPCCH time slots.
On reception of the TPC field in the UL DPCCH, the Node B processes the command depending on the DPC_MODE
and calculates PTPC(k):
DPC_MODE = 0 => at each slot:
PTPC(k) = TPCDLStepSize if TPC = up
PTPC(k) = -TPCDLStepSize if TPC = down
DPC_MODE = 1 => each 3 slots:
PTPC(k) = TPCDLStepSize if 3 last TPCs are up
PTPC(k) = -TPCDLStepSize if 3 last TPCs are down
nInSyncInd (FDDCell*)
nOutSyncInd (FDDCell*)
tRlFailure (FDDCell*)
RL Failure RL Restore
SIRAV > -3dB => InSyncInd
The uplink radio link sets are monitored by the Node B, to trigger radio link failure/restore procedures. Once the radio link sets
have been established, they will be in the in-sync or out-of-sync states.
When the radio link set is in the in-sync state, after receiving nOutSynchInd consecutive out-of-sync indications, the Node B
shall:
start timer tRlFailure;
upon receiving nInSynchInd successive "in sync" indications from Layer1:
Stop and reset timer tRlFailure;
if tRlFailure expires:
The Node B shall trigger the RL Failure procedure and indicates which radio link set is out-of-sync. When the RL
Failure procedure is triggered, the state of the radio link set will change to the out-of-sync state.
The RNC receiving a Radio Link Failure Indication message from the Node B will trigger the call release (call drop
radio in this case) if no radio link remains in "in sync" state.
When the radio link set is in the out-of-sync state, after receiving nInSynchInd successive in-sync indications Node B shall trigger
the RL Restore procedure and indicate which radio link set has re-established synchronization. When the RL Restore procedure is
triggered, the state of the radio link set will change to the in-sync state.
Similar Radio Link Control is implemented in DL in the UE thanks to UeTimerCstConnectedMode object parameters:
n315 UE constant is analog to nInSynchInd
n313 UE constant is analog to nOutSynchInd
Activation Rules:
(nOutSyncInd * 10) + tRLFailure < (n313 * 10) + t313 + PA off
t315 > rrcReestPSMaxAllowedTimer
UE Node B RNC CN
bad SIR
nOutSyncInd
UL synchronization
failure on last RL
isPsRrcReestablishAllowed
tRlFailure isCSRrcReestablishAllowed
(RadioAccessService)
Iu Release Command
Radio Link
Deletion Req
Radio Link
Deletion Resp Iu Release Complete
A call drop is triggered as soon as the loss of the last RL is detected by the RNC.
isPsRrcReestablishAllowed (resp. isCSRrcReestablishAllowed) is the parameter used to activate or de-
activate the RRC Connection Re-establishment feature for CS (resp. PS).
UE Node B RNC CN
bad SIR
isPsRrcReestablishAllowed
t313 isCSRrcReestablishAllowed
N313 n313
DL synchronization (UeTimerCstConnectedMode) (RadioAccessService)
failure on last RL
T313
false
expiry
Cell Update
(radio link Iu Release Request
T314 Radio cnx with UE lost
failure)
Iu Release Command
Radio Link
Deletion Req
t314 Radio Link
(UeTimerCstConnectedMode) Deletion Resp Iu Release Complete
A call drop is triggered as soon as the loss of the last RL is detected by the RNC.
Note: t314 must be set equal to a non-zero value so that the UE performs a Cell Update procedure.
UE Node B RNC CN
timerPoll isPsRrcReestablishAllowed
X same RLC block isCSRrcReestablishAllowed
X not decoded (RadioAccessService)
maxDat
Maximum number of RLC
re-transmission reached
false
resetTimer
X RLC Reset
not decoded maxNbrOfResetRetrans
X
Maximum number of RLC Reset
re-transmission reached
Iu Release Request
DL RLC Error SRB/TRB timerPollPeriod
maxDat
Iu Release Command resetTimer
Radio Link maxNbrOfResetRetrans
Deletion Req (DlRlcAckMode)
Radio Link (UlRlcAckMode)
Deletion Resp Iu Release Complete
UE Node B RNC CN
false
RLC Reset X
not decoded
X
Maximum number of RLC Reset
re-transmission reached
Cell Update
(rlc unrecoverable Iu Release Request
error) UL RLC Error SRB/TRB
Iu Release Command
Radio Link
Deletion
Iu Release Complete
P ICH pda
te
P-C Cell
U
Last RL lost detection
rrcReestPSMaxAllowedTimer
Cell Update rrcReestCSMaxAllowedTimer
(failure Cause, CPICH_Ec/No) (RadioAccessService)
CPICH_Ec/No >=
rrcReestablishPSThreshold
CPICH Ec/No
Yes No
rrcReestablishPSThreshold
rrcReestablishCSThreshold
(RadioAccessService)
UE Node B RNC CN
UL synchronization
failure on last RL
tRlFailure
rrcReestPSMaxAllowedTimer
rrcReestCSMaxAllowedTimer true
(RadioAccessService)
Cell Update (radio link failure)
stopped
Radio Link Setup
RB Reconfiguration
Complete
1 6 48 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Power Management
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
A similar scenario can occur in case the UE detects a DL Radio Link Failure: the RNC receives a Cell Update from the
UE without expecting it.
This scenario applies only for (CS+PS) calls case because there is no TRB established in RLC ACK mode for CS calls
but TM mode is used. For CS call case, the RLC unrecoverable error can occur on SRB only, leading to a call drop.
PS Only (FACH ) Drop Call Drop Call Drop Call Not Supported
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
9
S R9 Dedicated Channel
U MT
Dedicated Channel
Dedicated Channel
Conversational CS
Streaming CS
Streaming PS
Interactive
Interactive PS
Background PS
178 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Management
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
Call Management is a set of reactive mechanisms performed during the call in order to improve the system capacity,
by adapting the resources used by the users to different situation (traffic activity, real traffic, RF conditions, etc.):
AlwaysOn: Downsize and upsize of user resources are performed depending on user traffic activity in DL & UL
(two-step mechanism depending on whether there is low traffic or no traffic at all).
RBRate Adaptation: Downsize and upsize of user resources are performed depending on user real time traffic
and buffer occupancy in DL & UL.
iRM Scheduling: Downsize and upsize of user resources are performed depending on user radio conditions in
DL.
iRM Preemption: Downsize of user resources are performed depending on cell load and user priority in DL.
Preemption: Downsize of user resources are performed depending on CAC failure and call priority in DL & UL. If
downsize is not enough to free resources, some calls can be release (low priority users).
AMR Rate Change: Downsize and upsize of user resources (codec) are performed depending on radio
conditions (UL interference & DL power) and Node B resources (Baseband, Iub) in DL & UL.
Most of these Call Management features operate only on UMTS PS I/B RABs since no Guaranteed Bit Rate is defined
for such traffic classes. However, iRM Scheduling is also available for PS Streaming services so as to avoid call drops
when UE moves in poor radio quality areas. Preemption process applies to any types of call wherease AMR rate
change applies to Multi-Mode AMR calls only.
Throughput Threshold
AO Step 1
T1
AO timers
CELL_PCH
CELL_DCH CELL_FACH or PMM-idle
URA_PCH
Dedicated radio resources are not optimal to support packet services with sporadic traffic. In order to find the best
trade-off between efficient resource usage and subscriber comfort, the Always-On concept developed is composed
of three steps.
After a first period of low activity (T1), the bearer is reconfigured to a predefined downsized bearer configuration,
which consumes less radio resources.
If traffic activity is detected, the bearer is upgraded back to its initial configuration (or to a degraded one if network
congestion is meanwhile detected).
If no traffic activity is detected during a second period of time (T2), then the radio bearer is released but the RRC
connection remains as well as the Iu Connection in order to speed up the needed radio bearer setup in case or user
traffic resumption.
If neither traffic nor signaling activity is observed during a third period of time (T3) then the RRC and Iu connection
are released but the following context info remains between UE and Network:
the PDP context at the SGSN
the PPP (or IP) link between UE and ISP
the SGSN-GGSN tunnel
When downsize criteria is met, the Always-On downsized RB (FACH) is determined at the OAM, thanks to the
following parameters:
DL downsized RB: alwaysOnDlRbSetFachId (AlwaysOnConf object)
UL downsized RB: alwaysOnUlRbSetFachId (AlwaysOnConf object)
User Traffic
Volume
AO step2 AO step1
Throughput Threshold
AO Step 1
traffic
resuming
O kbps Throughput Threshold
AO Step 2
TTT
AO timers
CELL_PCH
or CELL_FACH CELL_DCH
URA_PCH
RRC Context
In Cell_PCH or URA_PCH states, although the connection is no more active, the mobile keeps its PDP context active.
Therefore, a traffic resume is done either:
By the mobile, which re-establishes a connection to the network.
Orby the network by paging the mobile, which would have the effect of creating a new connection. The
dataflow is the same as the mobile initiated resume except for the paging phase.
In Cell_FACH state, the RNC assess the user data throughput and decides to perform an AO upsize to DCH radio
bearer if the high user throughput is detected.
step1AverageWindow(AoOnFachParam) step2AverageWindow(AoOnFachParam)
step1DlUlThroughputThreshold
(AoOnFachParam)
step2ThroughputThreshold(AoOnFachParam)
T1 T2
The AO downsize step1 condition is based on DL and UL traffic volume monitoring on non-sliding time windows. The
downsize criterion is met if:
(TBsizex NbTB) / Step1AverageWindow < Step1DlUlThresholdThroughput
during at least TimerT1.
With:
NbTB: Number of Transport Blocks transferred during the time window
TBsize: size of a L1 Transport Block (in bits)
AO Downsize is performed when UL and DL criteria are met.
The AO downsize step2 decision is based on DL and UL traffic volume monitoring on non-sliding time windows. The
release criterion is met if:
(TBsize x NbTB) / Step2AverageWindow < Step2ThresholdThroughput
at least during TimerT2 seconds
The UE may keep or not its RRC connection or not depending on the usage of the Cell_PCH/URA_PCH states or not.
Reporting
UE RLC/MAC Buffer Occupancy Reporting event4a
event4a Report
repThreshold (AoOnFachParam)
step1DlRlcBoThreshold(AoOnFachParam)
Upsize step1TimerBetween2Reports
(AoOnFachPara
m)
The DL upsize condition relies on the same kind of mechanism. As the sum of Buffer Occupancies of RBs multiplexed
onto the FACH exceeds a certain threshold for a given UE, the SRNC considered the DL upsize condition as fulfilled.
The parameter Step1TimerBetween2Reports is used to avoid sending unnecessary upsize required event reports
during the execution of the upsize procedure. This parameter sets the minimum time between the emissions of two
events "upsize required" by the RNC-IN.
CELL_DCH CELL_FACH
Nominal RB Fallback RB
RNC isOamCappingOfDataAllowed
(RadioAccessService)
= False
RB Reconfiguration
= True
maxDlEstablishmentRbRate
reconfigTimeFachDch maxUlEstablishmentRbRate
(FachRetransmissions)
(CacConfClass)
RB Reconfiguration
reconfigRetriesFachDch maxDlRateTransitionToDchDlTriggerDchAndDch
(FachRetransmissions) maxDlRateTransitionToDchUlTriggerDchAndDch
maxUlRateTransitionToDchDlTriggerDchAndDch
RB Reconfiguration x N maxUlRateTransitionToDchUlTriggerDchAndDch
(DchRateCapping)
RB Reconfiguration Complete
From UA7 onwards, the AO success rate during the Cell-Fach to DCH can be improved by repeating the message RB
Reconfiguration:
reconfigTimeFachDch: defines the timer for retransmission of reconfiguration messages in Cell FACH due to
response message timeout for Cell_FACH to Cell_DCH transition.
reconfigRetriesFachDch: defines the maximum number of reconfiguration message retries to transmit for
Cell FACH to Cell_DCH transition.
The rate limitation is applied only on a DCH transport channel for initial state and can be modified later by other
algorithms like RRA.
isSib11IntraFreqOneShotAllowedPerCell isSib11IntraFreqOneShotAllowed
isIntraFreqOneShotDchAllowed
(FDDCell) softHoAddThresholdOneShot
intraFreqOneShotFilterCoefficient (RadioAccessService)
(MeasurementConfClass)
g.
co nfi RNC
e nt
ur em
eas
otm
h
n eS
it hO Cell_FACH
1 1w
SIB RRC/BCH (SIB11 with One Shot Meas. setting)
l
C el
vi n g RRC/RACH (Event4A)
Se r
Cell_DCH
isSib11MeasReportingAllowed
RRC/DCH (Measurement Report Cells (up to 3 best))
(FDDCell)
The objective of this feature is to improve the mobility management of the UE once it arrives to DCH state for Idle
state - RRC Establishment, FACH to DCH Transition and PCH to DCH Transition. The focus is on FACH to DCH
transition.
From UA7 onwards, if isSib11IntraFreqOneShotAllowed is set to true, then the info to be included on SIB 11 is
the OneShot periodic measurement. This info replaces the normal 1a,b,c events by the Intra-Frequency One Shot
measurement.
Once the UE has this information thru SIB11, it can send Measurement reports in a very early stage (before call
management is ready to process it), in this case the event will be stored and processed as soon as possible by Call
management.
On reception of a One-Shot report following a transition from Cell_FACH to Cell_DCH, the RNC will update the iRM
RL color based on CPICH Ec/N0 measurements and determine whether to trigger SHO on the neighbor cells based
on a configurable threshold softHoAddThresholdOneShot.
For UEs that do not support the measurement configuration via the SIB11, the RNC can send an RRC Measurement
Control message with the following settings; periodic, the Reporting Quantity EC/Io with Amount of Reporting set to
1. This is done if the isIntraFreqOneShotDchAllowed parameter is set to True.
The One shot Ec/Io report has the same meas. Id as the inter-RAT (id=03), this is not problematic since the
intrafrequency one shot report has an amount of reporting =1, and it will be sent in an early stage of the call when
normally there is no 2D event configured.
Once the 2D event is triggered, another Measurement control will be sent with the same id but will overwrite the
intrafrequency settings of one shot with new inter-RAT IE.
isSib11IntraFreqOneShotDchAllowed is an RNC flag. This flag is supposed per FDDcell from UA07.1 and
corresponds to the isSib11IntraFreqOneShotDchAllowedPerCell parameter.
In UA07.0, a workaround was defined later in order to allow the activation of One shot per cell. The
reserved0.FddCell parameter is used to allow cell level control of this feature.
Note: The OneShot Measurement is only valid for Cell Reselection without HCS.
Copyright 2012 Alcatel-Lucent. All Rights Reserved.
TMO18255_V2.0-SG-UA08-Ed1 Module 1.7 Edition 1
Section 1 Module 7 Page 16
2 Always on
2.8 RRC states transitions
T2 T3
CELL_DCH CELL_FACH
Nominal RB T1 Fallback RB
Note: PCH DCH occurs in PS MUX, when PS traffic resumes or at CS RAB establishment
1 7 17 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Management
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
As explained in TS 25.331, "The RRC states within UTRA RRC Connected Mode reflect the level of UE connection
and which transport channels that can be used by the UE."
Whenthe RNC receives a RAB assignment request, the corresponding Radio Bearer is by default allocated in
CELL_DCH.
Then, later on during the call, a UE can be moved between CELL_DCH and CELL_FACH based on user activity
(i.e. user traffic volume monitoring), that can be controlled by the operator thanks to inactivity timers.
Since CELL_FACH makes use of RACH and FACH, which are common transport channels (shared between all the
users of the cell), CELL_FACH is only suited to non real-time data services (i.e. Interactive or Background) and can
even be used to transmit small amounts of user data. However, it cannot be used for real-time traffic, such as voice
or video telephony, which are always supported in CELL_DCH.
Always-On is the Alcatel-Lucent PS call management feature responsible for choosing the best radio resources
according to the amount of traffic the subscriber has to transmit.
From UA05.0, the Always-On mechanism supports these two RRC states: URA_PCH and CELL_PCH.
PCH sates (i.e. CELL_PCH and URA_PCH) are useful for data subscribers who can fallback to one of these states
when they are completely inactive:
Sinceno cell resources are allocated to UE in these states, i.e. no dedicated physical channel is allocated to the
UE, they have no impact on the cell capacity.
Nevertheless,subscribers benefit from a quicker re-establishment time compared to when in Idle mode and the
UE battery consumption is low, i.e. equivalent to when the UE is in Idle mode.
Dat
a to t
ran
CELL_DCH s mit CELL_FACH
When CELL_PCH is used, the transitions between URA_PCH and the other states are the following:
Transition from CELL_PCH to URA_PCH:
When in CELL_PCH, the transition to URA_PCH occurs when the user has performed a minimum number
of CELL UPDATE procedures. Therefore, this transition is based on the Cell Update signaling load and not
on the user traffic activity. Hence, this transition is not directly related to AO.
nbOfCellUpdatesFromCellPchToUraPch is used to control the transition from CELL_PCH to URA_PCH
state in case the both are activated. It represents the threshold value for the number of cell update
procedures (with cause Cell reselection) initiated by the UE in CELL_PCH state (for a maximum duration
of CellPCHtoIdleTimer) for the RNC to trigger a state change to URA_PCH for this UE.
Transition from URA_PCH to CELL_FACH:
In case some data need to be transmitted, the UE is transferred to CELL_FACH:
In uplink, access is performed by RACH.
In downlink, UTRAN sends a paging request message (PAGING TYPE1).
Transition from URA_PCH to idle through CELL_FACH:
Once in URA_PCH, if the subscriber is completely inactive, i.e. no traffic during a certain period
(URAPCHToIdleTimer), then the UE is further moved to Idle mode.
Transition to CELL_FACH is required to perform the RRC signaling connection release.
URA_PCH
(T2= Ina = UraPchEnabled
fach ctive us
T oU
raPcer
hTim
e r)
Dat
a to t
ran
CELL_DCH s mit CELL_FACH
When CELL_PCH is not used, the transitions between URA_PCH and the other states are the following:
Transition from CELL_FACH to URA_PCH:
When in CELL_FACH, the amount of user traffic is monitored in both uplink and downlink directions.
When there is no traffic during a certain period of time (FACHToURAPCHTimer) and CELL_PCH is not
enabled, the UE is moved to URA_PCH.
The transition criteria are the same as those used for transition to idle mode, i.e. traffic volume
measurement on DTCH in both uplink and downlink directions.
Similar to the transition from Cell_FACH to DCH the AO success rate can be improved from UA07
onwards by repeating the message RB Reconfiguration:
reconfigRetriesFachPch: defines the maximum number of retransmissions of RB
Reconfiguration message due to response message timeout.
reconfigTimeFachPch: defines the timer for retransmission of the RB Reconfiguration messages.
The transitions between CELL_PCH and the other states are the following:
Transition from CELL_FACH to CELL_PCH:
When in CELL_FACH, the amount of user traffic is monitored in both uplink and downlink directions.
When there is no traffic during a certain period of time (FACHToCellPCHTimer), the UE is moved to CELL_PCH.
The transition criteria are the same as those used for transition to idle mode, i.e. traffic volume measurement on
DTCH in both uplink and downlink directions.
Like for the transition from CELL_FACH to URA_PCH the AO success rate can be improved from UA7 onwards by
repeating the message RB Reconfiguration using reconfigRetriesFachPch and reconfigTimeFachPch
parameters.
Transition from CELL_PCH to URA_PCH through CELL_FACH (if URA_PCH state is used):
Once a UE is in CELL_PCH, and if URA_PCH is enabled, the RNC increments a counter that counts the number of cell
updates.
When the number of cell updates has exceeded a certain limit (NumberOfCellUpdatesFromCellPchToUraPch),
the RNC moves the UE from CELL_PCH to URA_PCH.
Transition to CELL_FACH is required to perform the transition signaling.
Transition from CELL_PCH to CELL_FACH: In case some data need to be transmitted, the UE is transferred to CELL_FACH:
In uplink, access is performed by RACH.
In downlink, UTRAN sends a paging request message (PAGING TYPE1).
Transition from CELL_PCH to Idle mode, through CELL_FACH:
Once in CELL_PCH, if the subscriber is completely inactive, i.e. no traffic during a certain period
(CellPchToIdleTimer), then the UE is further moved to Idle mode.
Transition to CELL_FACH is required to perform the release signaling.
Without direct transition from Cell/URA_PCH to DCH, direct RRC state transition from URA_PCH or CELL_PCH to
CELL_DCH is supported only for multi-RAB call or upon a mobile terminating CS RAB request.
This feature extends the direct transition support to all RAB combinations and to all Always-On upsize triggers.
It is based on:
Establishment Cause and/or
Traffic Volume Indicator (TVI)
which are included in the CELL UPDATE message, or through CELL_FACH
The decision to go directly to CELL_DCH or through CELL_FACH is based on the Establishment Cause and/or the
Traffic Volume Indicator (TVI) included in the Cell Update message.
This feature is an enhancement of the AO, base feature RRC & ALWAYS-ON STATES MANAGEMENT. The RRC &
ALWAYS-ON STATES MANAGEMENT is in charge of monitoring the radio resources allocated to a UE with regards to
the actual traffic activity of the user.
FEATURE BENEFITS
Direct transition from PCH to DCH improves the user perceived latency when traffic resumes from CELL_PCH or
URA_PCH. The transition delay versus the two-step transition is reduced from approximately 900 to 550 ms.
A direct transition from PCH to DCH involves a lower number of exchanged signaling messages between the UE and
the RNC versus a two-step transition. Thus the signaling load of the RNC control plane can be reduced.
Direct transition can avoid the intermediate step in CELL_FACH and thus reduces the number of users in CELL_FACH
state.
Establishment cause
RNC
RadioAccessService
supportEstabCauseAndTrafficVolInd
In 3GPP 25.331 Rel.5, the Establishment Cause has been added to the CELL UPDATE message. This can be used
by the RNC to determine the target RRC state, i. e. in case of Originating Conversational Call, the RNC can move
the UE to CELL_DCH directly (in the current solution, the call setup is done on Cell_FACH for PS I/B) and
therefore improves the call establishment success rate.
RadioAccessService
supportEstabCauseAndTrafficVolInd
In 3GPP TS 25.331 Rel-6, an optional IE which contains a TVI has been added to the Cell Update message. This IE
is set to TRUE when the amount of data in the UE buffer is above a configured threshold, while the absence of this
IE means the amount of data is lower. This can be used by the RNC to decide between CELL_DCH or CELL_FACH
when traffic resumes on a PS RAB. When the RNC receives a Cell Update message with TVI set to TRUE then it
moves the UE to CELL_DCH directly.
The Traffic Volume Indicator is set to TRUE when the criteria for event based traffic volume measurement reporting
is fulfilled, while the absence of this IE means the criterion is not fulfilled.
RA:
PS Interactive
alling e.g. IMS
high layer sign psSigAoTimerCoeff
P signalling
signalling, VoI (RadioAccessService)
"
t to "signalling
The SI IE is se
The second sub-feature provides configuration capability that permits operators to activate/deactivate the Always
On feature for UE having Interactive RAB, with the Signaling Indication (SI) set to signaling. The operator will also
be able to set different AO timers value for PS Interactive Signaling RAB by setting a desired value for the
configurable parameter.
In the existing implementation, the Always-On feature is by default disabled if SI=signaling. The Signaling
Indication is set to signaling by the PS Core network to inform the UTRAN that a RAB carries high layer signaling
(IMS signaling, VoIP signaling, etc). Hereupon, the Interactive RAB with SI=signaling is supposed to carry
signaling used by end-party entities to setup/release for example the streaming bearer, the initial thought that
allowing AO on such a RAB could introduce latency in the streaming session setup.
However, recent feedbacks from different customers stated that on the other hand keeping all SI=signaling RAB in
CELL_DCH is a waste of resources. For streaming, it may be pertinent not to trigger transition to CELL_FACH. But
for many other IMS based applications or for proprietary applications using the Signaling Indicator, keeping the UE
in CELL_DCH would result in an expensive dimensioning for Node B baseband resources (number of CEs, HSPA
connections).
The AO timers for PS Interactive Signalling RAB will be set according to the regular AO timer multiplied by
"psSigAoTimerCoeff".
URA_PCH
Cell Reselection
CELL_FACH
New
URA
URA No data
Update to transmit
Admission control
Although no cell resources are allocated to a UE in URA_PCH, the RNC has to maintain the RRC and Iu connections,
to keep a UE context as well as to process the URA Update procedure.
Therefore, the RNC controls the maximum number of simultaneous UEs in URA_PCH and once the limit is reached, a
UE is moved to Idle mode instead.
Mobility
In URA_PCH state the location of a UE is known at UTRAN Registration Area (URA) level.
A URA is an area covered by a number of cell(s), which is only known by the UTRAN.
The UE performs a Cell Reselection and upon selecting a new UTRA cell belonging to a URA that does not match the
URA used by the UE, the UE moves to CELL_FACH state and initiates a URA Update towards the network.
After the URA Update procedure has been performed, the UE state is changed back to URA_PCH if neither the UE
nor the network has any more data to transmit.
CELL_PCH
Cell Reselection
New
CELL
CELL_FACH
No data
Cell Update to transmit
Admission control
Although no cell resources are allocated to a UE in CELL_PCH, the RNC has to maintain the RRC and Iu connections,
to keep a UE context as well as to process the Cell Update procedure.
Therefore, the RNC controls the maximum number of simultaneous UEs in CELL_PCH and once the limit is reached a
UE is moved to Idle mode instead.
Mobility
In CELL_PCH state the location of a UE is known at UTRA cell level.
The UE performs Cell Reselection and upon selecting a new UTRA cell, it moves to CELL_FACH state and initiates a
Cell Update procedure in the new cell.
After the Cell Update procedure has been performed, the UE state is changed back to CELL_PCH if neither the UE
nor the network has any more data to transmit.
Mobility over Iur
If as a result of the Cell Reselection process, a UE initiates a CELL UPDATE message in a cell being controlled by an
RNC (CRNC) different from the SRNC, then an Alcatel-Lucent CRNC releases the RRC connection, i.e. RRC
CONNECTION RELEASE is sent with cause Directed Signaling Connection Re-establishment. The UE will then re-
establish the RRC connection under the new RNC, what should be transparent to the subscriber since it was
inactive.
The same procedure applies if an Alcatel-lucent SRNC receives a Cell Update message from a UE that has re-
selected a cell controlled by another RNC vendor.
New
CELL
CELL_FACH
No data
Cell Update to transmit
Mobility
In CELL_FACH state, the location of a UE is known at cell level.
The UE performs Cell Reselection and upon selecting a new cell, it initiates a Cell Update procedure in the new cell
and stays in Cell_FACH state.
Mobility over Iur
If as a result of the Cell Reselection process, a UE initiates a CELL UPDATE message in a cell being controlled by an
RNC (CRNC) different from the SRNC, then an Alcatel-Lucent CRNC releases the RRC connection, i.e. RRC
CONNECTION RELEASE is sent with cause Directed Signaling Connection Re-establishment. The UE will then re-
establish the RRC connection under the new RNC, what should be transparent to the subscriber since it was
inactive.
The same procedure applies if an Alcatel-lucent SRNC receives a Cell Update message from a UE that has re-
selected a cell controlled by another RNC vendor.
isFachToDchEnhancementAllowed
(RadioAccessService)
From UA07, onwards isFachToDchEnhancementAllowed determines whether handling of cell reselection during
FACH to DCH transition is supported.
This scenario occurs during a transition of a PS RAB from Cell_FACH to Cell_DCH where the RNC has submitted an
RRC RB Setup or RRC RB Reconfiguration message to the UE and an RRC Cell Update message (Cell update cause =
cell reselection or re-entered service area) interrupts the transition.
To successfully transition the UE to Cell_DCH on the newly selected cell, the RNC will roll back the operations on the
old cell and attempt anew the transition to Cell_DCH on the new cell.
If the Cell_FACH to Cell_DCH transition is initiated on a cell with poor radio conditions, and at a certain time the UE
is no more reachable, and it manages to reach back to the network thru a Cell-Update message with cause cell
reselection or re-entered service area, the RNC will:
Give Cell update message priority over the ongoing Reconfiguration from FACH to DCH.
Cancel any changes ongoing for resources to previous cell
Continue with the transition from FACH to DCH on the new cell using the same triggering information on the
initial attempt.
This procedure is only attempted once, if the cell update fails, then all resources on new and old cell are released
URA_PCH
pchRrcStates = pchRrcStates =
CellPchEnabled UraAndCellPchEnabled
CELL_PCH
fachToCellPchTime cellPchToIdleTime
(Inactiver user) (No signaling trafficr: Paging, Cell Update)
nbOfCellUpdatesFromCellPchToUraPc
h procedures)
(Counter of Cell Update
CELL_FACH Idle
AO Step 2 (T2) Step 3 (T3)
fachToUraPchTimer UraPchToIdleTime
(Inactive user) (No signaling traffic :r Paging, Cell Update)
URA_PCH
(pchRrcStates = {uraPchEnabled,
URA_PCH uraPchAndCellPchEnabled})
activated
URA_PCH activated (only) (pchRrcStates = {uraPchEnabled)
URA_PCH & CELL _PCH (pchRrcStates =
activated
CELL _PCH {uraPchAndCellPchEnabled})
activated (pchRrcStates = {cellPchEnabled,
uraPchAndCellPchEnabled})
isAlwaysOnAllowed (DlRbSetConf)
isAlwaysOnAllowed (UlRbSetConf)
= disabled = releaseOnly
neither RB downsize
nor RB release = degraded2AlwaysOnOnly no RB downsize
based on user inactivity but RB release
based on user inactivity
AO Step 3
RB downsize
then RB release
based on user inactivity
The isAlwaysOnAllowed parameter in DlRbSetConf and UlRbSetConf determines the behavior of each Radio
Bearer when the always-on downsize is triggered. It can take the following values:
Degraded2AlwaysOnOnly means that the downsize is allowed and the target radio bearer is the one determined
by the parameters AlwaysOnDlRbSetDchId / AlwaysOnUlRbSetDchId in case of downsize in cell DCH State or
AlwaysOnDlRbSetFachId / AlwaysOnUlRbSetFachId in case of downsize in cell FACH State.
ReleaseOnly means that there will be no intermediate downsize for this Radio Bearer. The Radio Bearer will be
released when the release conditions are fulfilled.
Disabled means that the always-on feature is disabled for this Radio Bearer.
isAlwaysOnAllowed = ? pchRrcStates = ?
isAlwaysOnAllowed = ? pchRrcStates = ?
isAlwaysOnAllowed = ? pchRrcStates = ?
PchRrcStates = none
PchRrcStates = UraAndCellPchEnabled
uraPchToIdleTimer
URA_PCH
AO Step3
CS RAB CS RAB
setup release
Down CS+PS
CS+PS I/B Inactivity CS + PS I/B 0/0 Inactivity
(CELL_DCH
CS + PS I/B8/8
(CELL_DCH) PMM-idle
(Cell_DCH)
Up
PS traffic resuming
Multi RAB PS I/B R99
isAlwaysOnAllowed (PS RAB) = releaseOnly pchRrcStates = any value
PS I/B
(CELL_DCH
DCH, HS -DSCH or E -DCH ) CELL_FACH
Admission Control
UE is kept in in CELL_DCH until
CAC FACH succeeds
MaxNumberOfUserPerMac
C
AO downsize Failure
(CacOnFachParam)
Bucket Occupancy
trbEstThreshold
(CacOnFachParam
CELL_FACH )
URA1
URA2
URA1
URA3 URA1 cell
URA2 cell
URA3 cell
disabled
true
degraded2AlwaysOnOnly
false
releaseOnly
AlwaysOnTimer
timerT1, timerT1ForHsdpa, timerT1ForHsdpaAndEDch
timerT2, timerT2ForHsdpa, timerT2ForHsdpaAndEDch
fachToCellPchTimer, cellPchToIdleTimer
fachToUraPchTimer, uraPchToIdleTimer
The Radio Bearers used for the downsized state are provided in the AlwaysOnConf object, including the type of
downsize (Cell_DCH or Cell_FACH).
The list of user services that are eligible to Always On is given through the isAlwaysOnAllowed parameter in
DlUserService and UlUserService objects.
The isAlwaysOnAllowed parameter in DlRbSetConf and UlRbSetConf objects determines the behavior of each Radio
Bearer when the always-on downsize is triggered. It can take the following values:
degraded2AlwaysOnOnly means that the downsize is allowed and the target radio bearers are determined by
the parameters of the AlwaysOnConf object.
releaseOnlymeans that there is no intermediate downsize for this Radio Bearer. The Radio Bearer is released
when the release conditions are met.
Disabled means that the Always-On feature is disabled for this Radio Bearer.
Assumptions
UE is R99
isAlwaysOnAllowed (alwaysOnConf) = True
alwaysOnUlRbSetDchId (alwaysOnConf) = PS_8K
alwaysOnDlRbSetDchId (alwaysOnConf) = PS_8K
isAlwaysOnAllowed (PS_128K_IBxSRB_3_4K) = True
isAlwaysOnAllowed (CS_64KxPS_128K_IBxSRB_3_4K) = True
isAlwaysOnAllowed (PS_128K_IB) = degraded2AlwaysOnOnly
isAlwaysOnAllowed (CS_64K) = disabled
timerT1 (AlwaysOnTimer) = 15s
timerT2 (AlwaysOnTimer) = 20s
timerT2ForMultiRab (AlwaysOnTimer) = 20s
PchRrcStates (RadioAccessService) = UraAndCellPchEnabled
nbOfCellUpdatesFromCellPchToUraPch (RadioAccessService) = 3
fachToCellPchTimer (AlwaysOnTimer) = 20s
fachToUraPchTimer(AlwaysOnTimer) = 20s
cellPchToIdleTimer (AlwaysOnTimer) = 60mn
uraPchToIdleTimer (AlwaysOnTimer) = 60mn
PS I/B 128K
128kbps
CS 64/64
64kbps
0kbps
5s
isUlRbRateAdaptationAllowed
isDlRbRateAdaptationAllowed
(RadioAccessService)
RB Resizing (UL/DL)
Estimated
Traffic Monitoring (UL/DL) Throughput
(DL/UL)
32 64 128 384
RB Rate Adaptation is applicable to UL and DL Interactive and Background PS. It introduces RB rate
downsizing/upsizing based on user estimated average throughput.
RNC monitors DL and UL traffic and determines if the current RB bit rate needs to be downsized or upsized to
accurately match the actual traffic:
Downsizing
The RNC targets the bit rate as closely as possible to the estimated throughput.
Upsizing
Uplink: The RNC targets the bit rate immediately above the current bit rate (step-by-step upsize).
Downlink: The RNC targets any rate (multi-stages upsize), based on user throughput and RLC buffer
occupancy. The targeted RB bit rate should never exceed the Reference RB bit rate.
DL and UL rate adaptation are performed independently.
T0 T1 T2 T3 T4
time
raUnitPeriodTime
raNbOfSample
(DlRbRaTrafficMonitoring
)
Rate[0]=0 Rate[0]=0 Rate[0]=0 Rate[0]=N0/T Rate[0]=N1/T
Rate[1]=0 Rate[1]=0 Rate[1]=N0/T Rate[1]=N1/T Rate[1]=N2/T raUnitPeriodTime
Rate[2]=0 Rate[2]=N0/T Rate[2]=N1/T Rate[2]=N2/T Rate[2]=N3/T raNbOfSample
(UlRbRaTrafficMonitoring
)
K 1
1 1 K 1
R
K
Rate k S 2
K 1 k 0
Rate k R 2
k 0
Confidence Level
throughput Estimate raMaxConfidenceInt
RLC-SDU
raModifiedStudentCoef
throughput
(DlRbRaTrafficMonitoring
)
The traffic monitoring function consists in calculating the average throughput over a time window and estimating
the confidence level of the observed throughput.
The algorithm used is the same in DL and in UL. The average throughput is estimated at RLC level excluding
retransmissions and acknowledgements.
The algorithm first periodically computes the user throughput over a period of time T (raUnitPeriodOfTime) as Rate
=N/T where N is the number of RLC-SDU bits transmitted for the first time during T.
Traffic estimates are then based on a sliding window of size K (raNumberOfSample).
The estimation of the average throughput R and of the throughput variance S is derived over the last K samples
Rate[k], where each value R[k] corresponding to a throughput value calculated during a period of time T (see above
slide formulas corresponding to an example with K = 3).
The estimated throughput is supposed to follow a Student-t distribution with K degrees of freedom. The throughput
estimate is considered reliable if the probability of the real throughput being out of the interval of confidence is
smaller than a determined threshold (see above slide formulas).
When the throughput estimate is considered reliable, the RB rate adaptation resizing process is triggered, otherwise
no action is taken.
DL6
DL384 DL256 DL128 DL32
4
isDlRbRateAdaptationAllowedForThisDlRbSetConf
no downsize isThisRbRateAdaptationDlRbSetTargetAllowed
dlRbRateAdaptationDownsizeThreshold
DL256 (DlRbSetConf)
TDOWNDL384
MIN (Adapted RB, IRM RB)
downsize to DL256
DL128
TDOWNDL256 Allocated RB
DL iRM
downsize to DL128
DL64 Reference RB
TDOWNDL128
DL32 downsize to DL64
TDOWNDL64
downsize to DL32
time
The RB Rate Adaptation process can downsize an RB from the current RB rate down to any smaller RB (all
transitions towards a smaller RB are possible except to PS 8k, which is not eligible as target).
Based on Traffic Monitoring, the RNC takes the decision to downsize when the following criteria, which are
periodically checked, are verified:
The observed average throughput is lower than a defined threshold (dlRbRateAdaptationDownsizeThreshold).
The confidence level of the estimated average throughput is good enough to consider the observation as
reliable.
The RB adaptation process can downsize an RB from the current RB rate down to any RB with lower bit rate but the
allocated RB is always constrained by the iRM table selection.
no downsize isUlRbRateAdaptationAllowedForThisUlRbSetCo
nf
isThisRbRateAdaptationUlRbSetTargetAllowed
ulRbRateAdaptationDownsizeThreshold
(UlRbSetConf)
UL128
TDOWN384
downsize to UL128
UL64
TDOWN128
UL32 downsize to UL64
TDOWN64
downsize to UL32
time
The RB adaptation process can downsize a Radio Bearer from the current RB rate down to any smaller rate (all
transitions towards smaller RB are possible except to PS 8 kbps).
Based on Traffic Monitoring, the RNC takes the decision to downsize when the following criteria, which are
periodically checked, are verified:
the observed average throughput is lower than a defined threshold (ulRbRateAdaptationDownsizeThreshold).
the confidence level of the estimated average throughput is good enough to regard the observation as reliable.
Same criteria and mechanisms as for DL RB Rate Adaptation downsizing apply.
Multi-stages Upsize avoids successive reconfigurations intermediate bit rates in order to reach directly the most
suitable RB rate.
The RB adaptation process can upsize an RB from the current RB rate up to any RB with higher bit rate but the
allocated RB is always lower than or equal to the Reference RB and is constrained by the iRM table selection.
Based on Traffic Monitoring, the RNC takes the decision to upsize according to the following criteria, which are
periodically checked:
The observed average throughput is higher than a threshold (dlRbRateAdaptationUpsizeThreshold).
The confidence level of the estimated average throughput is good enough to consider the observation reliable.
RLC-SDU buffer occupancy (in %) is higher than a threshold (raSduQueueThreshold).
If the Multi-Step DL Upsize algorithm is activated
(dlRbRateAdaptationUpsizeAlgorithm = multiStageUpsize and not stepByStepUpsize)
then the RNC selects the target RB according to the DL RLC-SDU buffer occupancy. It compares the current value of
the RLC buffer occupancy (in bytes) to a threshold in order to find the highest RB for which the following condition
is met: RLC-SDU Buffer Occupancy (in bytes) raSduQueueThresholdBytes
If no RB higher than the current RB meets this condition, the upsize is not performed, it means that little data is
waiting for transmission.
no upsize isUlRbRateAdaptationAllowedForThisUlRbSetCon
f
isThisRbRateAdaptationUlRbSetTargetAllowed
ulRbRateAdaptationUpsizeThreshold
(UlRbSetConf)
UL128
upsize of UL128
TUP128
UL64
upsize of UL64 TUP64
UL32
upsize of UL32 TUP32
time
ul4AThreshold
(UlRbRaTrafficMonitori
ng)
measQtyAverageTim
e
(DchUlBoTrafVolMea
s)
ul4ATimeToTrigger rcMeasPendingTriggerTi
(UlRbRaTrafficMonitorin me
g) (DchUlBoTrafVolMeas)
UL DCH1 RB UL DCH2 RB
From UA07 onwards, UL upsizing based on UL throughput is enhanced by a mechanism based on UE Buffer
Occupancy.
Measurement is set up on the UE to monitor the RLC Buffer Occupancy (BO). The Transmit power of the UE is
also monitored as an additional measurement if enabled (isUeTxPowerOn4AAllowed=true).
The BO is monitored thru an average over measQtyAverageTime; as soon it goes above a given threshold
(ul4AThreshold) during a given period of time (ul4ATimeToTrigger), an Event4A is sent to the RNC.
Upon reception of a 4A Measurement Report from the UE, the RNC determines the selected data rate as a
function of the iRM algorithm, OAM maximum Step size ul4AMaxRateStep and the UE Tx power headroom
described in the following two slides.
The advantage of this type of measurement is a quicker reaction of RNC to resource allocation needs compared
to decision based on throughput variation but we may face some excessive usage of resources due to upsize to
the highest RB.
RNC ul4AMaxRateStep
(UlRbRaTrafficMonitoring)
Current UL RB
UL iRM
Reference RB
If the usage of UE Tx power is not enabled (isUEtxPowerOn4AAllowed Set to false), then the Upsize procedure
will target an RB based only on the threshold ul4AMaxRateStep.
This RB reconfiguration will be limited by the iRM table and reference RB rate.
UEtxPoweroffsetFromMax4A
(MeasurementConfClass)
Current UL RB
Headroom(dB) = PowerOffsetMax
2
- UEtxPoweroffsetFromMax4A
The aim of this algorithm is to assure that once the UETxPower is equal or greater than the
UEtxPoweroffsetFromMax4A, then Rate_TxHeadroom will be equal or smaller than Current_Rate, which means
no upsize will be taken.
If isUEtxPowerOn4AAllowed is set to true then after receiving the Ue MR, the RNC will:
Determine the Current_Rate and the Max_UL_rate from ul4AMaxRateStep.
PowerOffsetMax = min((UlUsPowerConf:maxAllowedUlTxPower, UEmax) - UE Transmitted
Calculate
Power (taken from MR IE).
Calculate the Headroom(dB) = PowerOffsetMax - UEtxPoweroffsetFromMax4A
Rate_TxHeadroom= [Current_Rate * 10^(Headroom (dB)/10) rounded down to the nearest supported rate]
Selected_Rate = Min (Rate_TxHeadroom, ul4AMaxRateStep)
Once this algorithm is finished, and if Headroom is Positive, a UL Upsize RB Reconfiguration is done.
boThresholdDlDch
(DchDlBoTrafVolMeas)
boAverageTimeDlDch
(DchDlBoTrafVolMeas)
boTimeToTriggerDlDch boPendingTriggerTimeDlDch
(DchDlBoTrafVolMeas) (DchDlBoTrafVolMeas)
DL DCH1 RB DL DCH2 RB
From UA07 onwards, DL upsizing based on DL throughput is enhanced by a mechanism based on RNC Buffer
Occupancy.
This DL buffer measurement is performed using an average over boAverageTimeDlDch. Setting the
boThresholdDlDch to 0 disables the Algorithm in DL.
The reconfiguration will be on a step-by-step upsize. So if the Current_DL_Rate is PS64, the uspize will trigger an
RB reconfiguration to DL PS 128 if allowed by OAM settings.
RNC
Current DL RB
Next to current RB
Buffer
Allocated DL RB MIN (Adapted RB, IRM RB)
DL iRM
Reference RB
The Buffer Occupancy (BO) measurements allow the upsize of RB on a step-by-step procedure for DL.
The activation flags of throughput RB rate adaptation dont need to be enabled, as this mechanism is independent of
previously described throughput mechanism.
Once the BO fulfils the threshold defined, the RB reconfiguration is triggered towards the next PS I/B Radio Bearer
available and as long as the iRM Table and Reference RB are respected.
rbRateAdaptationPingPongTimer ueTrafficVolumeInhibitTimer
(RadioAccessService) (RadioAccessService)
MR BO MR BO MR BO
UeTrafficVolumeInhibitTimer
In DL as in UL, there is the same anti ping-pong mechanism in order to avoid continuous RB reconfigurations in case
of abnormal traffic conditions.
The UETrafficVolumeInhibitTimer timer will be set after the last RB reconfiguration. Another RB Reconfiguration
can only take place after this timer expiry. This timer is not considered in case RB Reconfigurations triggered by
other causes.
Important:
Before UA07, rbRateAdaptationPingPongTimer was triggered whenever a RRA mechanism occurred (Buffer or
Traffic related).
From UA07 onwards, this timer is still triggered on any RRA but does not prevent the BO mechanisms, introduced in
UA07, from taking place. Only UeTrafficVolumeInhibitTimer, which is triggered only by a BO event, will be
verified before processing any additional BO event for RRA.
The
TheUL/4A
UL/4ATVM
TVM((Traffic
TrafficVolume
VolumeMeasurement)
Measurement)isisset
setup
up/released
/released
through
through a Measurement Control Message sent from the RNCtotothe
a Measurement Control Message sent from the RNC theUE
UEif:
if:
UE
UEhas
hasDCH
DCHininUplink
Uplinkand
andthe
theRB
RBin
inuse
useisisnot
notthe
theMax
Maxassigned
assignedby
bythe
the
RNC & the UE has no 4A already setup
RNC & the UE has no 4A already setup
UE
UEhas
hasDCH
DCHininUplink
Uplinkand
andthe
theRB
RBin
inuse
useisisthe
theMax
Maxassigned
assignedby
bythe
theRNC
RNC
and the UE has 4A already setup OR E-DCH in Uplink
and the UE has 4A already setup OR E-DCH in Uplink
The
TheDL
DLTVM
TVM((Traffic
TrafficVolume
VolumeMeasurement)
Measurement)isisset
setup
up/released
/released
through an internal measurement control if:
through an internal measurement control if:
UE
UEhas
hasDCH
DCHin
inDownlink
Downlink&&the
theDL
DLRate
Rateisisnot
notthe
theMax
Maxassigned
assignedby
bythe
the
CN
CN&&PS
PSI/B
I/BRAB present&&boThresholdDlDch
RABpresent boThresholdDlDchisisnon
nonzero
zero
DL
DLRate
Ratereaches
reachesthe
themaximum
maximumbit
bitrate
rateassigned
assignedby
bythe
theCN
CN
From UA07 onwards, activation of the isMpdpRbAdaptationAllowed flag allows the RRA to be executed over
multi-RAB PSs (multiple PDPs).
RA
S
UlRbSetConf
UlRbRaTrafficMonitoring
ul4AMaxRateStep
ul4AThreshold
DlRbSetConf
ul4ATimeToTrigger
DlRbRaTrafficMonitoring
DchDlBoTrafVolMeas
boAverageTimeDlDch
Dedicatedconf/MeasurementConfClass boPendingTriggerTimeDlDch
boThresholdDlDch
boTimeToTriggerDlDch
DchUlBoTrafVolMeas
measQtyAverageTime
isBOTriggerForRbAdaptationAllowed
isUEtxPowerOn4AAllowed ueTxPowerOffsetFromMax4A measQtyQty
ueTrafficVolumeInhibitTimer boHoldoffTimeInitialDlFachDc rcMeasPendingTriggerTime
h
1 7 55 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Management
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
isUeTxPowerOn4AAllowe
= True
d (RadioAccessService)
Current_Rate=64kbps
UE power class=3 (24dBm)
AllowedUlTxPower = 24dBm
ul4AMaxRateStep(RBsetconf64)=384
UEtxPoweroffsetFromMax4A=3dB
Calculate the next bit rate for the following current UE Tx Power reported value:
14.1dBm
22.1dBm
Intermediate RB flipFlopUpDowngradeTimer
(RadioAccessService)
Fallback RB Upgrade
irmSchedDowngradeTxcpMaxBitRate
(RadioAccessService)
NodeB
DL 384 kbit/s
DL 128 kbit/s
DL 64 kbit/s
The iRM Scheduling mechanism is based on two sequential procedures triggered to adapt user throughput when he
goes alternately through good and bad radio conditions:
iRM Scheduling Downgrade reduces bit rate when radio conditions are getting bad.
iRM Scheduling Upgrade increases bit rate when radio conditions are getting better.
iRM Scheduling Downgrade is based on Transmit Code Power: trigger is based on a measurement done by the Node
B. The dedicated measurement is performed on the primary cell and concerns the Downlink Transmitted Code
Power.
The trigger based on TxCP dedicated measurement can be applied if the primary cell is handled by the serving RNC
or on a DRNC since iRM Scheduling on TxCP is supported over Iur from UA05 release.
irmSchedDowngradeTxcpMaxBitRate is the parameter specifying the fallback RB bit rate in case of iRM
Scheduling downgrade.
The flipFlopUpDowngradeTimer parameter allows avoiding ping-pong phenomena between RB upgrade and
downgrade.
iRM Scheduling/ RB rate adaptation dependency:
Incase if RB rate adaptation is enabled for the service, after iRM scheduling downgrade, the service is flagged
as ineligible for rate adaptation upsize. When an event B is reported, the iRM scheduling upgrade is triggered,
so the service come back eligible for RB rate adaptation upsize. Hence bit rate upsize will not be performed
immediately by iRM scheduling but rather with RB rate adaptation algorithm if necessary.
The isIrmSchedulingOverIurAllowed parameter has to be set to true on both SRNC and DRNC for the support
of iRM Scheduling over Iur to be effective.
Moreover, the global activation flags for both SRNC and DRNC on the RadioAccessService object have to be set to
True: isIrmSchedDowngradeTxcpAllowed and isIrmSchedulingUpgradeAllowed.
- thresholDdelta
(DlIrmSchedDowngradeTxcp)
Primary Cell
Event A
Threshold
timeToTrigger (DlIrmSchedDowngradeTxcp)
For iRM Scheduling Downgrade based on TxCP, the detection of degradation in radio conditions relies on the
monitoring of the DL Transmitted Code Power (TxCP). The TxCP trigger is based on NBAP Dedicated Measurement
(type Event A) performed by the Node B handling the primary cell.
When the transmitted code power (TxCP) rises above a threshold (TxCP threshold) during the hysteresis time
(timeToTrigger), a Dedicated Measurement Report is sent by the Node B to the RNC (Event A).
Upgrade
64 384 128
Transmitted Code Power Event B1 Event B2
Report Report
pcpichPower + maxDlTxPower
Primary Cell
- (DlIrmSchedulingUpgrade)
highB1ThresholdDelta
Event B2
128 Threshold
mediumB2ThresholdDeltaOverHighB
1
(DlIrmSchedulingUpgrade) Event B1
384 Threshold
When a UE uses the RB assigned by IRM Scheduling downgrade, the RNC configures two types of NBAP dedicated
measurements by event B report for this UE on the primary cell.
So each time the primary cell changes, the RNC terminates measurements on the old primary cell and initiates
measurements on the new primary cell.
Cell color
RL Condition
iRM tables
Requested RB
iRM RB
OLS
Min Granted RB
Event Bx
Event Bx
Processing Power RB
Any user becomes eligible for iRM Scheduling upgrade as soon as it experiences an iRM Scheduling downgrade. This
means that after the RB reconfiguration bringing about the downgrade, the RNC configures and activates the NBAP
dedicated measurement report on the primary cell, so as to track the transmitted code power (see section 4).
On reception of the NBAP Dedicated Measurement Report, the SRNC executes the RAB matching function taking into
account that the Power RB (H or I), corresponding to the event reported (B1 or B2), will be the highest rate able to
be allocated to this mobile.
On reception of the NBAP Dedicated Measurement report, the RNC proceeds to the Synchronous Radio Link
Reconfiguration (SRLR) if the Granted RB is different from the current one. So the anti ping-pong timer
flipFlopUpDowngradeTimer is started.
This timer should allow slow ping-pong phenomena between upgrading and downgrading if observed. At its expiry,
a NBAP dedicated measurement can be initiated if in the meantime an iRM scheduling downgrade has been
performed for the mobile.
Intermediate RB
DlUserService
xxx_PS_384K_STR_yyy
Fallback RB Upgrade
xxx_PS_256K_STR_yyy
NodeB xxx_PS_128K_STR_yyy
DL 64 kbit/s
Since high bit rate RBs are radio resources consuming, enhanced RRM is required to optimize radio resources usage.
iRM Scheduling Downgrade
Downgrading - similar to I/B but RNC selects the fallback bit rate that is equal or immediately above the
GBR
Upgrading - similar to I/B but but RNC selects the fallback bit rate that is equal or immediately above the
GBR
Always-On and RB Rate Adaptation are not applicable to PS streaming RABs.
RNC
1..1
Iur RadioAccessService irmSchedDowngradeTxcpMaxBitRate SRNC
Threshold Threshold
defined:
1..30 1..*
defined:
as absolute
value in dBm
DlUserService DedicatedConf as relative
value (to Pmax)
at RNC / 1..15 in dB
DlUserService 1..1 PowerConfClass FDDCell
level at Cluster
IrmSchedulingDowngradeTransCodePower 1..40 (PowerConfClass
DlUsPowerConf ) / DlUserService
level
1..1
LowRbA DlIrmSchedDowngradeTxcp
thresholdTransCodePowerDefinitionParam thresholdDelta
timeToTrigger timeToTrigger
Used when primary cell is on a drift RNC Used when primary cell is on the
serving RNC
RNC
1..1
RadioAccessService
1..*
1..30
Iur DedicatedConf SRNC
DlUserService
Threshold 1..15 Threshold
defined: 0..1
FDDCell PowerConfClass defined:
IrmSchedulingUpgrade
as absolute 1..40
value in dBm as relative
DlUsPowerConf value (to Pmax)
at RNC / in dB
1..1 1..1
DlUserService
level HighRbB1 MediumRbB2 DlIrmSchedUpgrade at Cluster
(PowerConfClass
intermediateRate
highBitRate
deltaThresholdTransCodePower
highB1ThresholdDelta ) / DlUserService
thresholdTransCodePower highB1TimeToTrigger
timeToTrigger deltaTimeToTrigger mediumB2ThresholdDeltaOverHighB1 level
mediumB2TimeToTriggerOverHighB1
Used when primary cell is on a drift RNC Used when primary cell is
on the serving RNC
EventA
Network
User Initiated Initiated
Service ChangeService Change (NI-SCUDIF)
(UI-SCUDIF)
This is the case where one of 3G users (originator or terminator) manually invokes a change in call type for an in-
progress call, from VT to speech or from speech to VT. There are several possible reasons for this call change (e.g.
privacy, performance, cost).
In order to invoke this procedure, the UE shall send a MODIFY message containing the BC-IE to change to. This BC-
1 7 65
IE must be one
W-CDMA R99 of those
Algorithms already
Description negotiated at call setup.
COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
Call Management
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
The basic premise of NI-SCUDIF is that under certain conditions the network makes the determination that a service
change from VT to speech (or vice versa) is required, and invokes the signaling required to change the call type
from 64KUDI to speech (or speech to 64KUDI).
When the radio access network detects a lack of sufficient resources to sustain the multimedia RAB configuration,
RNC shall inform the visited MSC by sending a RANAP Modify Request message to the visited MSC. The visited MSC
shall then:
Initiate an In-Call Modification procedure to speech towards the UE it serves using the MODIFY message.
Invoke the service change procedure towards the remote side.
Trigger the RAB modification by sending a RANAP RAB Assignment with the RAB requested to be modified to
the RNC.
Event A
pcpichPower + Report
maxDlTxPowerPerOls_Gold_OLS
Event A
Threshold
timeToTrigger
Upgrade
Primary Cell
- triggerThresholdDelta
Event B
Threshold
Event B Timer
TimeToTrigger
A VT call downgraded to Speech call may be upgraded (back) to a VT call based on:
RNC
RadioAccessService
DedicatedConf
PowerConfClass
DlUsPowerConf
ServiceChangeTxcpVideoToSpeech ServiceChangeTxcpSpeechToVideo
ForThisUserService ForThisUserService
timeToTrigger timeToTrigger
triggerThresholdDelta triggerThresholdDelta
RNC
RadioAccessService
DlUserService
ServiceChangeTxcpVideoToSpeech ServiceChangeTxcpSpeechToVid
ForNeighbouringRnc eoForNeighbouringRnc
timeToTrigger timeToTrigger
triggerThresholdAbsolute triggerThresholdAbsolute
These parameters define the TxCP threshold absolute value and TTT SCUDIF Video Telephony to Speech (and vice
versa) for cells managed by the neighboring RNC.
RNC
RadioAccessService
serviceChangeUserInitiated
serviceChangeRncInitiated
isVtServiceSupportedOn2G
NeighbouringRNC DlUserService
isScudifAllowedForDrns isThisDlUserServiceEligibleToNiScudif
These parameters define the TxCP threshold absolute value and TTT SCUDIF Video Telephony to Speech (and vice
versa) for cells managed by the neighboring RNC.
isIrmPreemptionAllowedForGoldUsers
isIrmPreemptionAllowedForSilverUsers
Worst isIrmPreemptionAllowedForBronzeUsers
Black
normal2congestedPLCThreshold
congested2normalPLCThreshold
White normal2congestedCLCThreshold
congested2normalCLCThreshold
Power
Cell Color CEM Color normal2congestedDlCEMThreshold
Color
congested2normalDlCEMThreshold
(IrmPreemptionCacParams)
The purpose of iRM Preemption is to keep some resources available when the cell becomes loaded. iRM Preemption
is a reactive process performed when all other preventive congestion solutions are not sufficient to free OVSF codes
and power resources quickly enough to maintain sufficient accessibility to the network.
The iRM Pre-emption has the following characteristics:
Is
a preventive mechanism (does not react on access failures but try to anticipate the congestion to avoid
them).
Is only applicable to DCH and not to HSDPA/E-DCH.
Is only applicable to DL.
Consists in downgrading current users if their allocated bit rate is higher than the one they would get in a RED
cell with a Radio Link Red (DL).
Doesnot provide any upsize mechanism for downgraded users. They may retrieve the initial requested radio
bearer after any reconfiguration (CS release, CS establishment when a PS connection is on-going, iRM
Scheduling Upgrade, AO Upsize, etc.) except the AO Downsize and iRM Scheduling Downgrade procedures.
Is an asynchronous process (all users not downgraded at the same time when the cell is congested).
May be activated/de-activated by OLS.
Thus iRM Preemption completes the existing congestion prevention iRM RAB to RB Mapping feature by implementing
a reactive congestion management at the cell level.
Note: IRM pre-emption feature activation requires that parameters isIrmOnRlconditionAllowed and
isIrmOnCellColourAllowed set to TRUE.
normal2CongestedDlTLCThreshold
congested2NormalDlTLCThreshold
Black (IrmPreemptionIubTransportLoadParameters
)
White
Power
Color Cell Color CEM Color Worst
Cell 1
Cell N Black
+
Black Worst
White =
White
Active Set Color
Cell Color
The transition from a normal to a congested state is computed when one of the following thresholds is crossed:
normal2CongestedCLCThreshold (for codes)
normal2CongestedPLCThreshol (for power)
normal2CongestedDlCEMThreshol (for CEM load)
normal2CongestedDlTLCThreshold (for Iub)
In order to avoid any ping pong effect between the congested and normal states, due to strong variations in the
radio resources allocation, the hysteresis cycle relies on additional thresholds characterizing the congested to normal
transition through the parameters:
congested2NormalCLCThreshold (for codes)
congested2NormalPLCThreshold (for power)
congested2NormalDlCEMThreshol (for CEM load)
normal2CongestedDlTLCThreshold (for Iub)
In the case of soft handover, the active set color is derived from the color of each cell.
Link C
olo r iRM Preemption
Radio fallBackRbRate MIN Downgraded
RB Bit Rate (RB1)
+ Gold
iRMRbRate
OLS
Yes UE is candidate
UE RB Bite Rate > RB1 ?
for iRM Preemption
The target Radio Bearers for iRM Preemption are defined using:
the iRM Tables: iRMRbRate as a function of (DlRbSetConf, OLS)
fallBackRbRate as a function of DlRbSetConf
They correspond to the Radio Bearers UEs would have received if UEs were admitted as the Radio Link Condition
was Bad and the Cell Color was Red.
Therefore, users eligible to iRM preemption are users whose current DL Bit Rate is higher than the one defined for
bad radio conditions and loaded cells.
As iRM tables are constructed making use of A/R Priority, iRM Preemption offers the possibility to preempt users
according to their OLS level.
UE 1
UE 2
UE N
Black
White
time
AS Color
irmPreemptionColorCheckingTimer (IrmPreemption)
As soon as a downlink PS I/B radio bearer is allocated to a UE, the iRM preemption timer assigned to each eligible
mobile is armed. At each UE timer expiration, the cell preemption color is checked, and if the cell is congested, the
eligible user is preempted if the following condition based on the bit rate comparison is fulfilled:
If
current DL RB bit rate > iRM preemption downgraded RB bit rate
Then
preemption is processed and the downgrade is performed.
R99
1 Step / 2 Steps Eligible Services
Preemption
Emergency Call
Queuing
RAB Assignment HSDPA HSUPA Speech
allowed
RRC connection Video telephony Preemption
request CS Streaming Capability
Always-on Upsize PS Streaming Preemption
Radio Link Addition Queuing not Signaling PS Vulnerability
Inter-Frequency allowed Eligible Interactive Priority (*)
intra-RNC Radio Link procedures PS Interactive
setup (IMCTA
Incoming CAC, does not trigger
relocation PS Background
iMCTA Alarm)
preemption in the target RNC (*) + OLS at user level
RadioAccessService
Preemption
PreemptionCapabilityProcedureInfo
Pre-emption
preemptionCapabilityForRrcEstabCause Procedure
on CAC Failure
OtherThanEmergencyCall
RAB Assignment ON
When a CAC failure occurs during one of the following procedures, the procedure goes on either by processing a
failure case or is queued (see below the comment for each procedure).
Simultaneously, a pre-emption action may be launched in order to free resources in each congested cell.
The freed resource might be used by the call that triggered the Pre-emption or not as part of the best effort
algorithm implemented.
Note 1: an incoming relocation in the target RNC shall not trigger pre-emption. Queuing is forbidden for relocation.
Note 2: an iMCTA upon service reason shall not trigger pre-emption nor in the source cell nor the target cell.
RNC
NodeB
FDDCell
NodeB DL radio resources not
available NodeB
NodeB UL radio resources not CAC
isCellPreemptionNbapCacFailureAllowed available Failure
NodeB resources not available
isCellPreemptionCodePowerCacFailureAllowed RNC DL code resources not
RNC
available
CAC
RNC DL power resources not
Failure
available
The S-RNC may decide to launch preemption in a cell when it faces up a CAC failure during the following resource
allocation procedures:
NBAP procedures: NBAP Radio Link setup failure, NBAP Radio Link addition failure, NBAP Synchronized Radio
Link Reconfiguration failure, NBAP Radio Link Reconfiguration failure using one of the following NBAP failure
cause: DL radio resources not available, UL radio resources not available and Node B Resources
Unavailable
These resource allocation procedures concern DCH, E-DCH or HS-DSCH resources allocation.
Others functions as HSxPA fallback or iMCTA may also solve the CAC failure situation depending on the trigger
eligibility and the feature activation. The order of the procedures is the following:
1-HSxPA fallback 2-iMCTA 3-preemption.
The resources allocation requests done through RNSAP procedures are not eligible to preemption.
Iub and Iur resource allocation failures dont call the pre-emption function.
Note: The NBAP failure cause Node B Resources Unavailable identifies a resource allocation failure without
indication of the direction which may be downlink, uplink or downlink & uplink.
When a CAC failure occurs, the selection of the users to be pre-empted depends
on the failure cause. Some examples below:
(*) XXX stands for DCH or E-DCH, YYY stands for DCH or HS-DSCH
RNC
RadioAccessService
PreemptionAllowedInf
o
isPreemptionAllowedForHsdpa DCH ON
HS-DSCH ON / OFF
E-DCH ON / OFF
isPreemptionAllowedForEdch
Pre-emption Capability
Preemption Pre-emption Vulnerability
preemptionMode OR
0..5
PreemptionFrequency/FD
Dx
PreemptionPcPvServiceClass/Emergency PreemptionOmcrPcPvInfo
PreemptionPcPvServiceClass/Speech preemptionCapability
PreemptionPcPvServiceClass/VideoTelepho preemptionVulnerability
nyPreemptionPcPvServiceClass/PS Streaming
PreemptionServiceInfo
preemptionPriorityOfService
Video Telephony Y N 3
CS Streaming Y Y 4
PS Streaming Y Y 5
Signalling PS
Y * N 6
Interactive
PS Interactive N Y 7
PS Backgroung N Y 8
Bronze
Bronze
Bronze
Silver
Silver
Silver
Silver
Gold
Gold
Gold
CAC P-Priority P-Priority P-Priority P-Priority
Services Failure k
n n-1 k
- Priority +
1st Service Preemptabl 2nd Service Preemptabl User
Incomin filtering e services filtering e services filtering
g service list (1st) (CAC list (2nd)
(PV, Priority) (OLS)
Failure)
For same service priority and same OLS, the expected gain in terms of radio bit rate is used as the third
criterion
In each congested cell, a search of P-Services to be pre-empted carrying a traffic channel is processed by applying a
downgrade or/and a release. The preemption criteria for each P-Service are taken into account.
The preemption process ends when the expected resource quantity to be de-allocated is achieved or when there is
no P-Service to pre-empt anymore.
The selection induces the following steps:
TheRNC selects all P-Services which are vulnerable (P-Preemption Vulnerability equal to pre-emptable) with a
P-Service priority lower or equal to the P-Service requesting the pre-emption
According to the CAC failure cause the P-services which dont allow the de-allocation of the resource requested
are filtered:
When the CAC failure cause is for Node B reason, a specific filtering applies:
The P-Service candidate to be pre-empted is filtered if it has not the same transport type as the
service requested.
When the CAC failure cause is for code or power reason, all P-Services using HSDSCH or DL DCH
resources are candidate to be-pre-empted.
Then the RNC:
only keeps the users with an OLS lower or equal to the OLS of the user requesting the pre-emption if
they are using different P-Service(s).
only keeps the users with an OLS strictly lower than the OLS of the user requesting the pre-emption if if
they are using the same Pservice(s).
Mono-Step Multi-Step
preemptionDowngradeReleaseSte preemptionDowngradeReleaseSte
ps ps
(Preemption) (Preemption)
DCH HS-DSCH DCH
E-DCH
Any value
Then the RNC starts to preempt the user with the lowest OLS within the lowest service priority by applying first a
rate downgrading action if eligible.
When there are no more users of the lowest OLS, the RNC goes on with upper OLS.
Then when there is no more user to downgrade, a second step of pre-emption using a release (if allowed for the
service) may apply to the users not already selected for downgrading (lowest OLS first) and to the users ineligible to
the downgrading (see note 1).
When there are no more users in this Service priority, the RNC goes on with the upper Service.
Note 1: There is no selection order between downgraded users and users ineligible to downgrading.
A user is ineligible for downgrading due to the Service (example: speech), the GBR, the transport type. When pre-
emption is processed for a given CAC failure, a service eligible for downgrading then release may only be either
downgraded or released: When it is downgraded, it will be candidate to the release at the next pre-emption function
call.
Note 2: For a given priority and a given OLS, a downgrading or a release applies by the highest expected gain in
term of radio bit rate. So the pre-emption order is Service, OLS and then radio bit rate gain whatever the CAC failure
reason.
Note 3: A release action may apply to the following Services:
Services
which are never eligible to rate downgrading:
CS Speech, CSD, CS streaming, PS Conversational
Services
which may be eligible to release according to an OAM parameter:
PS streaming, PS Interactive, PS Background
A service established on HS-DSCH may only be released.
A service established on E-DCH may only be released.
RBs selected
for downgrading
RadioAccessService
Preemption
preemptionFloorBitRateInDownlink
preemptionFloorBitRateInUplink
Filtering on
Target bit rate
RB bit rate > preemptionFloorBitRateInXX
RBs selected
for downgrading RBs selected
Keep the Max bit rate RBs for downgrading
An uplink target rate and/or a downlink target rate are set by OAM for each service per OLS.
The RB and the requested target rates are given by the pre-emption function to the RAB matching function which
selects the bearers with the nearest and lower or equal rate to the target rate.
When the RABb matching selects a service configuration with a sum of rate higher than the previous service
configuration, the call is not modified (this use case may apply when the call has others services which are upgraded
by the Rab matching function).
For a given priority and a given OLS, a downgrading or a release applies by the highest expected gain in term of
radio bit rate. So the pre-emption order is Service, OLS and then radio bit rate gain whatever the CAC failure reason.
Yes
Preemption
PreemptionDeallocationIn
fo
preemptionDeallocationFactorInDownlink
[High,Low]
preemptionDeallocationFactorInUplink
[High,Low]
preemptionDeallocationThresholdInDownlink
preemptionDeallocationThresholdInUplink
An estimation of the resource needed has to be done in the following CAC failure cases:
NodeB: DCH resource allocation failure
Code: DCH or HS-DSCH resource allocation failure
Power: DCH or HS-DSCH resource allocation failure
When the CAC failure occurs at Node B level for HS-DSCH or E-DCH resource allocation, a one to one release action is
processed (i.e. a P-Service established on HS-DSCH or E-DCH is released) without taken into account the resource quantity used.
In order to have a common estimator of all resources to be de-allocated, whatever where the CAC failure occurs, the estimation
of resources needed is based on:
the current radio bit rates if the CAC failure concerns DCH resources allocation onNode B or RNC sides (power and code).
The resource quantity to be de-allocated in a congested cell is based on the sum of radio estimated downlink rates needed
to establish all P-Services of the call in this cell.
the fair bit rate if the CAC failure concerns HSDPA resources allocation at RNC side (power and code). The resource quantity
to be de-allocated in a congested cell is based on the sum of fair bit rates needed to establish all P-Services of the call in
this cell. The fair bit rate of a P-Service is calculated by the fair sharing function and is either the GBR, the MinBR (defined
by OAM) if non null or the minHsDschReservationForCac (defined by OAM).
The resource de-allocation calculation uses a de-allocation factor (defined by OAM) to major the quantity of resources to be
freed. The unit is the kbits/sec. The resource to be freed is defined by the formula:
Ul resource rate to be freed= Ul Radio rate to be allocated * Ul de-allocationfactor
Dl resource rate to be freed= Dl Radio rate or fair bit rate to be allocated * Dl deallocation factor
The de-allocation factor depends on the resource quantity to be freed:
If the resource quantity <= threshold, the de-allocation factor used is the high factor value defined by OAM
If the resource quantity > threshold, the de-allocation factor used is the low factor value defined by OAM
Preemption
PriorityReallocationRegularReallocation
i ce
-Serv
r P
Pe
preemptionQueuingReallocationPriority = 0 or 1
resource allocation request
(PreemptionServiceInfo) 0
1
retry timer
max number of retries
preemptionQueuingReallocationTimer preemptionQueuingReallocationRetryMaxNumber
(PreemptionQueuingReallocation) (PreemptionQueuingReallocation)
T1 T2 T3 T4 T5 T6
Time
Enough resources are Enough resources are
available for S1 or S2 Available for S1 or S3
Assumptions
preemptionDeallocationFactorInDownlink [High,Low] = [200,120] (unit=%)
preemptionDeallocationThresholdInDownlink = 16 (kbps)
Question:
Estimate (in kbps) the quantity of DL resources to be de-allocated in order
to serve:
a PS Streaming call at 128 kbps downlink bit rate.
a CS AMR speech call at 12,2 kbps bit rate.
UL Cell Load DS
Iub
RNC
Iub DS Load
DL Tx CP
Differentiation per subscriber OLS
DL Power
Load
Max AMR
Rate
Re
IsAmrRateControlDuringTheCallAllowed
di
re
(RadioAccessService)
ct
io
n
isAmrRateControlOnULCellLoadAllowed
isAmrRateControlOnDlPowerLoadAllowed
isAmrRateControlOnIubDsLoadAllowed
(RadioAccessService)
(DlUserService)
Principle
Before release UA04.2, the bearer service support for AMR was restricted to mono-mode AMR, i.e. bearer service
support for AMR 12.2 and support of silent mode (aka SCR (VAD/DTX)) with use of version 1 of Iu UP.
The features 18717 AMR-NB multi-mode support and 30229 Iu UP SMpSDU V2 provide the bearer service
support with use of version 2 of Iu UP for a number of narrowband AMR speech services and the support of a
number of functions to allow core network Transcoder / Tandem Free mode of Operation (TrFO/TFO) speech calls.
Multimode AMR support offers:
Downlink capacity gain in terms of OVSF code resource. SF 256 can be used for AMR Low Rate configuration.
Downlink capacity gain in terms of radio resource. The required signal to interference ratio depends on the
required throughput. Therefore, a lower AMR throughput will require less power.
UL coverage gain. The required transmitted power will be reduced for lower AMR rate.
AMR Rate Control
In UA05.0, the RNC does not initiate Iu rate control towards the CN except in relocation scenarios.
If Iu UP used is version 2, the RNC may receive Iu rate control from the CN in case of TrFO/TFO. When that
happens, the RNC triggers an RRC TFC control indicating the new max allowed rate in the uplink.
SRB2 is used to carry the RRC TFC control message.
Max AMR
RNC UL: Rate
Rate control on UL cell load
UL Cell Color
Max Bit Rate is selected
delayBetweenAmrRateControls
During the Call
(RadioAccessService)
RNC
Rule: Eligible DL AMR Rates
The DL AMR rate is based on the observed Iub DS load, DL Power load and
DL TX Code Power and is selected according the following rule:
DL Requested MBR rate > DL Requested GBR
During the call, no Radio Bearer adaptation based on cell load (power, code) is performed. The allocated RB (speech
part) is kept unchanged. During the call, the RNC will only control the AMR rate. The principle is the following:
At the establishment, once the UE is eligible for AMR rate control, the RNC will perform:
In downlink, any rate control done. The maximum rate 12,2Kbps is given at establishment.
In uplink, a rate control is performed on the UL Cell Load criteria which give the maximum rate at call
establishment.
During the call, driven by rate control triggers, the RNC adjusts AMR rates by initiating RRC TFC control on
uplink and Iu UP rate control on downlink.
Procedure description:
Depending on the outcome of the IUB load table, the RNC initiates:
DL Rate control: Iu UP Rate Control frames sent to the core network.
UL rate control: RRC TFC control messages sent to the UE.
Whenever applicable, the rate is decreased step by step. This means that several Iu UP rate control / RRC TFC
control messages may be successively sent.
Iu b D S O L S M a x A M R ra te
tra n s p o rt
C o lo u r
G o ld e .g . 1 2 .2
G R E E N S ilv e r e .g . 1 2 .2
B ro n z e e .g . 7 .9 5
G o ld e .g . 1 2 .2
Table is played Y E L L O W S ilv e r e .g . 7 .9 5
B ro n z e e .g . 5 .9
After any mobility event, to help determine G o ld e .g . 5 .9
the new max UL/DL rate R E D S ilv e r e .g . 4 .7 5
B ro n z e e .g . 4 .7 5
In static state, determine possible UL/DL
rate change by checking the Active Set Iub
DS DL load color every configurable period.
DL: Only downgrade is possible
UL: downgrade & upgrade are possibles
A specific iRM table is introduced to determine the max allowed AMR rate based on the OLS and the Iub DS DL load
color of the Active Set.
The Iub DS load is only calculated for the DL assuming that in most of the cases the traffic over Iub DS is rather
symmetric or asymmetric services with higher rate in the DL.
In order to avoid any ping-pong effects due to the application of different criteria, the downlink rate upgrade is not
triggered based on Iub DS load color change. Only the trigger of TxCP measurement report is used to upgrade the
rate of an AMR call.
A specific UL iRM table is introduced to determine the max allowed AMR rate
based on the OLS and the UL cell load color of the cells in the active set. This
table is played as follows:
At admission to help After every primary cell In static state, determine possible
determine the max update, to help UL rate change by checking the UL
initial UL rate based on determine the new max cell load color of the Active Set
the UL cell load color UL rate based on the UL every configurable period (Cf.
of the cells of the cell load color of the cells parameter
Active Set of the Active Set delayBetweenAmrRateControls)
G o ld e .g . 1 2 .2
UL cell load color = max G R E E N S ilv e r e .g . 1 2 .2
(UL radio load color, UL B ro n z e e .g . 7 .9 5 Restriction:
CEM load color) G o ld e .g . 1 2 .2
Y E LL O W S ilv e r e .g . 7 .9 5 Multi-service: The AMR rate
With green < yellow < red B ro n z e e .g . 5 .9
control based on radio
G o ld e .g . 5 .9
R E D S ilv e r e .g . 4 .7 5 criteria (UL Cell Load) does
The UL Cell Load B ro n z e e .g . 4 .7 5
not apply for multi-service
does not take into
account the load
generated by the E- Applicable to UL only
DCH traffic. downgrade & upgrade are possibles
T o ta l D L T x O L S M a x A M R ra te
P o w e r lo a d
C o lo r
The DL Total Tx G R E E N
G o ld
S ilv e r
e
e
.g . 1 2 .2
.g . 1 2 .2
Restriction for Multi-service
Power does not B ro n z e e .g . 7 .9 5
with PS on DCH:
include the power of Y E L L O W
G o ld
S ilv e r
e
e
.g . 1 2 .2
.g . 7 .9 5 Multi-service : The RNC
the PA used by B ro n z e e .g . 5 .9
initiated AMR rate control
HSDPA R E D
G o ld
S ilv e r
e .g . 5 .9
e .g . 4 .7 5 based on radio criteria (Total
B ro n z e e .g . 4 .7 5
DL Tx Power load) does not
apply if PS RAB(s) mapped
on DCH (whereas it applies if
PS RAB(s) mapped on
Applicable to DL only HSDPA)
Only downgrade is possible
1 7 95 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Call Management
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
DL Tx CP 3. Crossing of the
1. Crossing of the Threshold triggers DL
Threshold triggers DL Rate Rate Change Request
Change Request after a
TimeToTrigger Iu UP Rate Control
RateIncTTT (new Max DL rate) to
Rate Inc Th the Core Network
NBAP 4. Rate Change
Event B
DL TX Min Power
5.9 12.2 t
Thresholds Rules for Downlink TxCP criteria
maxDlTxPower
= maxDlTxPowerPerOls[CurrentOls].maxDlTxPower Tim e To Trigger value determ ination for Dow nlink TxCP criteria
RateDecThreshold= pcpichPower + maxDlTxPower dlAmrRateDecreaseTxcpTrigger.thresholdDelta RateDecTTT = dlAm rRateDecreaseTxcpTrigger.tim eToTrigger
RateIncThreshold= pcpichPower + maxDlTxPower dlAmrRateIncreaseTxcpTrigger.thresholdDelta RateIncTTT = dlAm rRateIncreaseTxcpTrigger.tim eToTrigger
Where
dlAmrRateDecreaseTxcpTrigger.thresholdDelta <= dlAmrRateIncreaseTxcpTrigger.ThresholdDelta
At establishment of a speech bearer eligible to rate control, the RNC configures NBAP dedicated measurement
reporting on the primary cell in order to track the DL Transmitted Code Power.
Thus, the ALCATEL-LUCENT RNC can detect deterioration or improvement of radio conditions through NBAP
dedicated measurements on the transmitted code power.
DL Tx CP criteria: Dedicated measurements configuration
In order to detect a rate decrease / increase condition, the RNC configures dedicated measurement reporting with
the following characteristics.
Measurement quantity = transmitted code power (Radio Link)
Reporting mode = event triggered
Event =A/B
Threshold1 = RateDecThreshold
Threshold2 = RateIncThreshold
Timeto trigger = RateDecTTT / RateIncTTT. This time to trigger indicates the time during which the threshold
condition should be fulfilled before the UE sends the event to the RNC.
RNC
RadioAccessService
delayBetweenAmrRateControls
isAmrRateControlOnUlCellLoadAllowed
DLUserService UlIrmTableCellLoadConfClass
isAmrRateControlOnUlCellLoadAllowed
IrmAmrRateList
RadioAccessService
delayBetweenAmrRateControls
isAmrRateControlOnDlPowerLoadAllowed
DLUserService DlIrmTablePowerLoadConfClass
isAmrRateControlOnDlPowerLoadAllowed
IrmAmrRateList
RadioAccessService
delayBetweenAmrRateControls
isAmrRateControlOnIubDsLoadAllowed
DLUserService DlIrmTableIubDsLoadConfClass
isAmrRateControlOnIubDsLoadAllowed
IrmAmrRateList
RadioAccessService
delayBetweenAmrRateControls
DLUserService DedicatedConf
PowerConfClass
Iur Iub DlUsPowerConf
maxDlTxPowerPerOls
DlAmrRateDecreaseAbsoluteTxcpTrigger DlAmrRateDecreaseTxcpTrigger
Threshold thresholdDelta
timeToTrigger timeToTrigger
DlAmrRateIncreaseAbsoluteTxcpTrigger DlAmrRateIncreaseTxcpTrigger
Threshold thresholdDelta
timeToTrigger timeToTrigger
RNC and Node B support the modification of the following QoS parameters:
Streaming traffic class cant be changed but in a RAB combination that includes a streaming RAB, the I/B part
can be modified:
(CS) + PS Streaming + PS Interactive to/from (CS) + PS Streaming + PS Background
Traffic Class change may lead to AsConf change, thus an SRLR procedure and it may result in change of OLS,
HSDPA SPI, E-DCH SPI, MLP and minBR.
When the Core Network changes the Traffic Class, a new scheduling priority (SPI) is deduced by the RNC and is
provided to the HSDPA and HSUPA schedulers.
isPsRabModificationFullFeatureAllowed: RAB modification (full feature) activation flag which allows parameters
TC, MBR, ARP, THP and SI of a Ps Streaming or Ps I/B RAB to be modified
This UA07 feature is the extension of the UA06 feature PM31039 PS CN RAB Modification, available in the Korean
market. The isPsRabModificationAllowed parameter was introduced with the UA06 feature, and must be set
accordingly, so that the UA07 feature operates as expected.
UE Node B SRNC CN
Establishment of RAB (ID1, QoS1) and associated RBs on SRNC and Node B
Step 1
Establishment of Replacement AAL2 Bearers
The SGSN sends RAB Assignment Request for
ERQ RAB Id (ID1)
ECF If the RAB Id exists, this will be considered as
an attempt to invoke RAB Modification
(only applicable with ATM transport and bearers mapped on DCH) Otherwise it is an initial RAB Assignment
Qos 2 is the new QoS information, and may
DCH-FP RL DL/UL Synchronization include changes in one or more of the
parameters supported for modification
NBAP Radio Link Reconfiguration Commit
RAB Matching is executed, irrespectively of the
RRC Radio Bearer Reconfiguration value of the new RAB parameters, which is
required because the Granted RB may be
RRC Radio Bearer Reconfiguration Complete different from the Reference RB, possibly as a
result of the cell colour change
Release of Initial AAL2 Bearers
REL
RLC
MBR changes may apply to UL and DL for Interactive, Background and Streaming RABs.
For an MBR modification, the new value is taken into account to limit the user throughput in the RNC if the
source conformance mechanism is enabled (a reduced MBR can be applied in uplink and/or downlink to a user
generating a very high data volume leading to an unfair resource usage).
An ARP or THP modification may result in change of OLS, HSDPA SPI, E-DCH SPI, MLP and minBR.
For RABs mapped on DCH, an SRLR will occur if the new Granted bit rate is different from the current bit rate.
For RABs mapped on HS-DSCH or E-DCH, whenever an SPI occurs, the Node B must be notified of the new SPI
value (the UE is notified for E-DCH).
UE Node B SRNC CN
Establishment of RAB (ID1, QoS1) and associated RBs on SRNC and Node B
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
1 Network selection 7
1.1 PLMN selection 8
2 Cell selection in idle mode 9
2.1 Cell selection criteria 10
2.2 UE power compensation 11
3 Cell reselection in idle mode principles 12
3.1 General concept 13
3.2 Mobility in idle mode, Cell_FACH, Cell_PCH and URA_PCH 14
3.3 Idle mode neighboring list 15
3.4 Cell reselection eligibility criteria 16
3.5 High mobility detection 17
4 Cell reselection in idle mode without HCS 18
4.1 Measurements rules without HCS 19
4.2 Level ranking criterion without HCS 20
4.3 Quality ranking criterion without HCS 21
4.4 Cell ranking algorithm 22
4.5 Triggering algorithm 23
4.6 Exercise: multi-layer cell structure, HCS not used 24
5 Cell reselection in idle mode with HCS 26
5.1 Principles 27
5.2 Measurements rules with HCS in low mobility 28
5.3
185
HCS quality level threshold criterion 29
5.4
W-CDMAMeasurements rules
R99 Algorithms Description Mobility with HCS in high mobility
COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
in Reselection 30
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
5.5 Level and quality ranking criteria with HCS 31
5.6 HCS cell filtering in low mobility 32
5.7 HCS cell filtering in high mobility 33
5.8 Cell ranking algorithm 34
5.9 Triggering algorithm 35
5.10 Exercise: multi-layer cell structure, HCS used 36
6 Cell reselection in non-DCH connected mode 38
6.1 SIB 4 and SIB 12 broadcast 39
6.2 SIB3 & SIB11 parameters & objects 40
6.3 SIB4 parameters & objects 41
6.4 SIB12 parameters & objects UMTS FDD neighbor 42
6.5 SIB11 & SIB12 parameters & objects GSM neighbor 43
6.6 Exercises 44
6.6.1 Exercise 1: mono-layer topology 45
6.6.2 Exercise 2: bi-layer topology 46
6.6.3 Exercise 3: tri-layer topology 48
7 3G to 4G inter-RAT cell reselection 51
7.1 Support of SIB19 52
7.2 LTE neighboring definition 54
7.3 Measurement rules 55
7.4 LTE neighboring cell criteria 56
7.5 Priority parameters 58
7.6 Other parameters 59
8 Cell status and reservation 60
8.1 Cell status and reservation process 61
9 Location registration 62
9.1 LAC/RAC/SAC 63
PCH
-CC
/P
MIB
mobileCountryCode (RNC/Plmn)
mobileNetworkCode (RNC/Plmn)
mobileCountryCode (CsCoreNetworkAccess)
mobileNetworkCode
(CsCoreNetworkAccess)
mobileCountryCode (PsCoreNetworkAccess)
mobileNetworkCode
(PsCoreNetworkAccess)
mobileCountryCode (FDDCell)
MCC MNC MSIN mobileNetworkCode (FDDCell)
The different UMTS networks are identified uniquely in the world by the PLMN identifier composed of:
the Mobile Country Code (MCC)
the Mobile Network Code (MNC)
For one carrier, once the cell search procedure is completed, the UE has found the strongest cell and knows its
scrambling code. It is then possible to decode the Primary CCPCH.
The MNC and MCC are part of the system information broadcast on the P-CCPCH (in the Master Information Block
or MIB).
The UE then decodes the received PLMN identifiers and determines whether or not the PLMN is permitted according
to the lists of preferred and forbidden PLMN (stored in the UE).
If the PLMN is permitted and chosen, the cell selection parameters are used by the UE to determine which cell to
camp on.
FDDCell
H
PIC
P-C
CellSelectionInfo
l
C Ce l
FDD
qQualMin
qRxLevMin
S Criteria
Squal and SRxlev are the two quantities used for cell selection criteria.
If the criteria are fulfilled, the UE moves to the camped normally state where the following tasks will be performed:
Select and monitor the indicated PICH and PCH.
Monitor relevant System Information.
Perform measurements for the cell reselection evaluation procedure.
If the criteria are not fulfilled, the UE will attempt to camp on the strongest cell of any PLMN and enter in the
camped on any cell state where it can only obtain limited service (emergency calls). The following tasks will be
performed in the camped on any cell state:
Monitor relevant System Information.
Perform measurements for the cell reselection evaluation procedure.
Regularlyattempt to find a suitable cell trying all radio access technologies that are supported by the UE. If a
suitable cell is found, the cell selection process restarts.
RNC
Srxlev > 0
CPICH_RSCP > qRxLevMin + Pcompensation
RadioAccessService NodeB
Pcompensation
=
max (sibMaxAllowedUlTxPowerOnRach P_MAX, 0) DedicatedConf FDDCell
powerConfId
UE Class P_MAX
3 +24 dBm
P3 P3 P3 P3 P3 P3
P2 P2 P2
Copyright 1996 Northern Telecom
P1
Cell Reselection without HCS differs from UA04.2 only by the fact that High Mobility Detection is used in the
reselection triggering timer.
Cell Reselection with HCS was introduced in UA05.
Cell Filtering
using HCS Priority
Cell Reselection without HCS differs from UA04.2 only by the fact that High Mobility Detection is used in the
reselection triggering timer.
Cell Reselection with HCS was introduced in UA05.
CP CH
/ P-C
11
SIB
l
C el
vi n g
Se r
FDDCell Neighboring List
UMTSFddNeighbouringCell List
GsmNeighbouringCell List
An algorithm is used to declare and control correctly the list of neighboring cells in order to differentiate between
the configuration of idle mode/cell_FACH mode neighbors (sent in SIB11) and cell_DCH connected mode neighbors.
The idle mode/cell_FACH mode neighboring list is a subset of the cell_DCH connected mode neighboring list. The
differentiation is set through the sib11OrDchUsage parameter on each umtsFddNeighbouringCell. Note that this
parameter is only used when sib11NeighboringFddCellAlgo is set to manual.
It is recommended to set sib11OrDchUsage to sib11AndDch for less than 96 neighboring cells (either GSM or
FDD) if UA07 feature 34291 support of 64 UMTS neighbors is activated (via isEnhancedSib11Allowed). Else, it
should be limited to 48 neighboring cells.
sib11OrDchUsage must be set to dchUsage for the other remaining neighboring cells.
HCS usage: HCS requires its related information elements to be added to sib-11 thus leading to a shrinkage of the
space available for neighbor data. When the cell flag isHcsUsed is set to TRUE, this feature can roughly support up
to 87 cells (serving cell + 31 FDD neighboring Intra + 32 FDD neighboring Inter + 23 I-RAT neighboring cells) in
the SIB11.
UMTSFddNeighbouringCell
Max(MaxAllowedUlTxPower - P_max, 0)
(UmtsNeighbouringRelation)
GSMNeighbouringCell
Max(MaxAllowedUlTxPower - P_max, 0)
(GSMCell)
Srxlev >
0
Once the criteria for GSM or UTRAN/FDD neighboring cells tracking and measurements based on CPICH_Ec/No are
applied, a criteria S is applied on the measured GSM or FDD neighboring cells to assess their eligibility to cell
reselection.
To be eligible, the intra and inter-frequency FDD cells must fulfill criteria very similar to what is used for Cell
Selection. But this time these relationships shall be verified on the neighbor cell. This means the measurements are
made on this neighbor cell, and the parameters are those defined in the neighboring relationship.
To be eligible, the inter-system GSM cells must fulfill criteria shown in the above slide. Any cell (serving and any
GSM or UTRAN/FDD neighboring cell), which fulfills these criteria, will be part of the list of cells for ranking.
FDDCell
UE not in High Mobility state
CellSelectionInfo
Nb of Reselection >
nCr CrMgt
during tCrMax
nCr
tCrMax
UE enters High Mobility state tCrMaxHyst
High and low mobility UEs are distinguished thanks to the rate of Cell Reselection.
Srxlev
Intra-frequency No measurement
sSearchHcs
Intra-frequency
Inter-frequency
Inter-frequency
sHcsRatGsm
Intra-frequency
Inter-frequency Inter-frequency
GSM GSM
isHcsUsed =
sSearchHcs False
FDDCell
sHcsRatGsm
CellSelectionInfo
sIntraSearch
sInterSearch
sSearchRatGsm
UMTSFddNeighbouringCell
FDDCell
GSMNeighbouringCell
The cell level ranking criterion is used to rank the cells prior to the reselection. When HCS is not used, the behavior
is the same as before UA5.0.
The serving cell and all the neighboring cells being eligible (S criteria) are ranked accordingly to the RL criteria, as
defined below:
RLs = Qmeas,s + Qhysts; for the serving cell
RLn = Qmeas,n - Qoffsets,n; for any GSM or UTRAN/FDD neighboring cells
Where Qmeas is the CPICH_RSCP for the FDD case. For GSM cells, RxLev is used instead of CPICH RSCP in the
mapping function.
Where Qhysts is the qHyst1 parameter of the CellSelectionInfo object.
Where Qoffset is the qOffset1sn parameter of the GSMcell object.
FDDCell UMTSFddNeighbouringCell
qHyst2 (CellSelectionInfo)
qOffset2sn (UmtsNeighbouringRelation)
The cell quality ranking criterion is used to rank the cells prior to the reselection. When HCS is not used, the
behavior is the same as before UA5.0.
The serving cell and all the FDD neighboring cells being eligible (S criteria) are ranked accordingly to the RQ criteria,
as defined below:
RQs = Qmeas,s + Qhysts; for the serving cell
RQn = Qmeas,n - Qoffsets,n; for any UTRAN/FDD neighboring cells
Where Qmeas is the CPICH Ec/No measurement.
Where Qhysts is the qHyst2 parameter of the CellSelectionInfo object.
Where Qoffset is the qOffset2sn parameter of the UMTSFddNeighbouringCell object.
FDDCell GSMCell
Best cell is a ..?
CPICH_Ec/No CPICH_RSCP
qualMeas = ..?
Best GSMCell
is reselected
Best FDDCell
after Second Ranking
is reselected qualMeas (CellSelectionInfo)
tReselection
x speedDependScalingFactorTReselection GSMNeighbouringCell UMTSFddNeighbouringCell
x interRatScalingFactorTReselection Inter-freq
(UE in High Mobility)
tReselection
interFreqScalingFactorTReselection
interRatScalingFactorTReselection speedDependScalingFactorTReselection
(CellSelectionInfo) (CrMgt)
Neighb.GC
Macro GSMcells
Neighb.GA Neighb.GB
Neighb.MC
Macro FDDcells F2
Neighb.MA Neighb.MB
Neighbo.ma Serv.mc
Micro FDDcells F1 Neighbo.mb
Assumptions
UE class 3
qualMeas = qualMeasEcno
qQualmin (Serving and Neighboring Cell 3G) = - 16 dB
qRxLevMin (Serving and Neighboring Cell 3G) = - 115 dBm
qRxLevMin (Neighboring Cell 2G) = - 104 dBm
sibMaxAllowedUlTxPowerOnRach = 24 dBm
maxAllowedUlTxPower (Neighboring Cell 3G) = 24 dBm
maxAllowedUlTxPower (Neighboring Cell 2G) = 33 dBm
sIntraSearch = 8dB
sInterSearch = 6dB
sSearchRatGSM = 4dB
sSearchHcs = 0dB
sHcsRatGsm = 0dB
isHcsUsed = False
P3 P3 P3 P3
P1
hcsPrio (HcsCellParam)
hcsPrio (UmtsNeighbouringHcsCellParam)
hcsPrio (GsmHcsCellParam)
HCS priorities are broadcasted in SIB3 for the serving cell and SIB11 for the neighboring cells.
3GPP assumes that a cell with hcsPriority=7 has higher priority than another cell with hcsPriority=0.
Actually, one shall consider HCS priority in conjunction with HMD and opertors strategy, as depicted in:
When high-mobility state is detected, the UE will try to reselect a cell with a lower HCS priority.
When high-mobility state is NOT detected, the UE will try to reselect a cell with higher HCS priority.
HCS rules regarding priorities and HMD are presented in the following pages.
Srxlev
Intra-frequency Intra-frequency
Inter-frequency Inter-frequency
hcsPrion >= hcsPrios hcsPrion > hcsPrios
sSearchHcs
All Intra-frequency
All Inter-frequency
Srxlev
GSM
No measurement
hcsPrion >= hcsPrios
sHcsRatGsm
All GSM
sSearchRatGsm sLimitSearchRat
Squal
sLimitSearchRat
(HcsCellParam)
With isHcsUsed set to True, Use of HCS IE broadcasted in SIB11 is set to Used.
When HCS is used, measurement rules are based on the same thresholds as when HCS is not used (sIntraSearch,
sInterSearch, sSearchRatGsm, sSearchHcs and sHcsRatGsm) plus a new parameter sLimitSearchRat
which is broadcasted in SIB3.
When using HCS, the difference in the neighboring measurement rules relies on the filtering of the measured cells
based on high-mobility state detection.
When the UE is not in high-mobility state, measurements are triggered on higher priority neighboring cells.
NeighbouringCell
FDDCell hcsPriorityn = hcsPrioritys
hcsPrioritys
NeighbouringCell
hcsPriorityn <> hcsPrioritys
Hn = Qmeasn qHcsn
qHcs (HcsCellParam)
qHcs (UmtsNeighbouringHcsCellParam) H criterion for FDD or GSM Neighboring Cell
qHcs (GsmHcsCellParam)
of different HCS priority than Serving cell
1 8 29 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Mobility in Reselection
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
HCS introduces a new criterion, so-called Quality Level Threshold H criterion, which is used to determine whether
prioritized ranking according to hierarchical cell reselection shall apply.
Qmeass and Qmeasn = CPICH_Ec/N0 or CPICH_RSCP for serving cell and FDD neighboring cells based on
qualMeas parameter.
Qmeasn = BCCH_RSSI (or BCCH RxLev) for GSM neighboring cells.
Note: According to 3GPP 25.304, the real formula of Hn for a neighboring cell of hcspriority which is different from
hcsPriority of the serving cell is Hn = Qmeasn qHcsn temporaryOffsetn * W(t)
but Alcatel-Lucent recommends not to use temporaryOffsets parameters. So temporaryOffsetn is always set to 0.
UE in high-mobility state
Srxlev
Intra-frequency
Inter-frequency
hcsPrion <= hcsPrios
sSearchHcs
All Intra-frequency
All Inter-frequency
Srxlev
GSM
hcsPrion <= hcsPrios
sHcsRatGsm
All GSM
sSearchRatGsm sLimitSearchRat
Squal
When the UE is in high-mobility state, measurements are triggered on lower priority neighboring cells.
NeighbouringCell
FDDCell hcsPriorityn = hcsPrioritys
hcsPrioritys
NeighbouringCell
hcsPriorityn <> hcsPrioritys
Rn = Qmeasn qOffsetsns,n
Like when HCS is not used, both Level and Quality Ranking criteria can be used depending on the setting of the
qualMeas parameter.
Note: According to 3GPP 25.304, the real formula of Rn for a neighboring cell of hcspriority equal to hcsPriority of
the serving cell is Rn = Qmeasn qOffsetsnsn temporaryOffsetn * W(t)
but Alcatel-Lucent recommends not to use temporaryOffsets parameters. So temporaryOffsetn is always set to 0.
At least
UE is in no no
one cell has
High Mobility H criterion
state ? >=0 ?
yes yes
Once the H criterion has been computed for the serving cell and each neighboring cell, the UE performs ranking of
all cells that fulfill the S criterion, among:
when high-mobility state has NOT been detected (the higher priority, the smaller size):
All measured cells with the highest hcsPrio among the cells that fulfill H>=0
All measured cells, not considering hcsPrio levels, if no cell fulfills H>=0
When high-mobility state has been detected (the lower priority, the bigger size):
All measured cells with the highest hcsPrio that fulfil H>=0 and with a lower hcsPrio than the
serving cell
else:
All measured cells with the lowest hcsPrio that fulfil H>=0 and with a higher or equal hcsPrio than
the serving cell
else:
All measured cells without considering hcsPrio
empty list
Keep all neighboring FDD and GSM
cells
of the lowest hcsPriority as
not empty list Candidate Cells Ranking
candidates
among those having R criteria
1. hcsPriorityn >= hcsPrioritys
2. H criterion >=0
empty list
Keep all neighboring FDD and GSM
cells
as candidates
without ordering them
Once the H criterion has been computed for the serving cell and each neighboring cell, the UE performs ranking of
all cells that fulfill the S criterion, among:
when high-mobility state has NOT been detected (the higher priority, the smaller size):
All measured cells with the highest hcsPrio among the cells that fulfill H>=0
All measured cells, not considering hcsPrio levels, if no cell fulfills H>=0
When high-mobility state has been detected (the lower priority, the bigger size):
All measured cells with the highest hcsPrio that fulfil H>=0 and with a lower hcsPrio than the
serving cell
else:
All measured cells with the lowest hcsPrio that fulfil H>=0 and with a higher or equal hcsPrio than
the serving cell
else:
All measured cells without considering hcsPrio
Candidate
Neighboring First Ranking RL Same Ranking
FDD and GSM cells as without HCS
(CPICH_RSCP & RxLev)
for Cell
Reselection
FDDCell GSMCell
Best cell is a ..?
CPICH_Ec/No CPICH_RSCP
qualMeas = ..?
Best GSMCell
is reselected
Best FDDCell
after Second Ranking
is reselected
Same triggering
as without HCS
tReselection
x speedDependScalingFactorTReselection GSMNeighbouringCell UMTSFddNeighbouringCell
x interRatScalingFactorTReselection Inter-freq
(UE in High Mobility)
tReselection
interFreqScalingFactorTReselection
interRatScalingFactorTReselection speedDependScalingFactorTReselection
(CellSelectionInfo) (CrMgt)
Macro GSMcells
Neighb.GC
hcsPriority = 0
Neighb.GA Neighb.GB
The Cell Reselection Control feature enables a more flexible cell reselection control from the network in a
Hierarchical Cell Structure (HCS).
HCS layers management offers several solutions to manage the traffic demand and its associated noise rise.
For example, traffic may be split between the two or three layers in order to minimize the global noise rise, or it
may be split depending on the type of service used.
The later solution is conceivable if the microcell layer deployment aims at offering higher rate services continuously
within an area.
As a matter of fact, high data rate services require smaller cell sizes than low data rate services and therefore may
be continuously offered within an area only by the use of microcell sites (as illustrated on the above slide).
Moreover, since a microcell layer offer better indoor coverage quality than macro layers, this is well suited to high
data rate services, which are more likely to be indoor applications.
Assumptions
UE class 3
qualMeas = qualMeasEcno
qQualmin (Serving and Neighboring Cell 3G) = - 16 dB
qRxLevMin (Serving and Neighboring Cell 3G) = - 115 dBm
qRxLevMin (Neighboring Cell 2G) = - 104 dBm
sibMaxAllowedUlTxPowerOnRach = 24 dBm
maxAllowedUlTxPower (Neighboring Cell 3G) = 24 dBm
maxAllowedUlTxPower (Neighboring Cell 2G) = 33 dBm
sIntraSearch = 8dB
sInterSearch = 6dB
sSearchRatGSM = 4dB
sSearchHcs = 0dB
sHcsRatGsm = 0dB
isHcsUsed = True
sLimitSearchRat = 4dB
temporaryOffest = parameters not used
isDynamicSibAlgoWithSbAllowed
t
as (RadioAccessService)
e ll adc
C ro
rv ingers b
e t
: S me
I B 4 para
S n DCH DCH
o
l e c ti lls
- se
g Ce rs DCH DCH DCH
re rin ete
SIB1 SIB1
h bo ram +
1
+
1
eig pa t SIB1
: N tion dcas
DCH 1 DCH
2 DCH DCH
B 1 lec oa SIB1 + SIB1
SI e-se br DCH
1 DCH 1
DCH
r SIB1 + + SIB1
1 DCH Serving DCH 1
+ SIB1 FDDCell SIB1 +
DCH 1 1 DCH
DCH + + DCH
DCH SIB1 DCH
DCH 1+ DCH
SIB1 DCH SIB1
Cell_FACH 1 1
Cell_PCH + +
URA_PCH DCH DCH
DCH DCH DCH
DCH DCH
Prior to UA06.0, Cell Selection and Reselection information was only broadcast to UE in SIB3/SIB11 whatever mode
(Idle, URA_PCH, Cell_PCH and Cell_FACH), SIB3 (resp. SIB11) containing serving cell information (resp. neighboring
cell).
UA06.0 introduced the support of SIB4/SIB12 in order to have different Cell Selection and Reselection setting
between Idle and connected modes (Cell_PCH, URA_PCH and Cell_FACH).
Enabling SIB4 and SIB12 has a direct impact on the System Information size since many information (especially
about neighboring cells) are duplicated.
When all scheduling information cannot be coded in one MIB segment, SB1 Scheduling Block (SB) is used to support
the exceeding segments, as defined per 3GPP. isDynamicSibAlgoWithSbAllowed allows the use of such an SB1.
RNC
NodeB
tReselection
FddCell qualMeas
UmtsNeighbouring qHyst1
CellSelectionInfo qHyst2
qQualMin
qRxLevMin
UmtsFddNeighbouringCell sIntraSearch
sInterSearch
sSearchRatGsm
UmtsNeighbouringRelation CrMgt
qOffset1sn
qOffset2sn
qQualMin HcsCellParam
qRxLevMin hcsPrio
neighbouringCellOffset
qHcs
sLimitSearchRat
UmtsNeighbouringHcsCellParam
hcsPrio
qHcs
Idle Mode
Connected mode
Root = FddCell
CellSelectionInfo CellSelectionInfoConnectedMode
CellSelectionInfo.FachMeasOccasion CellSelectionInfoConnectedMode.FachMeasOccasion
CellSelectionInfo.CrMgt CellSelectionInfoConnectedMode.CrMgt
CellSelectionInfo.HcsCellParam CellSelectionInfoConnectedMode.HcsCellParam
No Equivalent CellSelectionInfoConnectedMode.CellSelectReselectInfoPchFach
CellAccessRestriction CellAccessRestrictionConnectedMode
CrMgt hcsPrio
qHcs R5/R6 UEs
HcsCellParam sLimitSearchRat
cellSelectReselectInfoPchFach
tReselectionPch
tReselectionFach
qHyst1Pch
cellAccessRestrictionConnectedMode qHyst1Fach
qHyst2Pch
qHyst2Fach
UmtsNeighbouring NodeB
Fddcell
UmtsFddNeighbouringCell
UmtsNeighbouringRelation
qOffset1sn
qOffset2sn
FddNeighCellSelectionInfoConnectedMode qQualMin
qRxLevMin
cellIndivOffset
UmtsNeighbouringHcsCellPara
hcsPrio
m
qHcs
UMTSNeighbouringRelation. UMTSNeighbouringRelation.FddNeighCellSelectionInfo
UmtsNeighbouringHcsTemporaryOffset ConnectedMode.UmtsNeighbouringHcsTemporaryOffs
et
GSMNeighbour NodeB
Fddcell
GsmNeighbouringCellqRxlevMin
GSMCell qOffset1sn
gsmCellindivOffset
GsmCellSelectionInfoConnMode qRxlevMin
qOffset1sn
gsmCellindivOffset
GsmgHcsCellParam hcsPrio
qHcs
UMTSNeighbouringRelation FddNeighCellSelectionInfoConnectedMode
maxAllowedUlTxPower cellIndivOffset
neighbouringCellOffset maxAllowedUlTxPower
qOffset1sn qOffset1sn
qOffset2sn qOffset2sn
qOffsetMbms qQualMin
qQualMin qRxLevMin
qRxLevMin
By Respecting the following strategy, fill in tables in next pages with the
appropriate values for the Mono, Bi and Tri Layer Topologies:
Fast reactivity
Lower threshold to delay the triggering inter freq or inter system measurements
1 8 44 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Mobility in Reselection
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
Neighboring Definition:
3G 3G 3G
2G
Only Sib4 is used
Subsidiary Question:
Give two values for SInterSearch when InterFreq or inter-Rat is preferred in the iMCTA
setting
Idle Mode Connected Mode Objective
qHyst2 2 dB Prevent ping pong between cells
on connected mode
sInterSearch 6 dB Will depends on iMCTA Alarm
priority setting
sSearchRatGsm 2 dB No 2G reselection, first fallback
to interfreq reselection
2G
Only Sib4 is used
F1 & F2 settings
Idle Mode Connected Mode Objective
qHyst2 2 dB Prevent ping pong between cells
on connected mode
sInterSearch 6 dB F1 F2 reselection allowed for
load sharing
sSearchRatGsm 2 dB No 2G reselection, first fallback
to interfreq reselection
F3 3G (R99) 3G (R99)
2G
Because of load sharing between F1 & F3 (R99 cells), the strategy is to allow
the UE to select the best cell (intra + inter-frequency F1+F3) in idle mode.
For connected mode (PCH / FACH), the strategy consists in keeping the UE on
its current layer, except for R99: with load sharing purpose, it is accepted that
the UE goes from F1 to F3 (and F3 to F1), but not on F2
F1 & F3 settings
Idle Mode Connected Mode Objective
qHyst2 2 dB Prevent ping pong between cells
on connected mode
sInterSearch 6 dB F1 F3 reselection allowed for
load sharing
sSearchRatGsm 2 dB No 2G reselection, first fallback
to interfreq reselection
F2 settings
Idle Mode Connected Mode Objective
qHyst2 2 dB Prevent ping pong between cells
on connected mode
sInterSearch 6 dB To delay the inter-freq mobility
sSearchRatGsm 2 dB No 2G reselection, first fallback
to inter-freq reselection
qOffset2sn 0 dB
F1 & F3
Better throughput when the PS call is established in (or moved to) the LTE
network thanks to the enhancement in SIB 19.
In UA08.1, the following enhancements on the 3G-4G cell reselection framework is introduced: Support up to 8 E-
UTRA frequency & priority list in SIB19 (changed from 2 to 8).
RNC
RNC NodeB
FDDCell
NodeB RadioAccessService
utraServingCellPriority
utraSearchPriority1
FDDCell utraSearchPriority2
isSib19Allowed utraThreshSLow
eUtraTargetDlCarrierFrequencyArfcn
sib19Enable eUtraTargetFrequencyMeasurementBandwidth
eUtraTargetFrequencyPriority
eUtraTargetFrequencyQrxlevmin
eUtraTargetFrequencyThreshxHigh
eUtraTargetFrequencyThreshxLow
eUtraTargetFrequencyDetection
Utran Release UA07.1.2 provides the support of cell reselection from UMTS towards LTE. This applies to LTE capable
UEs in UTRA RRC IDLE, CELL_PCH, URA_PCH modes.
Cell reselection in Idle mode from UTRA to E-UTRA RRC idle
Cell reselection in CELL_PCH from UTRA to E-UTRA RRC idle
Cell reselection in URA-PCH from UTRA to E-UTRA RRC idle
It uses a new inter-RAT cell selection algorithm based upon priority information, as introduced by 3GPP in Rel. 8.
The priority algorithm allows one to additionally prioritize RAT in order to favor a system technology such as E-
UTRAN (Rel. 8 feature).
MS (non E-UTRAN capable) which does not support the priority algorithm shall use the legacy algorithm.
The network shall provide priority information if E-UTRA cells or frequencies are included within the neighbor cell
list.
The priority information (for E-UTRA and UTRA cells) as well as other parameters are broadcast to the UE in a newly
introduced SIB 19. A Rel. 8 UE will be able to detect itself E-UTRA cells but it will read SIB11/12 as usual to identify
FDD and GSM cells.
The measurement rules below apply in Idle, URA_PCH, CELL_PCH states. Reselection from Cell_Fach states towards
LTE is not yet supported by 3GPP.
Srxlev and Squal of the serving FDDCell are compared respectively with the parameters utraSearchPriority1
and utraSearchPriority2:
If SrxlevServingCell <= utraSearchPriority1
or
SqualServingCell <= utraSearchPriority2,
the UE shall perform measurements of inter-RAT layers of lower or higher priority.
If SrxlevServingCell > utraSearchPriority1
and
SqualServingCell > utraSearchPriority2,
the UE may choose not to perform measurements of inter-RAT layers of lower priority, the UE shall search for
E-UTRA layers of higher priority.
Reselection criterion
when LTE cell priority is higher
Treselection Srxlev
nonServing cell
New cell reselected
Threshx,high
Time
If the priority of the E-UTRAN layer is higher than the current serving cell:
Cell reselection criterion: Srxlev[nonServingCell,x] of a cell on an evaluated higher absolute priority
layer is greater than Threshx,high during a time interval Treselection.
In this case the Target LTE Cell is selected regardless of the Rxlev and quality level on the serving FDD cell.
Reselection criterion
when LTE cell priority is higher
Srxlev Srxlev
Serving cell Treselection
nonServing cell
Sprioritysearch2
New cell reselected
Threshx,low
0
Threshserving,low Squal
Sprioritysearch1 Serving cell
Time
1 8 57 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Mobility in Reselection
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
If the priority of the E-UTRAN layer is lower than the current serving cell
Cell
reselection criterion: Srxlev[ServingCell] < Threshserving,low or Squal[ServingCell] < 0 and
the Srxlev[nonServingCell,x] of a cell on an evaluated lower absolute priority layer is greater than
Threshx,low during a time interval Treselection.
If more than one LTE cell meets the criteria, the UE shall reselect the cell with the highest SrxlevnonServingCell,x
among the cells meeting the criteria on the highest absolute priority layer.
RNC
NodeB
FDDCell
utraThreshSLow
eUtraTargetFrequencyPriority
eUtraTargetFrequencyQrxlevmin
eUtraTargetFrequencyThreshxHigh
eUtraTargetFrequencyThreshxLow
eUtraTargetFrequencyDetection
barred notBarred
other UE
allowed notAllowed reserved notReserved
All UEs are members of one out of ten randomly allocated mobile populations defined as Access Classes 0 to 9. The
population number is stored in the SIM. In addition, mobiles may be members of one or more out of 5 special
categories (Access Classes 11 to 15) also held in the SIM and allocated to specific high priority users as follows
(enumeration is not meant as a priority sequence):
Class 15 - PLMN Staff (VIP)
Class 14 - Emergency Services
Class 13 - Public Utilities (for example, water/gas suppliers)
Class 12 - Security Services
Class 11 - For Operator Use
An additional control bit known as "Access Class 10" is also signaled over the air interface to the UE. This indicates
whether or not network access for Emergency Calls is allowed for UEs with access classes 0 to 9 or without an IMSI.
Cell status and cell reservations are indicated with the Cell Access Restriction Information Element in the System
Information Message (SIB3) by means of four Information Elements:
Cell barred (IE type: "barred" or "not barred")
Cell Reserved for operator use (IE type: "reserved" or "not reserved")
Cell reserved for future extension (IE type: "reserved" or "not reserved")
The last element (Intra-frequency cell re-selection indicator) is conditioned by the value barred of the first element
(Cell barred)
LAC 2
RAC 1
SAC 2
SAC 1
Sac (FDDCell/CBSResource)
Broadcast Domain
1 8 63 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Mobility in Reselection
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
RNC RNC
Intra-NodeB SHO
Inter-RNC SHO Intra-RNC SHO (Softer)
Soft Handover (SHO) applies only to dedicated physical channels and refers to the case where more than one cell
has a link established with a UE. In this mode, the UE is connected to a set of cells known as the Active Set, where
it benefits from macro diversity.
Softer Handover is a special case of SHO where the cells communicating with the UE belong to the same Node B,
thus it can only be performed intra-RNC. The particularity of the softer handover comes from the fact that the radio
links coming from different cells of the Node B are combined together at the Node B level before being sent back to
the RNC.
In the Intra-RNC SHO case, the cells involved in the soft handover procedure belong to different Node Bs that are
connected to the same Serving RNC (that is, the RNC in charge of the RRC connection with the mobile). Radio Link
recombination is performed at the S-RNC level.
In the Inter-RNC SHO case, the cells of the active set are not all controlled by the S-RNC. This is where the notions
of Drift and Serving RNC come into play:
S-RNC is in charge of the RRC connection with the mobile.
D-RNC controls the Node B that does not belong to the S-RNC and for which a radio link needs to be
established with the mobile.
An Iur link between S-RNC and D-RNC is required to perform inter-RNC SHO.
From S-RNC and UTRAN transport perspectives, D-RNC acts as a router.
From UE and Core Network perspectives, the presence of D-RNC is transparent, that is, soft handover occurs as
usual.
FALSE
isEventTriggeredMeasAllowed
(FDDCell)
Measurement Control
(UETxPower 6A,6B)
From UA04.2, the mobility of a given UE is managed either in Periodical Mode or Full Event Triggered Mode.
The choice between these two modes is done by the RNC when the UE establishes a communication in CELL_DCH
state and is kept unchanged as long as the UE remains in CELL_DCH state. There is no switch between Periodical
Mode and Full Event Mode in CELL_DCH state, even when the Primary Radio Link is changed.
In the event-triggered reporting mode, the type of the triggered event becomes the main indication to compute the
Mobility decisions. The semantic of the received event indicates which decision has to be taken. In this mode, the
RNC provides the UE, in the RRC Measurement Control, with the means to compute the criteria to manage the
Mobility. The RNC has to perform the related action indicated by the received event.
The isEventTriggeredMeasAllowed parameter controls the activation of the Full Event Triggered feature on a per
cell basis. The reporting mode for the call is the one configured on the cell where the call is established, and is not
changed during the call duration (on CELL_DCH).
Warning: not ALL UE support E6A and E6B: MeasControl Failure cause: unsupported measurement shall be sent
by the UE but the call is kept
The above slide indicates in which order the various procedures are performed in the RNC when receiving a RRC
Measurement Report message from the mobile.
In this release, Alarm Hard Handover refers to the following mobility cases:
3G to 2G handover for CS
3G to 2G handover for PS
3G to 2G handover for CS+PS
3G to 3G inter-frequency inter-RNC handover
3G to 3G inter-frequency intra-RNC handover
NA2
NA2 NA1
NA3 A2 NA2
NA3 NA1 NA1
NA3
A3 A1
NA2 NA2
NA3 NA2 NA1
NA3
NA3 NA1
Prl Type1 or
typeOfCompoundingNeighbourListIntraFreq Type2
(FDDCell and NeighbouringRNC)
Two different CNL alogrithms are supported from UA06.0 that can be enabled/disabled using the
typeOfCompoundingNeighbourListIntraFreq parameter defined under the NeighbouringRnc and FDDCell
objects:
Prl stands for primary radio link and aims at disabling CNL at FDD cell level; the neighboring list only consists in
the primary cell neighborhood.
Type2 was the initial CNL algorithm introduced in UA4.1; the neighboring list classicaly consists in
concatenating Primary cell neighborhood then the best second legs until
maxCompoundingListSizeIntraFreq is reached.
Type1 is the newly introduced algorithm which aims at building intrafrequency neighbouring list based on
priorities and occurrence.
Basing the monitored set on the compound neighbor list rather than on the primary cell neighbors increases the
number of cells in the monitored set. So it is important to have a way to limit the size of the neighbor list, and
ensure that the monitored set comprises the best cells.
To achieve this, the algorithm consists in scanning the cells from the active set in decreasing order of CPICH Ec/N0
and adding their neighbors to the monitored set, until the number of cells in the list reaches the maximum size
allowed.
A cell is not added in the compounding list if it is already included in this list, so as not to have several instances of
the same cell in the list.
If maxNbOfMonitoredCellForNonShoCompoundList is set to zero, the compound list is only composed of the
primary cell and its neighborhood.
Type1 is the newly algorithm introduced in UA06.0 and is mainly based on:
neighbourCellPrio, between 0 and 62, defining for each FDD Cell a hierarchy within its neighborhood (0 is
the highest priority, 62 the lowest); note that two UMTSFddNeighbouringCell can NOT have the same
neighbourCellPrio.
cells in active set which are the cells from the ASET.
For each sponsoring cell, the RNC builds a neighboring cell list ordered by neighbourCellPrio. Then it builds the
final intra-frequency neighboring list by:
adding the sponsoring cells.
selecting the numOfPrimaryCellNeighbourIntraFreq first neighboring cells from Primary Cell neighboring
list.
and then performing the selection by number of occurrences.
In case of conflict, the RNC selects the neighboring cell whose sponsoring cell has the highest CPICH Ec/No, then
the one with the highest neighbourCellPrio until maxCompoundingListSizeIntraFreq is reached.
Iur case
In case a cell from the active set belongs to a Drift RNC, the Serving RNC may learn its intra-frequency
neighborhood using the information present in the RNSAP RadioLinkSetup/AdditionResponse message.
When type1 is selected, the RNC automatically assigns neighbourCellPrio by order of presence in the RNSAP
message (starting from 0).
It is recommended to set maxCompoundingListSizeIntraFreq to 32 to ensure that at least all cells of the
primary cell are sent. If maxCompoundingListSizeIntraFreq is 16 (for example) and the primary cell has 20
data filled neighbors, then only 16 neighbors will be sent to the UE.
RNC
NeighbouringRnc
RadioAccessService typeOfCompoundingNeighbourListIntraFreq {prl, type1, type2}
isCompoundingCellListActivated {True, False}
numOfPrimaryCellNeighbourIntraFreq [0..32]
maxCompoundingListSizeIntraFreq [16..32]
NodeB
FDDCell
typeOfCompoundingNeighbourListIntraFreq {prl, type1, type2}
numOfPrimaryCellNeighbourIntraFreq [0..32]
UMTSFddNeighbouringCell
neighbourCellPrio [0..62]
Cell4
Cell11
Cell12
Cell11 Cell14
Cell1 Cell31 Cell1
Cell15
Cell12 Cell21 Cell32 Cell41
Primary Cell Cell16
Cell2 Cell22 Cell33 Cell42
Cell17
Cell3 Cell23 Cell34 Cell43
neighborhood Build
Cell4 Cell51 Cell1 Cell44
Cell18
the list of cells
Cell13 Cell3 Cell2 Cell2
Cell19
for Intra-Freq
Cell14 Cell4 Cell4 Cell3
Cell51 Measurements
Cell15 Cell52 Cell35 Cell45
Cell21 in Cell_DCH
Cell22
Cell16 Cell53 Cell36 Cell46
Cell23
Cell17 Cell24 Cell37 Cell47
Cell52
Cell18 Cell25 Cell38 Cell48
Cell53
Cell19 Cell26 Cell39 Cell54
Cell3 Cell32
maxCompoundingListSizeIntraFreq = 26
Type1 is the newly algorithm introduced in UA06.0 and is mainly based on:
neighbourCellPrio, between 0 and 62, defining for each FDD Cell a hierarchy within its neighborhood (0 is
the highest priority, 62 the lowest); note that two UMTSFddNeighbouringCell can NOT have the same
neighbourCellPrio.
cells in active set which are the cells from the ASET.
For each sponsoring cell, the RNC builds a neighboring cell list ordered by neighbourCellPrio. Then it builds the
final intra-frequency neighboring list by:
adding the sponsoring cells.
selecting the numOfPrimaryCellNeighbourIntraFreq first neighboring cells from Primary Cell neighboring
list.
and then performing the selection by number of occurrences.
In case of conflict, the RNC selects the neighboring cell whose sponsoring cell has the highest CPICH Ec/No, then
the one with the highest neighbourCellPrio until maxCompoundingListSizeIntraFreq is reached.
Add relative
Delta
Drop Relative
Add Absolute Criterion
Delta
Add
Absolute
Threshold
Drop Be Add
st
Relative No Action Ce Relative
ll
Criterion Criterion
The active set update algorithm applies to all soft handover cases. Its purpose is to ensure that the strongest cells in
the UE environment will be part of its active set.
The algorithm is based on relative comparison between the best cell and each cell CPICH EC/N0 of the reported set.
Since UA04.1, the Active Set Update algorithm offers the possibility of using absolute thresholds for link addition and
link deletion criteria, providing additional tools to reducing call drop rates and improve the capacity of the network
from the perspective of radio power, code and RNC and Node B processing cost.
Note that absolute thresholds are optional and can be deactivated through parameters. Once activated, the criteria
for RL Addition/RL Deletion would be a logical OR between the relative and absolute criteria.
Cell Individual Offsets have also been supported since UA04.1. CIO is added to the measurements received from the
mobile before SHO conditions are evaluated, that is, Ec/No(i) + CellIndividualOffset(i) is compared to the add or
drop threshold (relative or absolute).
Note: Cell individual offset is taken into account by the RNC only if at least one of the flag enabling absolute
thresholds (add or drop) is set to true.
neighbouringCellOffset (UMTSNeighbouringRelation)
legDroppingDelta (SoftHoConf)
Drop Relative shoLinkAdditionAbsoluteThresholdEnable (SoftHoConf)
Delta Add Absolute Threshold
shoLinkAdditionCpichEcNoThreshold (ShoLinkAdditionParams)
shoLinkDeletionAbsoluteThresholdEnable (SoftHoConf)
shoLinkDeletionCpichEcNoThreshold (ShoLinkDeletionParams)
Drop Absolute Threshold
The RNC first identifies which is the best cell, that is, the cell with the highest CPICH EC/N0 of the reported set
(active set + monitored set).
Then for the cells belonging to the active set, the RNC applies the drop criteria:
Cells
not matching drop criteria are kept in the active set until the maximum number of cells in the active set is
reached.
Cells matching one of drop conditions are removed from the active set.
The drop criteria depend on the activation of the absolute add or drop thresholds.
If none of the cells of the current active set are eligible, the RNC keeps at least the best cell even if it does not meet
the criteria to be eligible.
The RNC identifies then the remaining cells (non-eligible cells) as cells to be removed from the active set. This
information will be transmitted in the active set update message.
When both relative and absolute criteria are used for the SHO Algorithm, it may happen that the relative and
absolute criteria trigger contradictory decisions for the same cell.
In order to avoid such a situation, it is necessary to add a supplementary check in order not to delete a radio link
which satisfies the link relative deletion threshold but is above the link absolute addition threshold.
neighbouringCellOffset (UMTSFddNeighbouringCell)
legAdditionDelta (SoftHoConf)
shoLinkAdditionAbsoluteThresholdEnable (SoftHoConf)
Add Absolute Threshold
shoLinkAdditionCpichEcNoThreshold (ShoLinkAdditionParams)
Keep Cell in
Monitored Set Add Relative shoLinkDeletionAbsoluteThresholdEnable (SoftHoConf)
Delta shoLinkDeletionCpichEcNoThreshold (ShoLinkDeletionParams)
maxActiveSetSize (UsHoConf)
Drop Absolute Threshold
The RNC first identifies which is the best cell, that is, the cell with the highest CPICH EC/N0 of the reported set
(active set + monitored set).
Then for the cells belonging to the monitored set, the RNC applies the add criteria:
Cellsmatching one of add delta & add absolute criteria are added in the active set until the maximum Active
Set size is reached.
Cells not matching any add criteria are ignored.
The addition criteria depend on the activation of the absolute add or drop thresholds.
When both relative and absolute criteria are used for the SHO Algorithm, it may happen that the relative and
absolute criteria trigger contradictory decisions for the same cell.
In order to avoid such a situation, it is necessary to add a supplementary check in order not to add a radio link
which satisfies the link relative addition threshold but is below the link absolute deletion threshold.
Event1A
Event1A
Event1A
maxActiveSetSize (UsHoConf)
amountRep1A
(FullEventRepCritShoMgtEvent1A)
CPICH_EC/No
Best Cell
New Cell
entering reporting range
timeToTrigger1A repInterval1A
(FullEventHOConfShoMgtEvent1A) (FullEventRepCritShoMgtEvent1A)
NA
10 LogM New CIO New W 10 Log M i (1 W ) 10 LogM Best ( R1a H 1a / 2)
i 1
cpichEcNoReportingRange1A
hysteresis1A
weight1A neighbouringCellOffset
(FullEventHOConfShoMgtEvent1A) (UMTSNeighbouringRelation)
typeOfCompoundingNeighbourListIntraFreq RNC
= TYPE 1
(FDDCell and NeighbouringRNC)
Event 1A TriggeringCondition =
detectedSetAndMonitoredSetCells detectedSetCellAddition (FddCell)= Enabled OR
OR (Automatic AND Cutoff List <> )
monitoredSetCells detectedSetCellAddition (FddCell) = Automatic AND Cutoff List =
From UA7 onwards, if the detectedSetCellAddition flag is set to ENABLED or AUTOMATIC, then the cells from the
Detected Set can be added to the Active Set.
If the combined intra frequency Cell Info List exceeds its maximum list size
maxCompoundingListSizeIntraFreq then cells with lower ranking priority need to be cut off from the list
which in consequence does no more represent the complete neighburhood of the active set.
Thecells having been cut off may reappear as detected set cells reported by the UE and will be re-identified by
the RNC. Those cells will no longer be ignored but be eligible to be added to the active set.
Thisis applicable when compounding algorithm Type1 is activated. Type2 does not support detected set cell
addition in the active set.
Compounding Neighboring List algo (typeOfCompoundingNeighbourListIntraFreq) must be set to
Type1
maxCompoundingListSizeIntraFreq MUST be set to any value strictly below 32.
If the detectedSetCellAddition flag is set to DISABLED the cells from Detected Set will not be used in the mobility
algorithms.
Event1B
Event1B
Event1B
Event1B
CPICH_EC/No amountRep1B
(FullEventRepCritShoMgtEvent1B)
Best Cell
timeToTrigger1B repInterval1B
(FullEventHOConfShoMgtEvent1B) (FullEventRepCritShoMgtEvent1B)
N
10 LogMOld CIOOld W 10 Log Mi (1 W ) 10 LogMBest (R1b H1b / 2)
A
i 1
cpichEcNoReportingRange1B
hysteresis1B
weight1B neighbouringCellOffset
(FullEventHOConfShoMgtEvent1B) (UMTSNeighbouringRelation)
Note: The above drawing shows an example assuming that CIO is set to 0 dB.
R99/R4 UEs are not able to use periodical measurement reporting after initial report.
Event1C
Event1C
Event1C
Event1C
maxActiveSetSize
(UsHoConf)
CPICH_EC/No amountRep1C
(FullEventRepCritShoMgtEvent1C)
AS Cell
New Cell
maxNbReportedCells1C
leaving reporting range InASCell (FullEventRepCritShoMgtEvent1C)
timeToTrigger1C repInterval1C
(FullEventHOConfShoMgtEvent1C) (FullEventRepCritShoMgtEvent1C)
hysteresis1C (FullEventHOConfShoMgtEvent1C)
neighbouringCellOffset
(UMTSNeighbouringRelation)
Event 1C is triggered when a new P-CPICH becomes better than an active P-CPICH.
Event 1C is used to replace an RL based on relative criteria when the Active Set is full.
Note: The above drawing shows an example assuming that the maximum AS size is set to 2 and that all the CIOs
are set to 0 dB.
= True
RNC
From UA7 onwards, if the isEnhanced1cHandlingAllowed flag is set to TRUE, the RNC will determine if the pilot
of any non-active cell to be added to the active set is weaker than the computed minimum signal strength (= pilot
strength of the strongest cell of the active set - cpichEcNoReportingRange1B)
If it is weaker, the candidate cell will not be added to the active set. A non-active cell will only be added if the
reported CPICH EcNo plus the Cell Individual Offset (CIO) is greater than threshold:
Where:
A1 = strongest active set cell
NA_x = non-active cell under consideration
R1b = cpichEcNoReportingRange1B
Event1E
maxActiveSetSize isEvent1EUsed
(UsHoConf) repeated! (FullEventHOConfShoMgt)
CPICH_EC/No
AS Cell
New Cell
entering reporting range
absolute threshold
leaving reporting range
maxNbReportedCells1E
(FullEventRepCritShoMgtEvent1E)
timeToTrigger1E (FullEventHOConfShoMgtEvent1E)
cpichEcNoReportingRange1E
hysteresis1E
(FullEventHOConfShoMgtEvent1E)
neighbouringCellOffset (UMTSNeighbouringRelation)
Event 1E is triggered when a new P-CPICH becomes better than an absolute threshold.
Event 1E is used to add an RL based on absolute criteria when the Active Set is not full.
In order to help the operator efficiently monitor its network and optimize its neighboring plan, it is possible to trigger
this event 1E based on both Detected Set and Monitored Set. However the cells from the Detected Set will not be
used in the mobility algorithms.
In order to achieve this, the isDetectedSetCellsAllowed parameter shall be set to True.
Event1F
Event 1F is not isEvent1FUsed
repeated ! (FullEventHOConfShoMgt)
CPICH_EC/No
AS Cell
timeToTrigger1F (FullEventHOConfShoMgtEvent1F)
cpichEcNoReportingRange1F
hysteresis1F
weight1F neighbouringCellOffset
(FullEventHOConfShoMgtEvent1F) (UMTSNeighbouringRelation)
Event 1F is triggered when an active P-CPICH becomes worse than an absolute threshold.
Event 1F is used to delete an RL based on absolute criteria.
shoAfterBlockingPhaseEnable
RNC
(RadioAccessService)
stored flushed
event1y
event1x
An SHO procedure may block other procedures or may be blocked by ongoing procedures.
On the one hand, a blocking phase consists in forbidding any procedure until an SHO procedure is completed, i.e.
after the RNC receives RRC ActiveSetUpdateComplete from the UE.
The shoAfterBlockingPhaseEnable parameter allows one to activate a mechanism to process events during such
a blocking phase:
When set to True, all the events received during the blocking phase are stored in a stack and processed once
on-going procedure is completed, thus preventing abusive call drops in case of unexpected add/delete.
When set to False, all these events are discarded, thus inhibiting mobility and degrading performances.
Event1A(cell1)
True False
Event1B(cell2)
Event1A(cell3)
Event1B(cell4)
1A(cell3)
1B(cell4) 1B(cell4)
1A(cell3) 1B(cell2)
Events 1A, 1C, 1D, 1E, 1J
Only lasts events can be stored
Event 1B
Only last OR several (*) events can be stored
Event 1F
Not stored (processed during blocking phase)
UE = R99
* Depends on UE release & isOne1bStorageAlloweed paramer
1 9 28 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Mobility in SHO
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
The RNC processes only the last event 1A, 1C, 1D, 1E or 1J received during the blocking phase. Appropriate action
is taken just after the completion of the other procedure. However, the RNC processes all events 1F received during
the blocking phase.
A specific handling has been provided for event 1B in order to adapt the RNC reaction to different UE behaviors. If
several events 1B have been received from the same UE during a blocking phase, the events related to a continuous
trigger condition (received from a UE with release later than R99) or the successive single samples representing
different trigger conditions (received from a R99 UE) will be repeated by the UE.
In order to avoid any loss of information on radio links to be deleted, the RNC should keep the received events until
the other procedure is completed; but queuing of all repeated events may lead to handling of outdated measured
results. They will be reactivated and processed sequentially when the blocking phase is terminated. If event 1A or
1C was stored in parallel then event 1B will be processed first.
However, in order to enhance the handling of event 1B by keeping only the last event 1B when multiple events 1B
are received in the blocking phase, the handling of event 1B received from a UE later than R99 depends on the
isOne1bStorageAllowed flag as follows:
the isOne1bStorageAllowed flag is set to TRUE, only the last event 1B received during a blocking phase
If
will be stored. It will be reactivated on completion of the other procedure. Compared to the behavior defined
above for R99 Ues, the processing will be in reverse order so that event 1B will be processed after event 1A or
1C if stored in parallel (in order to avoid a call drop if event B corresponds to a last RL).
the isOne1bStorageAllowed flag is set to FALSE, then the same behavior as defined above for R99 UEs
If
will apply.
primaryRlDelta (SoftHoConf)
The primary cell selection algorithm applies to all soft handover cases. The primary cell is used for monitored set
determination, but also as a pointer to mobility parameters. The primary cell selection algorithm is performed each
time a MEASUREMENT REPORT is received by the S-RNC.
The previous primary cell is compared with the candidate cell for primary minus a threshold defined PrimaryRlDelta.
The CPICH EC/N0 values used are the ones contained in the RRC MEASUREMENT_REPORT.
The Monitored Set should be updated each time the primary cell of active set changes. This is performed via the
RRC_MEASUREMENT_CONTROL message (with the measurement command set to modify) sent to the UE with the
cells to add and remove from the previous monitored set to form the new one.
The monitored set update usually follows the ACTIVE_SET_UPDATE message, but may also happen without any
ACTIVE_SET_UPDATE, when the active set content does not change, but, inside the active set, a cell becomes
strong enough to replace the primary cell.
Event1D
Event 1D is not
repeated !
CPICH_EC/No
timeToTrigger1D
(FullEventHOConfShoMgtEvent1D)
hysteresis1D (FullEventHOConfShoMgtEvent1D)
weight1D (FullEventHOConfShoMgtEvent1D)
neighbouringCellOffset (UMTSNeighbouringRelation)
The 1D event can be triggered by a cell in the Active Set for R5/R6 UEs or by a cell in the Active Set or in the
Monitored Set for R99/R4 UEs.
The primary cell determination is based on event 1D reception. Based on the reception of this event, the RNC stores
the new primary, and applies the current process used in case of change of primary cell.
Since events 1A and 1C are also configured, it is assumed that the new primary cell is already in the Active Set
when a 1D event is triggered. Typically, this will be ensured if the time to trigger 1D is greater than or equal to the
time to trigger events 1A or 1C. It should be noted that a monitored set cell that needs to be included in the active
set will trigger first a 1A event if its CPICH Ec/No is lower than the best cell in the Active set but entering in its
reporting range, or a 1C event if the Active Set is full and this cell is better than the worst in the Active Set.
An event 1D will typically be triggered by a cell better than the best in the active set. Therefore due to the triggering
conditions defined for these events, if the time to trigger a 1D event is greater than or equal to that for a 1A and 1C
event, the 1D will typically be triggered by a cell in the active set.
If the event 1D is triggered by a monitored cell, the RL will be added in the Active Set if not full.
If the Active Set is full and the 1D event is triggered by a monitored set cell, then the new best cell will be added in
the active set, replacing the worst active set cell.
In addition to the event 1D, a new primary cell will be defined if the current primary cell is removed due to reception
of RL deletion events.
Event 1D is not repeated by UE and therefore if lost during SRNS relocation it cannot be recovered. It is only
another event 1D for a different cell that will fix the issue of wrong primary cell. This is particularly bad for HSDPA
calls which follow the primary cell and do support SHO.
Event1AHoConfInSIB11
ShoLinkAdditionParams shoLinkAdditionCpichEcNoThreshold
ShoLinkDeletionParams shoLinkDeletionCpichEcNoThreshold
FullEventHoConfShoMgt
isEvent1EUsed timeToTrigger1A
isEvent1FUsed timeToTrigger1B
timeToTrigger1C
hysteresis1C timeToTrigger1D
FullEventHoConfShoMgtEvent1 hysteresis1D timeToTrigger1E
hysteresis1A timeToTrigger1F
hysteresis1B cpichEcNoReportingRange1A
hysteresis1E cpichEcNoReportingRange1B
hysteresis1F cpichEcNoReportingRange1E
cpichEcNoReportingRange1F
1 9 32 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Mobility in SHO
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
Measurement
Measurement Measurement
Reporting
ID Results
Quantity
Cell A
F1
Cell B Cell D
F1 F1
Cell C
SRNC F1
Measurement Id 1:
Cell A, B & C are candidates to SHO RNC1
(specified in the new IE)
1A, 1B, 1C, 1D, 1E and 1F (dedicated to RNC2
SHO)
Cell A, B & C could be candidates to HHO
in case IUR link is down
Measurement Id 16:
Cell D is candidate to HHO (specified in the new IE)
Event 1A is the only event provisioned as SHO can
NOT be performed without IUR
Intra-frequency Inter-RNC mobility without Iur may occur in the following situations:
Two different operators (PLMN) using the same frequency in the same area
National roaming agreements are needed
In a single PLMN, no Iur provisioned between 2 RNCs
E.g. due to IOT reasons between 2 different RNC manufacturers
In a single PLMN, the Iur interface is provisioned but is not in an enabled state.
Need to handle Intra-frequency Inter-RNC HHO
FRS 21302 was initially implemented in UA4.3K timeframe
FRS 33422 was duplicated for UA05.0
Code was implemented but commercially not supported
Not applicable to HSxPA calls
FRS 33814 has been created for UA06.0
Feature is fully supported
Also applicable to HSxPA calls
No influence of isIntraFreqInterRncHHOIurLinkDownAllowed
This flag is only used at RRC Measurement Report (MR) processing (cf. next slide)
Cells 1 2 3 4
DlUserService
RadioAccessService
isIntraFreqInterRncHHOAllowed = TRUE
IntraFreqTargetCellParams
isIntraFreqInterRncHhoOnIurLinkDownAllowed = TRUE
isHsdpaHhoWithMeasAllowed minimumCpichEcNoValueForHHO
RNC
isEdchHhoWithMeasAllowed minimumCpichRscpValueForHHO
UmtsNeighbouring UmtsNeighbouringRelation
neighbouringCellOffset
This best cell can NOT be added into the ASET as IUR is not
provisioned HHO is triggered
Cells 1 2 3 4
DlUserService
RadioAccessService
isIntraFreqInterRncHHOAllowed = TRUE
IntraFreqTargetCellParams
isHsdpaHhoWithMeasAllowed
isEdchHhoWithMeasAllowed minimumCpichEcNoValueForHHO
RNC
minimumCpichRscpValueForHHO
UmtsNeighbouring UmtsNeighbouringRelation
neighbouringCellOffset
RadioAccessService
isIntraFreqInterRncHHOAllowed
isIntraFreqInterRncHhoOnIurLinkDownAllowed
isHsdpaHhoWithMeasAllowed
isEdchHhoWithMeasAllowed DlUserService
DedicatedConf IntraFreqTargetCellParams
minimumCpichEcNoValueForHHO
minimumCpichRscpValueForHHO
UsHoConf
FullEventRepCritHhoMgt FullEventHoConfHhoMgt
FullEventRepCritEvent1AWithoutIur FullEventHOConfHhoMgtEvent1AWithoutIur
amountRep cpichEcNoReportingRange
maxNbReportedCells hysteresis
repInterval timeToTrigger
HLR
HLR GGSN PS GGSN
CS
UE UE
The SRNS Relocation procedure is used to move the RAN to CN connection point at the RNC from the source SRNC
to the target RNC. As a result of this procedure:
The Iu links are relocated from the Source RNC (S-RNC) to the Target RNC (T-RNC).
The target RNC becomes the SRNC.
The source RNC is released from the call.
The figure shows a PS domain SRNS relocation example where the S-SRNC and T-SRNC are connected to different
SGSNs (i.e. inter-SGSN Relocation).
Beforethe SRNS Relocation procedure, the UE is registered in the old SGSN. The source RNC is acting as the
serving RNC (SRNC).
After the SRNS Relocation procedure, the target RNC is acting as the serving RNC.
TheSRNS Relocation (UE not involved) is triggered when there are no links in the active set on the Source RNC
and all remaining links in the active set are on a single DRNC.
The SRNS relocation procedure implemented in Alcatel-Lucent UA05 UTRAN release is called Iur-based SRNS
relocation to differentiate it from the other variants of SRNS relocation defined by 3GPP 23.009.
The key benefits associated with Iur-SRNS relocation are as follows:
Reduction in the delay associated with the routing of the user plane flow via the Iur interface.
Capacity gain at RNC and Iur interface due to saving of Iur transmission resources
QoS improvement: better RRM, less inter-RNC HHO
RNC
NeighbouringRNC
RadioAccessService
is3Gto3GWithIurAllowed
timeToTrigger3Gto3GWithIur
isOutgoing3Gto3GWithIurAllowedForCsConversational
isOutgoing3Gto3GWithIurAllowedForCsCsStreaming
isOutgoing3Gto3GWithIurAllowedForCsPsInteractive
isOutgoing3Gto3GWithIurAllowedForCsPsBackground
isIncoming3Gto3GWithIurAllowedForCsConversational
isIncoming3Gto3GWithIurAllowedForCsStreaming
isIncoming3Gto3GWithIurAllowedForPsInteractive
isIncoming3Gto3GWithIurAllowedForPsBackground
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
RNC RNC
Node B
Node B Node B
Lte
GSMcell
2G to 3G HHO
Hard Handover gathers a set of handover procedures where all the old radio links are abandoned before the new
radio links are established (break before make).
Hard handovers are needed as soon as the UE needs to leave its serving UMTS carrier. It could be:
When the Iur interface is not present and the UE is changing RNC.
When moving to another UMTS carrier (inter / intra-RNC or inter / intra-PLMN): this case is defined as the
inter-frequency HHO.
When moving to another mode (e.g. TDD).
When moving to another technology (GSM, GPRS): this case is defined as the inter-RAT HHO.
When moving to UMTS from another technology (GSM, GPRS, LTE): this case is defined as incoming relocation
HHO.
When Soft Handover is not permitted (due to OAM constraint).
LTE layer
iMCTA alarm
trigger
GSM layer
CAC failure
trigger
FDD2 HSPA
Intelligent Multi-Carrier Traffic Allocation (iMCTA) allows the handover of a UE to another layer when in
Cell_DCH connected mode:
In case of coverage gap (iMCTA Alarm).
In case of RAB establishment failure. If call establishment is not possible because Call Admission Control (CAC)
fails establishing the traffic RAB (TRB) (iMCTA CAC).
After
Service Type change, Intra-frequency mobility or Always-On upsize based on load condition of
source/target cells, call type and UE/Cell capabilities (iMCTA Service).
iMCTA Alarm cannot be disabled as the objective is to save the call in case coverage is being lost. CAC and Service
can be enabled/disabled through OMC parameters. The recommended setting is to have iMCTA Alarm and CAC
enabled in order to improve Accessibility and Retainability.
CAC failure at
2. CAC Failure HSxPA
P1 PS call establishment
R99 C CS call
C iMCTA CAC
PNA
GSM
The iMCTA algorithm manages HHO handovers which may be triggered for several reasons:
save the call in case of loss of coverage: iMCTA triggered on HHO Alarm (case 1).
manage to establish a RAB that has experienced a CAC failure: iMCTA triggered on CAC failure (case 2).
optimize throughput by redirecting CS and PS calls to a preferred layer:
In case of a user service establishment or modification: iMCTA triggered on User Service (case 3).
In case of Primary Cell change: iMCTA triggered on Mobility Service (case 4).
For any iMCTA trigger cause, a load balancing can be applied in order to improve the overall network quality and
capacity:
For iMTCA triggers on User Service or Mobility Service:
Serving cell load can be checked to enable or disable the user redirection.
Target cell load can be checked to favor less loaded eligible cells.
For iMTCA triggers on HHO alarm or CAC failure:
Serving cell load is not checked.
Target cell load can be checked to filter out high loaded cell or favor less loaded eligible cells.
The iMCTA function only applies to call in connected mode (Cell DCH, E-DCH and HS-DSCH).
Call in Cell Fach (signaling or traffic), Cell PCH or URA PCH are not managed.
iMCTA configuration retrieval Neighboring cell searching & filtering for Radio
Access Selection for Measurements based HO;
Blind HO processing UA08: HSPA load criteria instead of DCH Load
Criteria for HSPA calls & load color BLACK for
DCH
Blind HHO execution UA08: limitation to 2 FDD carriers for IFREQ
Neighboring cells radio measurement
searching and filtering Pre-check for IFREQ measurements to avoid
triggering measurements for FDD overloaded
Configuration of CM & access layers
of UE measurements
iMCTA Algorithm
Mobility Service on
Primary Cell change
iMCTA algorithm
Stop timer
Event2D
Event2D
Event2F
CPICH_EC/No
CPICH_RSCP timerAlarmHoEvent2D
(FullEventHOConfHhoMgt)
Best Cell
The algorithm used to trigger alarm measurements while in Full Event mode is based on the following principles:
On reception of an event 2D (alarm condition), the RNC starts the timer timerAlarmHoEvent2D.
If,
at expiry of the timerAlarmHoEvent2D, no event 2F (indicating that the alarm conditions are no
more met) has been received, the alarm condition is confirmed, and iMCTA Alarm is invoked.
Atthe opposite, if an event 2F is received while the timer is still running, the timer is stopped, and the alarm
conditions are invalidated. iMCTA Alarm is not invoked.
/RSCP
DL: P-CPICH EcNo
isAlarmHHOUeTxPwrAllowed
UE Tx Power (RadioAccessService)
Primary Cell
UE Max Tx Power
UE Tx Power
iMCTA Alarm
ueTxPwrMaxThresholdOffset
(FulleventHOConfHHOMgtEvent6A)
ueTxPwrMaxThresholdOffset
timerAlarmHoEvent6A (FulleventHOConfHHOMgtEvent6B)
(FullEventHOConfHhoMgt)
6A
6B
Prior to UA07, alarm HHO was triggered by the RNC upon reception of event 2D thus only dependent on the DL
Radio quality.
UA07.1 onwards:
UL UE Tx Power can be taken into account to evaluate the radio quality and trigger the Alarm HHO.
The Feature is activated by isAlarmHHOUeTxPwrAllowed.
Benefit: avoid dropped calls when UE TX power is insufficient.
When the UE TX Power reported by the UE is above a threshold, an alarm Hard Handover (inter-frequency or
inter-system) is triggered in order to rescue the call on another carrier.
The actual Handover then follows the same procedures as for the Ec/No and RSCP triggers. The RNC will
configure an uplink trigger as new Alarm HHO criteria by using the internal measurements Events 6A/6B:
6A: The UE Tx power exceeds an absolute threshold.
6B: The UE Tx power falls below an absolute threshold.
A Alarm HHO uplink criterion is hit when the RNC received an Event 6A AND the alarm confirmation timer
elapses without receipt of Event 6B.
at expiry of the timerAlarmHoEvent6A, no event 6B (indicating that the alarm uplink conditions are no
If,
more met) has been received, the alarm condition is confirmed, and IMCTA Alarm is invoked.
Atthe opposite, if an event 6B is received while the timer is still running, the timer is stopped, and the alarm
conditions are invalidated. iMCTA Alarm is not invoked.
iMCTA Algorithm
Neighboring cells
NodeB searching and filtering
Configuration of CM &
of UE measurements
Primary FddIntelligentMultiCarrierTrafficAllocation
MS AS
mode
iMCTA
mode Alarm CAC User Mobility
(Primary Cell) Service Service
Once iMCTA trigger has been identified, Validity Checking aims at determining whether this specific iMCTA trigger is
enabled on the Primary cell, through the mode parameter that can have 4 different values:
AlarmOnly to have only iMCTA Alarm enabled
AlarmAndCac to have iMCTA Alarm and iMCTA CAC enabled
AlarmAndService to have iMCTA Alarm and iMCTA Service enabled
All to have all iMCTA triggers enabled
The mode parameter must be set at primary cell level, i.e.:
at FDD cell level when primary cell C-RNC is the Serving RNC, through the
FddIntelligentMultiCarrierTrafficAllocation object (so-called FddImcta).
at Neighbouring RNC level when primary cell C-RNC is a Drift RNC, through the
NeighbouringRncIntelligentMultiCarrierTrafficAllocation object (so-called NeighbouringRncFddImcta).
CN
DRNC neighbouringRncId NeighbouringRNC
AS Primary
FDDCell mode
MS Primary
iMCTA
mode Alarm CAC User Mobility
(Neighbour RNC) Service Service
iMCTA Service can only be processed when the Primary cell is located on the Serving RNC:
Therefore, iMCTA Service can only be activated on the FddImcta object.
mode must be never be set to AlarmAndService or All on the NeighbouringRncImcta object.
SRN isImctaOnAoUpsize RN
Primary
C C
FDDCell
userServiceSigToTrafficOnlyEnable RadioAccessServic
NodeB
FddImcta e
mobilityServiceForHsxpaEnable
mobilityServiceForNonHsxpaEnable
When the userServiceSigToTrafficOnlyEnable parameter is set to True, iMCTA User Service is only processed
consecutively to the establishment of the very first RAB, i.e. after transition from standalone SRB (DCH 3.4 or 13.6
kbps) to SRB+TRB (either CS or PS).
Therefore, this parameter allows one to limit the number of iMCTA User Service occurrences.
The mobilityServiceForHsxpaEnable parameter activates/deactivates the triggering of iMCTA Service when an
HSxPA-capable mobile performs Primary cell change.
The mobilityServiceForNonHsxpaEnable parameter activates/deactivates the triggering of iMCTA Service when
a non HSxPA-capable mobile performs Primary cell change.
FddIntelligentMultiCarrierTrafficAllocation/isImctaOnAoUpsize = TRUE
Iub Color
Worst
Cell Color
Yes
CEM Color
UL Cell Color < originatingCellColourThreshold
Radio Load
Color No
Worst
iMCTA Configuration Retrieval
CEM Color
ServiceSegmentationPriorityClass FDDCell
FDDImcta
ServiceForTrafficSegmentationPriority
Imcta
isServiceSegmentationTopPriority originatingCellColourThresholdconfclassref
OriginatingCellColourThresholdConfClass
OriginatingCellColourThresholdperservice
originatingCellColourThreshold
The cell load information is needed for iMCTA to state whether the Primary cell is eligible or not to iMCTA Service.
Comparing Primary cell iMCTA color with originatingCellColourThreshold is systematically performed by iMCTA
Service, for load balancing purposes but also for traffic segmentation (which is based on UE and NodeB HSxPA
capabilities). If operators priority is to perform traffic segmentation rather than load balancing,
originatingCellColourThreshold must be set to Green so as to systematically go further in the algorithm,
whatever Primary cell load.
Consequently, in UA05, it was impossible to have Load Balancing and Service Segmentation fully coexisting because
the former needs to have HHO only triggered when the cell is loaded whereas the latter needs to have HHO
triggered whatever cell load.
In UA06.0, such limitation was removed thanks to the introduction of isServiceSegmentationTopPriority, a new
flag defined per serviceType, which allows to bypass originatingCellColourThreshold during iMCTA Validity
checking, as presented hereafter.
NAMING CONVENTIONS
Cell fully compatible with UE capabilities
HSUPA cell and HSUPA UE
HSDPA cell and HSDPA UE
R99 cell and R99 UE
Cell partially compatible with UE capabilities
HSUPA (resp. HSDPA) cell and HSDPA (resp. HSUPA) UE
Cell NOT compatible with UE capabilities
HSxPA cell and R99 UE
R99 cell and HSxPA UE
Load threshold parameters used by iMCTA have been enhanced with the new
load color black
DL HSPA Load
isImctaUseHspaLoad
HSPA load is enabled if: isHspaCellLoadAllowedForImcra
(RadioAccessService)
iMCTA & iMCRA HSPA load are
activated
Fair sharing is activated isHsxpaR99ResourcesSharing
OnCellAllowed (FDDCell)
Rational: Feature UA07.1.2 81213 Load Balancing Between HSPA Carriers introduced a new HSDPA downlink load
criterion in order to manage load balancing of HSPA I/B traffic with or without MinBR (DL HSPA Load = maximum
(HSPA DL Power Colour, HSPA DL Codes Colour)); however, in UA07.1.2 it applies to IMCRA only.
iMCTA to use HSPA load Enhancement: this sub-feature introduces the HSPA load indicator usage for iMCTA
use. In parallel, load threshold parameters used by iMCTA have been enhanced with the new load color black.
iMCTA to use HSPA load is enabled if:
RadioAccessService/isImctauseHspaLoad = TRUE
isHspaCellLoadAllowedForImcra= TRUE & isHsxpaR99ResourcesSharingOnCellAllowed= TRUE
(PM81213);
Otherwise, iMCTA will use the aggregate DCH load.
In such case:
HSPA I/B calls will address the HSPA load criterion (DL HSPA Load = maximum (HSPA DL Power Color, HSPA DL
Codes Color)).
For DCH and HSPA GBR calls, the existing DCH load criterion is used further on.
For multi RAB calls (CS on DCH and PS on HSPA, or HSPA GBR and HSPA I/B), the RNC shall use the maximum
load color of DCH and HSPA load.
FDD1 P2 P2 P1 P1 P1 P1
FDD2 P3 P3 P3 P3 P3 P3
2G P1 PNA P2 P2 P2 P2
CAC priority Table
Cs Cs Cs Ps Ps CsSpeech
Service Type Speech Conversational Streaming Streaming Ib PlusOther
Access Type
FDD1 P2 P2 P1 P1 P1 P1
FDD2 P3 P3 P3 P3 P3 P3
2G P1 PNA P2 P2 P2 P2
Service priority Table
Cs Cs Cs Ps Ps CsSpeech
Speech Conversational Streaming Streaming Ib PlusOther
FDD2 P3 P1 P1 P1 P2 P1
FDD1 P2 P2 P2 P1 P1 P1
2G P1 PNA P3 P3 P3 P2
Once it has been checked that the invoked iMCTA trigger is enabled for the Primary cell, iMCTA Configuration Retrieval aims at
selecting the Priority Table to be used.
Priority Table is an important concept in iMCTA which must be associated with ServiceType (ST) and Priority.
An iMCTA Priority table defines for a specific iMCTA Trigger (Alarm, CAC or Service) and for each ST a priority level to be applied
on the different FDD and GSM neighboring cells.
When iMCTA Alarm is triggered, the algorithm retrieves the AlarmPriorityTable that is pointed either by FddImcta or
NeighbouringRncImcta object, depending on the Primary cell location.
From UA08 onwards, iMCTA Alarm & iMCTA CAC & iMCTA Service priority Tables are extended to contain 12 FDD
Frequencies.
Since iMCTA Service can only been triggered when Primary cell is located on the Serving RNC, the algorithm can retrieve up to 3
Priority tables that are pointed by the FddImcta object:
ServicePriorityGeneralTable
ServicePriorityTableForHsdpa (optional object)
ServicePriorityTableForHsupa (optional object)
The selection of the right Service Priority Table is based on UEs HSxPA-capabilities sent in the RRC Connection Setup Complete
message:
If UE is HSUPA-capable ServicePriorityTableForHsupa is retrieved if present
Otherwise ServicePriorityTableForHsdpa if present
Otherwise ServicePriorityGeneralTable
If UE is HSDPA-capable ServicePriorityTableForHsdpa is retrieved if present
Otherwise ServicePriorityGeneralTable
If UE is not HSDPA-capable nor HSUPA-capable ServicePriorityGeneralTable is retrieved
Unlike as for iMCTA Alarm and CAC, for iMCTA Service typically GSM and FDD access types should have different priority
because the Service trigger is not needed to save the call but to assign a better layer. Nevertheless, it is possible to
assign the same priority for GSM and FDD if desired.
OriginatingCellColourThresholdConfClass
RNC
0..11
NodeB
RadioAccessService OriginatingCellColourThresholdPerService 011
FDDCell
Imcta ServiceSegmentationPriorityClass 0..4
AlarmPriorityGeneralTableConfClass 0..11
CACPriorityGeneralTableConfClass 0..11
Frequency 1..14
ServicePriorityGeneralTableConfClass 0..11
priority
UmtsNeighbouringImcta
The originatingCellColourThreshold parameter can be configured per Service Type (ST). iMCTA
Alarm, CAC and Service Priority Tables are extended to contain 12 FDD Frequencies and 12 ConfClass:
New Object OriginatingCellColourThresholdConfClass 0..11;
New Object OriginatingCellColourThresholdPerService 011;
Updated Object Frequency 1..14 (FDD1-12; 2G; OtherFDD);
In UA08, AlarmPriorityTable may be composed of 12 FDD frequencies plus one input for 2G.
Service HO option
RNC
RadioAccessService
isChangeGsmIratHoCriterionAllowed
Imcta
serviceHoRanapIeEnable
The Service Handover option allows CS and PS Core Networks to inform UTRAN that GSM is preferred for this
service, through the optional serviceHO Information Element (IE) which is present in RANAP RAB Assignment
Request messages:
Basedon the 3 different values that ServiceHo IE can have (if present), the iMCTA algorithm may dynamically
change the GSM priority in the Priority table:
IE = should: GSM priority is overwritten with P0, i.e. GSM becomes the most preferred target Access.
IE = should not: GSM priority is overwritten with P6, i.e. GSM becomes the less preferred target
Access.
IE = shall not: GSM priority is overwritten with PNA, i.e. GSM is no more eligible to target Access.
The processing of this optional IE is not systematic and can be enabled/disabled through
serviceHoRanapIeEnable parameter.
should not is never taken into account when iMCTA Alarm or iMCTA CAC are triggered.
+
Primary cell is on SRNC
+
isImctaBlindHo
Blind HO activated
Blind HO (RadioAccessService)
+
blindHoType
=allowOnlyBho OR
{(blindHoType = blindHoType
minLevPrlBho) AND minLevPrlBho
(primary link level (FDDCell)
minLevPrlBho)}
In UA08, iMCTA may apply measurement-based handover, also known as Mobile Assisted HO (MAHO) or Blind HO,
also known as Database Assisted HO (DAHO). If no Twin cell is eligible for Blind HO, then iMCTA continues with
Measurement based HO if enabled.
Blind HO applies to:
AO upsize CAC & Service trigger (Cell/URA_PCH or Cell_FACH to Cell_DCH)
only Blind HO is enabled for AO upsize trigger, because there is no available measurement.
for AO upsize trigger, Blind HO is enabled independent of parameters RadioAccessService/
isImctaBlindHo, FddIntelligentMultiCarrierTrafficAllocation/blindHoType, and
FddIntelligentMultiCarrierTrafficAllocation/blindHoMode.
iMCTA Alarm, iMCTA CAC or iMCTA Service depending on configuration for parameter
FddIntelligentMultiCarrierTrafficAllocation/blindHoMode (alarm, AlarmCac, AlarmService, cac, cacService,
service, all) and only if the following criteria are met:
Primary cell is on SRNC (if the primary cell is on DRNC then only MAHO is applicable), AND
RadioAccessService/isImctaBlindHo is true, AND
FddIntelligentMultiCarrierTrafficAllocation/blindHoType is allowOnlyBho or minLevPrlBho and
primary link level is equal to or above the configured FddIntelligentMultiCarrierTrafficAllocation
/minLevPrlBho
Blind HO is processed in reference to a specific and parallel Twin cell list as did for RRC Re-direction while
Measurement based HO is in reference to the neighboring list.
Note: GSM twin list is new in UA08 whereas FDD twin list already existed in UA07.
Process filter based on priority tables: Similar as for neighbor cells a Twin cell is only considered if the
frequency is selected via the iMCTA priority table.
Process Load Based Target cell filtering: Since Measurement report processing does not apply for Blind HO, the
initial iMCTA phase for neighboring cell searching reuses some Measurement report processing concepts for Twin
Cells.
Process Target Twin Cell and RAT Selection: A single FDD Twin cell or GSM cell should be selected for Blind
HO.
If GSM and FDD Twin cells are applicable then the RNC shall select the RAT based on parameter is3GHandoverPreferred (true selects FDD;
false selects GSM);
In iMCTA CAC and iMCTA Alarm and AO upsize CAC trigger conditions, if FDD is applicable and two or more Twin cells are applicable then
iMCTA selects the target FDD cell based on the following criteria:
If UE is HSUPA capable, preference is given to HSUPA cells then HSDPA cells or if UE is HSDPA capable, preference is given to HSDPA cells.
Preference is given to the least loaded cell (Call type load or DCH load).
Preference is given to cell (s) belonging to the Carrier with the highest priority.
Round Robin based on downlink frequency (ARFCN).
In iMCTA Service and AO upsize Service trigger conditions, if FDD is applicable and two or more Twin cells are applicable then iMCTA
selects the target FDD cell based on the following criteria:
Cells belonging to the Carriers with the highest priority are kept.
Preference is given to the least loaded cell (Call type load or DCH load).
Round Robin based on downlink frequency (ARFCN).
If GSM is applicable and two or more Twin cells are applicable then iMCTA selects target GSM cell based the following criteria:
Preference is given to the least loaded cell.
Round Robin based on downlink frequency (ARFCN).
iMCTA Algorithm
Compounded DlUserService
Blind HHO execution
Neighboring
Neighboring cells
searching and filtering
Remove
neighboring cells isGsmCModeActivationAllowed
of Access or Frequency
marked as PNA isInterFreqCModeActivationAllowed
In Priority Table
Remove
neighboring cells
of GSM bands
not supported
by UE
Remove
neighboring cells
Inter-RAT and Inter-Freq
If CM disable for this DlUserService
and UE is mono-receiver
If Blind HO does not apply or if no twin cell is eligible for Blind HO, then iMCTA shall continue with the
Measurement-based HO process if enabled. Once the Priority table has been retrieved for the selected Service Type,
this function aims at filtering the inter-frequency and inter-RAT neighborhood provisioning of the FDD cell so as to
eventually keep the cells:
Belonging to an authorized PLMN
Compatible with the retrieved Priority Table, i.e. neighboring cells whose Access is PNA are removed
Compatible with GSM and UMTS bands supported by UE
Compatible with Compressed Mode capabilities:
UE capabilities sent in an RRC Connection Setup Complete message
CM activation flags per DlUserService
isGsmCModeActivationAllowed indicates if the compressed mode for GSM is allowed for this DlUserService.
isInterFreqCModeActivationAllowed indicates if the compressed mode for inter-frequency is allowed for this
DlUserService.
iMCTA Alarm
Compounded Generic
Neighboring Filtering
List
iMCTA CAC
Compounded Generic
Remove
Neighboring neighboring cells
Filtering
List of D-RNC
iMCTA Service
Remove
Compounded Generic neighboring cells
Neighboring Filtering of lower Priority
List than Primary Cell
As per 3GPP, neither the relocation procedure nor the RNSAP RL addition procedure supports the Transport Channel
addition/deletion. Therefore, for iMCTA CAC, every neighboring cell belonging to another RNC are removed.
For iMCTA Service, unlike iMCTA Alarm and iMCTA CAC, the neighboring cells whose priority is worse than the
Primary cell one are removed.
Feature principles
The ALU RNC will compute the Inter-Freq neighbor list if:
A new RRC Measurement Control message needs to be sent to the UE for inter-frequency measurements.
the primary cell changes or the active set is updated while an inter-frequency measurement is ongoing.
The ALU RNC will compute the Inter-RAT neighbour list if a new RRC Measurement Control message needs to be
sent to the UE for inter-RAT measurements.
The Compounding Neighbor List algorithm considers:
Occurrence of a cell within all neighborhoods.
Cell12
numOfPrimaryCellNeighbourInterFreq
Cell13
Sponsoring cells 1 to 4 are ordered by EcNo Cell14
Cell53
Neighboring cells
Cell54
Cell1 Cell2 Cell3 Cell4
with highest occurrence
Cell11 Cell21 Cell31 Cell41 Cell55
Cell31
Cell32
Common number
of occurrences maxCompoundingListSizeInterFreq
Cell12
Sponsoring cells 1 to 4 are ordered by EcNo numOfPrimaryCellNeighbourInterRAT
Cell13
Cell1 > Cell2 > Cell3 > Cell4 Cell14
Cell52
Cell53
Cell1 Cell2 Cell3 Cell4 Neighboring cells
Cell54
Cell11 Cell21 Cell31 Cell41 with highest occurrence
Cell55
Cell12 Cell22 Cell32 Cell42
Cell15
Cell13 Cell23 Cell33 Cell43
Cell16
Cell14 Cell51 Cell34 Cell44
Cell17
Cell15 Cell52 Cell35 Cell45
Cell18
Cell16 Cell53 Cell36 Cell46
Cell19
Cell17 Cell24 Cell37 Cell47
Cell21
Cell18 Cell25 Cell38 Cell48
Cell22
Cell19 Cell26 Cell39 Cell54
Cell31
Cell32
Common number
of occurrences maxCompoundingListSizeInterRAT
Two IFREQ:
It is a 3GPP limitation
The RNC sends to the UE max 2 carriers for iMCTA measurements if:
Measurement based HO (not Blind HO)
Feature activated
carrierPreSelectAlgorithm (FDDCell)
1. Least loaded cell from 1. Least loaded cell from 1. Biggest number of
Neighboring List Twin cells List neighbors
2. Highest Priority 2. Round Robin on DL 2. Round Robin on DL
3. Round Robin on DL frequency frequency
frequency
Rational:
3GPP 25.133 restricts the maximum number of carriers to be supported simultaneously by the UE to the current
carrier plus up to 2 additional FDD carriers, i.e. a total of 3 carriers. The UE behavior is undefined if more than 2
additional FDD carriers are to be measured. Tests with one UE type showed that arbitrarily some carriers were
ignored (not necessarily the carriers given first/last). QUALCOMM stated that the first two frequencies are selected.
For a deterministic behavior the RNC is required to restrict the number of additional (i.e. inter-frequency) FDD
carriers to be measured by a UE to 2.
Notes:
This feature is about inter-frequency (the UE can measure more than 2 cells in IFREQ).
This feature impacts the Neighboring cell searching and filtering part of the iMCTA algorithm.
is3GHandoverPreferred
NodeB UE
RRC Measurement Control
CM activation
+
Neighboring Cells
Inter-RAT
and/or
Inter-Freq
Once target Access has been selected, this function configures Inter-frequency or GSM measurements on Node B
and UE sides, by respectively sending the NBAP Compressed Mode Command and RRC Measurement Control
messages.
The alarm measurement results are reported in periodic mode.
In event mode, new measurements are configured to report Inter-Frequency/Inter-RAT measurements (Intra-
Frequency measurement are not impacted and still reported in event-triggered mode).
The UE is requested to report up to 6 neighboring cells amongst the monitored set. The monitored set is defined as
the set of FDD inter-frequency and GSM neighbors of the primary cell and is provided to the UE through a
MEASUREMENT CONTROL message first time the alarm measurement condition is fulfilled and on modification of
the monitored set.
InterFreqMeasConf
interFreqFilterCoeff
NodeB
Measurement Report
RRCMeasurement
Measurement Report rrcIntraFreqMeasurementReportingPeriod
rrcGsmMeasurementFilterCoeff
Measurement
Measurement Measurement
Reporting
ID Results
Quantity
Inter-RAT
and/or 6 Best Monitored cells
Inter-Freq
This function processes the measurements reported by the UE so as to possibly perform HHO. Processing is
different for Inter-frequency or GSM measurements. To improve the measurement report processing in UA08.1, a
measurement report processing delay is implemented.
In order to minimize algorithm changes, the same behaviour is adopted between periodic and full event modes. The
alarm measurement results are reported in periodic mode.
The only difference is that:
In
periodic mode, Inter-Freq/Inter-RAT are declared as additional measurements reported in the same RRC
measurement reports as Intra-Frequency (rrcIntraFreqMeasurementReportingPeriod).
Inevent mode, new measurements are configured to report Inter-Frequency/Inter-RAT measurements (Intra-
Frequency measurements are not impacted and are still reported in event-triggered mode).
In event mode and if both Inter-Frequency and Inter-RAT measurements are used, then the Inter-Frequency
measurement is configured as the main measurement in periodic mode and the Inter-RAT measurement is
configured as the additional measurement of the Inter-Frequency measurement.
The UE is requested to report up to 6 neighboring cells amongst the monitored set. The monitored set is defined as
the set of FDD inter-frequency and GSM neighbors of the primary cell and is provided to the UE through a
MEASUREMENT CONTROL message first time the alarm measurement condition is fulfilled and on modification of
the monitored set.
The Inter-freq measurement list contains cells belonging to multiple carrier frequencies
UE Measurements handled by the RNC after measRepDiscardTimerXXX expiry
or if target cell's technology is preferred AND target cell's Call Type color is GREEN
measRepDiscardTimerAlarmMultiRAT
measRepDiscardTimerCACMultiRAT
measRepDiscardTimerServiceMultiRAT
measRepDiscardTimerAlarmFDD
measRepDiscardTimerCACFDD
measRepDiscardTimerServiceFDD
(RadioAccessService)
Rational: the RNC uses a single 2D threshold for both IFREQ & IRAT measurements and it asks the UE to start
measuring both IFREQ & IRAT neighbor cells at the same time. Also, the RNC uses periodic IFREQ & IRAT
measurements with a periodicity of 0.5s, meaning UE sends a measurement report every 0.5 seconds. Field
observations show that UE reports first cells from the first frequency ranked on Measurement Command and takes
about 2-3s for completing the measurement on all the cells in the neighboring list. The measurement reports sent in
the mean time contain only partial measurements that may even not include cells from the preferred priority layer.
Report Processing Delay Enhancement: to improve the measurement report processing it is implemented a
measurement report processing delay; this implementation is enabled if imcta/measRepDiscardTimerXXX > 0
and only apply if:
Simultaneous IFREQ and IRAT measurements are needed, or
the IFREQ measurement list contains cells belonging to multiple carrier frequencies.
In such conditions, measurement reports are processed the following way:
on expiry of timer imcta/measRepDiscardTimerXXX, or
if target cell technology is preferred (parameter is3GHandoverPreferred) AND target cell Call Type color is
GREEN.
If CM is needed (mono-receiver UE) then CM activation time is additionally taken into account.
The timer measRepDiscardTimer is adjusted for the CM activation time (it starts on start of iMCTA guard timer
and is stopped whenever iMCTA guard timer is stopped). measRepDiscardTimer is adjusted the following way:
UE needs CM for FDD: (CModeConfiguration::cModeDeltaCfn + CModePatternSeqInfo::tgcfnOffset of FDD) * 10
UE needs CM for GSM: (CModeConfiguration::cModeDeltaCfn + CModePatternSeqInfo::tgcfnOffset of GSM
BSIC) * 10
Note: This feature impacts the Measurement report processing part of the iMCTA algorithm.
Preferred layer
False
No
e lls
If the simultaneous compressed mode is requested but not possible then the RNC uses the single compressed mode
as per parameter is3GHandoverPreferred.
So, it specifies whether to perform inter-Frequency (3G) or inter-RAT (2G) measurements if the RAB combination or
the Node B does not support simultaneous CM patterns.
is3GHandoverPreferred = TRUE will also result in choosing an FDD cell as target if the measurement report
contains eligible cells from both 2G and FDD. Consequently, more traffic will handover to other FDD layer(s).
Load Post-Filtering
FDDCell Filtering on
HSxPA capabilities
for iMCTA Alarm or CAC
UMTSNeighbouringRelation
neighbouringCellOffset
Cells with
If no cell is eligible a GSM best FDD
Cell will be selected if a priority
suitable GSM Cell is reported Cell with best EcNo
minimumCpichEcNoValueForHO
miniumCpichRscpValueForHO
(DlUserService)
minimumCpichEcNoValueForHoOffset
minimumCpichRscpValueForHoOffset
FDD1 Primary Cell (UMTSFddNeighbouringCell)
To be eligible to HHO, a FDD neighboring cell must be reported better than 2 thresholds, one for CPICH Ec/No,
another for CPICH RSCP.
In UA07, the RNC filters target cells for which:
CPICH EcNo < minimumCpichEcNoValueForHO
CPICH RSCP < minimumCpichRscpValueForHO
Yes No
(Trigger = IuB CAC) AND
(IuB colour Target cell < IuB colour Source cell)
Yes
<=
Cell Color imctaLoadBasedTargetCellColourThreshold
No
This filtering for iMCTA Alarm and CAC load based is only invoked if isImctaLoadBasedAllowed = TRUE.
In this case, to be eligible to HHO, an FDD neighboring cell shall have a Cell Load Color fulfilling the
imctaLoadBasedTargetCellColorThreshold.
For instance, if imctaLoadBasedTargetCellColorThreshold = Yellow and one DCH color of the neighboring cell
is Red, this cell is removed from the candidate list to HHO.
An additional criterion applies for iMCTA Alarm and CAC load based: the 2G layer should be set with the same
priority value (different from PNA) as the highest FDD priority (activation of simultaneous IFREQ/ IRAT
measurements) is mandatory allowing simultaneous FDD and GSM compressed modes. This way, it is guaranteed
that at least an inter-RAT handover is performed to rescue the call due to coverage loss or resource shortage when
all FDD neighboring cells are discarded due to overload.
In UA07.1, isImctaLoadBasedAllowed enables the cell load criteria when the reason for iMCTA is Alarm or CAC
failure.
No No
No No
R99
cells
Filtering on UE and reported cells HSxPA capabilities aims at optimizing UTRAN radio resources since only HSxPA
capable mobiles are able to use HSxPA radio resources.
Dont forget that for iMCTA Alarm or iMCTA CAC, Priority is the same between the different FDD carriers.
As an example, if UE is HSUPA capable, theRNC will keep the HSUPA-capable cells if present. Otherwise, the
HSDPA-capable cells if present. Otherwise, all cells.
This filtering can be seen as a best-effort filtering.
HSxPA UE stands for HSUPA or HSDPA UE, i.e. R6 or R5 UE.
gsmCellIndivOffset
RadioAccessService
2G Cell X
Color
inhibitTimer3g2g(Load)
inhibitTimer3g2g
RadioAccessService
minimumGsmRssiValueForHO
To be eligible to HHO, a GSM neighboring cell (BCCH Rxlev) must be reported better than a threshold.
Dl(Ul)GreenLoadValue
gsmDl(Ul)AvailableCapacityThresholdToRedColor
Dl(Ul)YellowLoadValue
Dl(Ul)RedLoadValue
inhibitTimer3g2gLoad
2G Cell X
Color HHO to HHO to
2G cell
HHO reject
2G cell GSM
from 2G
X X Unloaded 2G cell
inhibitTimer3g2g Loaded 2G cell
The aim of this feature is to make better 2G Target cell selection during an inter-system handover procedure,
thanks to a better knowledge of the 2G cell load, which is provided by the 2G BSC.
Moreover, the RNC takes the opportunity to provide the 2G BSC, during an inter-system handover procedure, with
the 3G cell load information so as to also allow it to improve the 3G Target cell selection. The 3G cell load
information is sent in the following RANAP messages:
Relocation required / old BSS to new BSS information
Relocation request ack / new BSS to old BSS information
Relocation failure / new BSS to old BSS information
During inter-system handover procedures, the feature must ensure the following functions when the feature is
activated:
Compute the UTRAN cell load information and send it to the BSC.
When the GSM cell load information is provided by the BSC, compute the GSM cell load color based on this
received cell load information.
If the 2G network supports the feature Unified RRM Step 2, the GSM cell load information is received by the
RNC in the following messages:
Relocation request / source RNC to target RNC
Relocation command / inter system information transparent container
Relocation preparation failure / inter system information transparent container
Filtering on
HSxPA capabilities
for iMCTA Service
Filtering on
load criteria
FDD Eligible Cells filtering is the same as for iMCTA Alarm or CAC.
Filtering on HSxPA capabilities for iMCTA Service is explained later in the document.
Filtering on load criteria is explained later in the document.
No
R99 cells
only
Unlike iMCTA Alarm and iMCTA CAC, a new filtering is introduced based on the hsxpaSegmentationEnable
parameter.
When this parameter is set to True, the algorithm removes all reported cells that are not compatible with UEs
HSxPA capabilities (sent in the RRC Connection Setup Complete message):
If UE is NOT HSxPA-capable, all HSxPA-capable cells are removed (either HSDPA or HSUPA).
If UE is HSxPA-capable (either HSDPA or HSUPA), all non HSxPA-capable cells are removed.
NO
Target Cell load Green or targetCellColourThreshold
YES
isServiceSegmentationTopPriority NO
AND
YES
Source cell NOT fully compatible
YES
YES
target cell fully or partially compatible
umtsFddNeighbouringCell isServiceSegmentationTopPriority
YES
AND
Source cell NOT fully compatible
umtsNeighbouringImcta Target cell is kept
NO
The idea of iMCTA Service is to improve quality of service or to optimize resource usage.
Therefore, all reported neighboring cells whose load is worse than targetCellColourThreshold must be removed.
This is applicable whatever Priority of the neighboring cell.
For instance, if targetCellColourThreshold= YELLOW and one DCH color of the neighboring cell is RED, this cell
is removed from the candidate list to HHO.
Note: an FDD cell belonging to a neighboring RNC is considered as RED as Serving RNC is not able to determine its
load color.
Comparing reported neighboring cells iMCTA color with targetCellColourThreshold is systematically performed
by iMCTA Service, for load balancing purposes but also for traffic segmentation (which is based on UE and NodeB
HSxPA capabilities).
If the operators top priority is to perform traffic segmentation rather than load balancing,
targetCellColourThreshold must be set to RED so as to systematically go further in the algorithm, whatever
neighboring cell load.
isServiceSegmentationTopPriority=True AND originating cell is NOT fully compatible with UE capabilities,
If
remove cells that are NOT compatible with UE capabilities:
Fully and partially compatible cells are kept
Flag defined per serviceType
If priority is equal to primary cell:
If cell load is GREEN or better than Primary cell load, keep the cell
Otherwise remove the cell except when isServiceSegmentationTopPriority=True AND originating cell is
NOT fully compatible with UE capabilities
forget that for that neighboring cells whose priority is worse than the Primary cell one have been
Dont
removed at Neighbouring Cell Searching and Filtering stage.
GSMCell
GSM
Eligible Cells GsmImcta
targetCellColourThreshold
Keep cells with
load <=
targetCellColourThreshold
2G Cell X
Colour
inhibitTimer3g2g
GSM Eligible Cells filtering is the same as for iMCTA Alarm or CAC.
GSM Cell Load must be understood here as a means to inhibit an HHO to a GSM cell to which a previously
attempted HHO has failed in the last inhibitTimer3g2g seconds:
A GSM target cell will have a RED Cell Color if a handover towards this cell has been rejected for load reasons
in the previous inhibitTimer3g2g seconds.
otherwise the Cell Color is GREEN.
The 2G cell load detection is based on receipt of the RANAP Relocation Preparation Failure message with a cause IE
value equal to Relocation failure in target CN/RNC or target system.
measurementGuardTimer
no candidate cell
measurementGuardTimer
no candidate cell
measurementGuardTimerFdd
measurementGuardTimer
no candidate cell measurementGuardTimer2g
HHO Decision is taken by the RNC according to the Measurement Reported on neighboring cells: HHO triggered can
be Inter-RAT normal or blind, Inter-Freq Intra-RNC or Inter-RNC.
The period during which iMCTA Alarm waits for neighboring reported cells is bounded by 2 different guard timers,
so-called measurementGuardTimerFdd and measurementGuardTimer2g, depending on the selected target
Access type.
Atguard timer expiration, if no neighboring cell is candidate for HHO (no reported cell has been reported or the
reported cells have been discarded):
For iMCTA Alarm:
RNC triggers 2G Blind HO if provisioned.
otherwise Compressed Mode is reactivated if Event 2F has not been received (iMTCA Alarm only).
For iMCTA CAC
RAB cannot be established and the RNC sends RANAP RAB Assignment Response (RAB failed list)
to CN.
For iMCTA Service
The UE remains on the initial frequency.
measurementGuardTimerFdd and measurementGuardTimer2g replace UA04.2 previous CM timers
(gsmCmodeReactivationTimer, fddCmodeReactivationTimer) and HHO timers (blindHhoGsmTimer,
blindHhoFddTimer).
If iMCTA Alarm or iMCTA CAC HHO to another FDD frequency or to GSM fails due to any reason (internal SRNC
failure, NBAP, RNSAP, RANAP or RRC failure message) and iMCTA Alarm or CAC fallback is enabled via parameter
imcta/imctaFallbackHoFailure then the RNC will once re-attempt HHO to another target cell.
The fallback is triggered only once. If the fallback attempt also fails, then no 3rd attempt is done.
The target for a fallback is selected as follows:
Measurement based HO: The 2nd best cell if reported by the UE.
Blind HO: The next Twin cell if configured (2nd best target).
RNC
RadioAccessService
maxCompoundingListSizeInterFreq [16..32]
maxCompoundingListSizeInterRAT [16..32]
NeighbouringRNC
NodeB numOfPrimaryCellNeighbourInterFreq
numOfPrimaryCellNeighbourInterRat
typeOfCompoundingNeighbourListInterFreq
typeOfCompoundingNeighbourListInterRat
FDDCell
typeOfCompoundingNeighbourListInterFreq
typeOfCompoundingNeighbourListInterRat
numOfPrimaryCellNeighbourInterFreq
numOfPrimaryCellNeighbourInterRat
GsmNeighbouringCell UMTSFddNeighbouringCell
neighbourCellPrio neighbourCellPrio
UsHoConf [0..21]
cpichEcNoThreshold
cpichRscpThreshold
counter
FastAlarmHardHoConf stepUp
stepDown
FullEventHoConfHhoMgt timerAlarmHoEvent2
D
FullEventHoConfHhoMgtEvent2D
timeToTrigger2D cpichEcNoThresholdUsedFreq2D
hysteresis2D cpichRscpThresholdUsedFreq2D
FullEventHoConfHhoMgtEvent2F
timeToTrigger2F cpichEcNoThresholdUsedFreq2F
hysteresis2F cpichRscpThresholdUsedFreq2F
NodeB RadioAccessService
isImctaLoadBasedAllowed
isCEMLoadUseRestrictedAllowed
isImctaFddLayerLoadPreCheckAllowed
FddCell Imcta
measurementGuardTimer2g
measurementGuardTimerFdd
serviceHoRanapIeEnable
inhibitTimer3g2g
ServicePriorityTableConfClass
ServiceSegmentationPriorityClass 1..7
Question: Fill up the Service Priority General Table and the Service Priority For
HSDPA Table next page
Assume that Service Priority General For HSUPA Table is not configured
3G FDD2
HSDPA HSxPA HSxPA HSxPA HSDPA 3G
FDD1
R99 R99 R99 R99 R99 R99 R99 R99 R99
GSM GSM GSM GSM GSM GSM
2G
B1 B3
C1 C3 C5
3G F3 -6dB -22dB
-5dB
C2 C4 C6
-12dB -20dB -3dB
Supposing the service used is CsSpeech, what would be the target Cell for HHO?
Supposing the service used is PsIb, what would be the target Cell for HHO?
Core Network
BCCH
1
3
activationHoGsmCsAllowed
activationHoGsmPsAllowed
Handover From UTRAN Command isInterFreqMeasActivationAllowed
2 or
(RadioAccessService)
Cell Change Order
isGsmCmodeActivationAllowed
isPatternAllowed (DlUserService)
(CmodePatternSeqInfo)
Inter-system HHO can occur following iMCTA Alarm, CAC or Service triggering. The selection between FDD and 2G
Access is part of the iMCTA algorithm, mostly based on UE capabilities, priority tables and available neighboring
cells.
2
iMCTA Algo ranking: G1,G2
G1 Best candidate Selected for HHO 3
Relocation Required G1
5
G2 Selected for HHO RNC BSC
4
Relocation Failure G1:
Cause: Lack of radio resources
1
Measurement
report CM
InterRat Cells
G1,G2,G3
G1 G3
Cell 1: Primary Cell G2
If the SRNC received a RELOCATION PREPARATION FAILURE message upon a RELOCATION REQUIRED message
for relocation preparation to a GSM target cell with one of the following causes:
No Radio Resources Available in Target cell (or)
Relocation Target Not Allowed (or)
Relocation Failure in Target CN/RNC or Target system
Then UTRAN can retry relocation preparation to the next best GSM cell if the UE has reported more than one cell
with sufficient received BCCH level.
This mechanism can be enabled via the isGsmIratHoToNextBestCellAllowed parameter introduced in UA07.
Else SRNC shall maintain the CS call on 3G with no further HHO attempt. A new 3G 2G HHO attempt may be
triggered again at the next 2G measurement report message received from the UE.
SRNC shall only trigger another relocation preparation procedure to the next best GSM cell (with lowest cell load
preferably (if available) and highest RSSI), based on latest 2G measurement report, if and only if:
An alternative GSM target cell exists as per the latest 2G measurement report.
If
the UE has sent a combined measurement report for inter-RAT and inter-frequency then the report contains
no suitable inter-frequency target cells.
If
eligible neighbors present in report for both inter-RAT and inter-frequency then HHO target layer selection
will be based on is3GHandoverPreferred.
The need for handover is still present, e.g. the alarm condition is still active.
SRNC
4 P-CPICH
is3Gto3GWithoutIurAllowedforCS
isIrmCacForInterFreqIntraRncEnable
is3Gto3GWithoutIurAllowedforPS
(RadioAccessService)
(RadioAccessService)
If the Inter-Freq Intra-RNC HHO takes place in a DRNC then the procedure is:
eithera handover over Iur with the neighboring RNC to be the DRNC which controls both the source and the
target cells.
or a SRNS relocation UE Involved with this DRNC to become the new SRNC.
d CN Rel
1
e q uire oca 2
R t io
ca tion a n d nR
equ
el o m R elo c
R C om .
est
i o n Req
t . Ac
oca 4 3
SRNC Rel k
Target RNC
P-CPICH
8
7
6
5
is3Gto3GWithoutIurAllowedforCS
isInterFreqHandoverOverIurAllowed is3Gto3GWithoutIurAllowedforPS
= False
(NeighbouringRNC) (RadioAccessService)
Prior to UA07, Inter-RNC HHOs are processed in the same way whether there is Iur or not, i.e. through a SNRS
Relocation UE involved procedure through the CN.
From UA07, Inter-RNC HHOs are processed using the SRNS Relocation UE involved if the
isInterFreqHandoverOverIurAllowed parameter is set to False.
P-CPICH
6
5
4
3
is3Gto3GWithoutIurAllowedforCS
isInterFreqHandoverOverIurAllowed is3Gto3GWithoutIurAllowedforPS
= True
(NeighbouringRNC) (RadioAccessService)
The radio link to the target cell is to be set up with the RNSAP Radio Link Setup or RNSAP Radio Link
Addition procedure depending on the DRNC being new or already existing in the call context. The radio link to the
source cell needs to be released with the NBAP Radio Link Deletion procedure.
2. Partial relocation from LTE by accepting only a part of RAB in RAB list contained in
the RANAP relocation request
Event A2
EventA2
Measured quantity
hyst A2 Theshold
LTE Cell
timeToTriggerA2
Event A2: Serving becomes worse than threshold
1 10 71 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Inter-Carrier Mobility in HHO
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
Feature 104489 LTE to UMTS HHO provides Handover from LTE to UMTS and support the following procedures:
LTE to UMTS PS HHO: incoming SRNS relocation request triggered for a UE connected in 4G radio network;
Inter-RAT Mobility events used to trigger mobility 4G to 3G:
Event A2: Serving becomes worse than threshold.
Event B2: Serving becomes worse than threshold1 and inter-RAT neighbor becomes better than
threshold2.
CS Fallback from LTE to UMTS: The CS fallback procedure only could be started following the successful
completion of PS service relocation.
Event B2
EventB2 3G Condition
4G Condition
Measured quantity
3G Hyst. 3G Th2
LTE Cell
4G Hyst. 4G Th1
3G Cell
timeToTriggerB2
Event B2 :Serving becomes worse than threshold1
and inter RAT neighbor becomes better than threshold2
1 10 72 COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
W-CDMA R99 Algorithms Description Inter-Carrier Mobility in HHO
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
Handover Required
isPartialRelocationFromLteAllowed
Relocation Request
UA08.1 introduces the capability for the
RNC to perform RAB Selection (RAB To Setup list IE prioritized)
to filter out some RABs before performing
Relocation Request Ack
RAB Matching
(RABs Failed To Setup List RAB
Selection & RAB Matching)
Forward Relocation Response
UTRAN supports maximum 3 PDP contexts in UA08 and E-UTRAN supports till 8 RABs for a given UE.
The Ranap Relocation Request from LTE through the target SGSN may contain more than 3 RABs to be set up.
Some RAB combinations/services may not be supported by the RNC. As a result, UA08.1 introduces the capability for
the RNC to perform RAB selection to filter out some RABs before performing RAB matching.
No
partialRelocationFr Existing RAB matching
omLTE == true? algorithm.
Yes
st
Keep the 1 conv. RAB. Move
other conv. RABs and all
Yes PS conv. RAB Yes streaming RABs to failed
conversational supported? RAB list
RABs included?
No No
Yes
#streaming st
Keep the 1 streaming RAB.
RABs > 1? Move other streaming RABs to
failed RAB list
No
Select 1st (up to) 3 RABs from remaining RAB to setup list. Add the rest RABs to the failed RAB list.
No
#RABs >0?
Peg counter NT_Iu_relocation_request_failures_ps
(537) screening
4Gto3GRejectionDueToRabMatchingFailure (12)
Move the last
RAB to failed
RAB list.
RAB Matching algorithm Relocation preparation fails.
Send RANAP relocation
failure.
No Yes
RAB Matching
succeessful?
No
Continue RAB setup for the
(#RABs ==2 && they are successful RABs.
Sig+Conv)? or Peg counter PS_Rab_Selection
(#RABs ==1 && it is
Conv.)?
CS fallback procedure
The CS fallback procedure could only be started following the successful completion of PS service relocation as
detailed in previous sections.
Regardless the Circuit Switched Fallback (CSFB) type (e.g. Mobile termination CS fallback or Mobile origination CS
fallback), the CSFB procedure is always beginning with an RRC UL INITIAL DIRECT TRANSFER message of the "CN
domain identity IE" set to "CS domain, with the embedded NAS message of IDT set to either Paging Response or
Service Request.
The CS service is then setup in UTRAN with PS service already established.
UA08.1 PM108661 introduces a new RNC timer (TmaxDelayForCsCallEstablishment). The timer starts only for
handover with cause set to CS fallback triggered, after the Relocation Complete message is sent to the SGSN. The
purpose of this timer is to check whether a CS call is successfully established or not after an LTE to UMTS relocation
request with cause CS fallback.
If a CS establishment request (Initial Direct Transfer - Service Request, or Paging Response, with CS domain) is for
an established PS call with a CS fallback, timer TmaxDelayForCsCallEstablishment will be running.
When the Initial Direct Transfer message is received by the RNC, this CS establishment is identified and treated as
LTE to UMTS Handover with CS fallback:
Timer TmaxDelayForCsCallEstablishment is stopped.
CS establishment is treated as normal CS call setup.
Otherwise, this CS establishment is treated as normal CS call setup.
For LTE to UTRAN handover for CS fallback, handling the subsequent CS establishment after the PS handover
completes the whole CS fallback procedure.
RNC
PsCoreNetworkAccess RadioAccessService
FromEutraRelocationCiphering isEutraToUtraHhoAllowed
isPartialRelocationFromLteAllowed
maxDelayForCsCallEstablishment
UA07.1.2/7.1.3
parameters
New UA08.1 parameters
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
1 iMCRA overview 7
1.1 Objectives of iMCRA 8
1.2 iMCRA-step 1 (UA07) 10
1.3 iMCRA-step 2 (UA08) 11
1.4 Twin cells definition 12
2 iMCRA step 1 13
2.1 Redirection types 14
2.2 UE capabilities and call type 15
2.3 Cell capabilities and FaType 16
2.4 Candidate target cell list selection procedure 17
2.5 Load balancing procedure 19
2.6 RRC redirection based on CAC 20
2.7 iMCRA: RAN model 21
2.8 Exercise 1 22
2.9 Exercise 2 23
3 iMCRA step 2 24
3.1 Principle 25
3.2 Call type calculation 26
3.3 Action list mapping 28
3.4 Carrier selection list processing 29
3.5 Load condition calculation 30
3.6 Target carrier selection
1 11 5
31
3.7
W-CDMACAC failure
R99 Algorithms case
Description
COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
Inter-Carrier Mobility at RRC Connection 34
UTRAN UA08 9300 W-CDMA R99 Algorithms Description
3.8 RAN model 35
3.9 Exercise 36
LTE layer
Why iMCRA?
Traffic Segmentation
UE & Cell Capability
GSM layer
RRC CAC Failure
Load balancing
FDD2 HSPA
To increase the network capacity and optimize HSxPA throughput, operators may deploy multi-layer configurations
with several layers structures:
Multi-layers
based on asymmetric topologies (dedicated layers for HSxPA or Data traffic separated from
dedicated layers for R99 or Conversational traffic).
Multi-layers based on symmetric topologies (shared layers for HSxPA and R99 traffic).
Several features are used in order to distribute the traffic between the different layers including 2G (GSM, CDMA)
and 4G (LTE):
Idle Mode & Hierarchical Cell Structure (HCS) allows strategies based on UEs camping homogeneously on
different frequencies or favor specific carriers or even prioritizing cell layers for mobiles in idle mode, Cell_FACH
and URA/Cell_PCH connected modes. The HCS cell reselection algorithm also takes into account UE speed so
that fast moving UEs can be placed in large cells to avoid excessive cell reselections.
Intelligent Multi-Carrier RRC Connection Allocation: iMCRA allows redirecting UE to FDD, GSM or LTE
cells at RRC connection establishment. iMCRA Step2 provides a powerful programmable RRC redirection action
matrix giving great flexibility to operators for a best use of their access carriers resources. It provides the
capability to process in parallel the RRC Connection Request message or the failure of SRB CAC based on:
iMCRA Service may be triggered by the RRC Connection Request message, enabling RRC redirection to
an FDD cell or the GSM RAT or the LTE RAT based on common or mixed RRC Redirection strategies.
iMCRA CAC may be triggered by the failure of SRB CAC induced by any CAC failure cause, enabling RRC
redirection to an FDD cell or the GSM RAT or the LTE RAT based on common or mixed RRC Redirection
strategies.
Intelligent
Multi-Carrier Traffic Allocation (iMCTA) allowing handover UE to another layer when in
Cell_DCH connected mode.
UE NodeB RNC
Call Type analysis
RRC Connection Request
(Establishment Cause)
iMCRA
Call setup in originating cell
Algorithm
When activated, iMCRA step2 replaces the following features (meaning the user can define an action list and a
carrier selection list to reproduce the function of those features):
iMCRA Step 1
3G/2G Redirection based on cell load (partially)
3G/2G Redirection for speech call
Load Balancing between HSPA carriers (partially, iMCRA part)
The twin cells are co-located inter-frequency cells plus cells from other Node
B for blind redirection referenced by Cell Id
NodeB1
twinCellList
FDD 1
(InterFreqHhoFddCell)
CellId xyz
FDD 2
FDD 3
FDD 4
FDD 5
NodeB 2
The twin cells are co-located inter-frequency cells plus cells from other Node B for blind redirection referenced by
Cell Id.
Cells FDD2, FDD3, FDD4 and FDD5 are twin cells of cell FDD1.
preferredFa
based on Frequency Allocation
Preferences
capaOnly capaAndEstCause
based on UE HSPA based on UE Capabilities
Capabilities only and Call Type
rrcRedirectionType
(InterFreqHhoFddCell)
Cac
None
based on RRC
Call Admission Failure Algorithm Disabled
During the RRC Connection Setup procedure, the RNC determines the preferred Frequency Allocation (FA) based
on:
UE
capability only (Redirection based on UE HSPA capabilities already available in UA06 plus segmenting
R6+ UEs, plus choice of less loaded target cell).
UEcapability and establishment cause (Redirection based on UE HSPA capabilities and call type already
available in UA06 plus segmenting R6+ UEs, plus choice of less loaded target cell).
Preferred FA (Select original cell or twin cell with lowest cell load).
Note: FA Frequency Allocation = Carrier
Optional FA classification: Conversational, Data or Other layer
OriginatingCellColourThresholdset to either GREEN/YELLOW/RED
GREEN: Full Load distribution; high number of redirections
RED: Load distribution only when originating cell gets overloaded
CAC (Redirection to twin cells with lowest load if CAC failure on originating cell)
None (Redirection disabled)
The UE capability combinations are based on the Access Stratum Release Indication IE in the RRC Connection
Request message for the identification of the different UE types.
The call type identification is based on the Establishment Cause IE in the RRC Connection Request message to
distinguish the different call types.
Note:
From UA07, the following causes Originating Subscribed Traffic Call, Registration and Originating High Priority
Signalling are mapped to the Data type but they may be removed from the list of Data causes that could perform a
redirection. Indeed, for short and low traffic procedures some operators may prefer to keep the UE on the
originating cell that had been selected by the UE for access.
From UA07, the spare RadioAccessService.reserved2 byte 2 (3 rd byte) is used to handle the configuration referred
above. For future releases, Boolean; {True, False} parameter
RadioAccessService.isOrigHighPrioSigAndRegistPreferredDataCall will apply.
RadioAccessService
RRCRedirectionConfClass 0..6
rrcRedirectionType
isRrcRedirectionInterFreq = True
layerPreferredForR99
Please note that an optional criterion introduced by UA06 feature 84900 is based on the preferred layer to
distinguish the R99 services potentially served on the preferred layer from the others. It is only applicable for
configurations using rrcRedirectionType = capaOnly or capaAndEstCause.
In such conditions, for all cells with layerPreferredforR99 = TRUE, the RNC shall assume HSPA Cell Capability =
DCH. For all cells with layerPreferredforR99 = FALSE, the RNC shall assume HSPA Cell Capability = HSPA.
When to decide if an RRC connection should be redirected to the DCH (R99) layer, the function check
layerPreferredforR99 for both cells, in addition to the existing criteria. Only when both twin cells are
configured as HSPA capable, layerPreferredforR99 is checked. If the current cell does not prefer R99 but the
twin cell does, redirect is selected. Otherwise, the current behavior is unchanged.
When to decide if an RRC connection should be redirected to HSPA layer, the function check
layerPreferredforR99 for both cells, in addition to the existing criterias. Only when both twin cells are
configured as HSPA capable, layerPreferredforR99 is checked. If the current cell prefers R99 but the twin
cell does not, redirect is selected. Otherwise, the current behavior is unchanged.
This option applies in case HSxPA is deployed on both layers (allowing both layers to carry HSPA traffic and keeping
RRC Traffic Segmentation enabled for R5+ UEs). It is possible to prefer some of the HSPA capable layers for R99
traffic by using parameter layerPreferredforR99 (it is possible to configure several cells with
layerPreferredForR99=TRUE and to configure R99-only cells, layerPreferredForR99 cells and HSPA cells):
DCH service originated on non-preferred cell will be redirected to the preferred cell.
HSxPA service originated on preferred cell will be redirected to the non-preferred cell.
This algorithm is enhanced from UA6.0, i.e. the RNC is able to find preferred HSxPA cells for DCH traffic in case
using RRC Traffic Segmentation based on UE Capabilities. This enhancement is only applicable if the originating cell
and operational twin cell is enabled for HSDPA (parameter hsdpaActivation).
Emergency calls are redirected in case of CAC failure only.
Conversational DCH
DCH Data DCH
Other Originating
UE Capability and Conversational DCH
Establishment Cause HSDPA Data HDSPA Same as in
capaOnly
Other Originating
Conversational DCH
HSUPA Data HSUPA
Other Originating
rrcRedirectOrigCellColorThreshold
Load Balancing is applied for any rrcRedirectionType (capaOnly, capaAndEstCause, preferredFA and cAC)
after the Candidate Target Cell List Selection procedure. This procedure will find the best target cell among the
candidate target cell list provided by the Candidate Target Cell List Selection phase.
If the originating cell is included in the set of candidate cells and the originating cell color is better than
rrcRedirectOrigCellColourThreshold then the call is established in the originating cell (note: not applicable for
rrcRedirectionType =cAC).
Otherwise, if two or more cells are selected, the following steps (with decreasing priority) are processed until
retaining one cell:
Preference is given to the less loaded cells.
The originating cell is selected if it is included in the list. Otherwise, select the twin cell with the round robin
algorithm based on frequency number (dlFrequencyNumber > originating one).
isRrcRedirectionInterFreq = True
rrcRedirectionType = cac
2
Second step: Load Balancing
If CAC failure occurs,
RNC select cells with lowest
cell color (green< yellow < red>
Black)
RRC Redirection can trigger re-try of the RRC Connection Allocation procedure when CAC failure occurs.
If CAC occurred during SRB establishment on the originating cell then the RNC considers redirection to a list of twin
cells. Emergency calls are also redirected in case of CAC failure.
RN
UA08 RAN model for iMCRA Step 1 C
RadioAccessService NodeB
isRrcRedirectionInterFreq
isOrigHighPrioSigAndRegistPreferredDataCall FddCell
layerPreferredForR99
twinCellList
RrcRedirectionConfClass 0..6
InterFreqHhoFddCell
In UA08.1, the TwinCellList parameter moves from the InterfreqHhoFddCell object to the FDDCell
object.
rrcRedirectOrigCellColourThreshold
faType FDD1
faType FDD2
faType FDD3
For both configuration scenarios, determine which is the target cell for the RRC
Establishment.
This function calculates the iMCRA call type associated to the RRC
Establishment Cause.
isRedirectionBasedOnEstablishmentCause
UE capability iMCRA call types
(basic call type, group
call type)
iMCRA Basic Call type and iMCRA Group Call type calculation can be split to several steps:
Step1 RRC message derives RRC cause Types
Step2 RRC cause types mapped to Basic call types
Step3 Basic call types simplify to Group call Types
iMCRA basic call types and their mapping to the iMCRA group call types:
Group Call Types MO CS CS DCH HSPA PS Data CS/PS
All
Group Group Group Group Group Group
Group Call
Basic Call Types Call Call Call Call Call Call
MO CS Voice Call X X X X X
MT CS Voice Call X X X X
MO CS Conversational Call X X X X X
CS
MT CS Conversational Call X X X X
CS Voice Call X X X X
CS Data Call X X X X
PSDCH Call X X X X
HSDPA Data Call X X X X
HSxPA Data Call X X X X PS
DC Data Call X X X X
LTE Data Call X X X X
Other Call X
MT CS Conversational Call: Mobile terminated CS voice or data call over DCH (UEs < Rel-6)
CS Voice Call: Mobile originated or terminated CS voice call on DCH, if the use of RRC establishment cause is
disabled (UEs > = Rel-6)
CS Data Call: Mobile originated or terminated CS data call on DCH, if the use of RRC establishment cause is
disabled (UEs > = Rel-6)
PS DCH Call: PS data call over DCH
DC Data Call: PS data call over HSPA and dual cell option
LTE Data Call: PS data call over HSPA and LTE redirection option
Other Call: Signalling connections (e.g. location registrations). If the operator has disabled the use of the RRC
establishment cause by setting isRedirectionBasedOnEstablishmentCause to FALSE, then Other Call
includes all calls for UEs < Rel-5 and all calls for Rel-5 UEs in CS domain (for all Rel-5 UEs in the PS domain
HSDPA Data Call is assumed).
High
1 RadioAccessService
Priority order
Imcra
Low
ImcraActionList
2
ImcraAction
imcraCallType
redirectionPercentage
carrierSelectionListId
This function provides a pointer to the iMCRA Carrier Selection List based on the iMCRA call type matching. The
calculated iMCRA call type shall be matched with configured call type in the action list per action (basic call types or
group call types).
The first action in the list has the highest priority. Thus basic call types should be configured first before actions for
group call types. For the redirection of a certain percentage of calls according to the "Redirect Percentage" field in
the action list, the following formula applies: Redirect if (RedirectPercentage + (#call * RedirectPercentage) % 1 >=
1); if the load condition is fulfilled and the call is selected according to the formula, a selected target carrier is
selected. Otherwise, the RNC will check the next load condition, if configured.
Maximum 50 Action Lists can be defined. Each Action List may contain maximum 15 actions.
Ca rrier Selectio n
Algorithm
N
Call Ty pe
in Action L ist?
Y N
Up date red irection
percentag e figure
N Redirection
req uired according
to p ercen tage
T arget carrier = figu re? T arget carrier =
o riginating cell originating cell
Y
End
The Carrier Selection List contains a load Rule per instance plus an
associated list of allowed target carriers
1 RadioAccessService
ImcraImct
a
CarrierSelectionList
CarrierSelectionRule
2
csrTargetCarrierList
CarrierSelectionCondition
High
csrCarrier
Priority order
csrLoadType
Low csrOperator
csrLoadValue
After matching the calculated iMCRA call type with one redirection action configured in the action list, RRC
Redirection Decision shall process its associated "Carrier Selection List.
The Carrier Selection List contains a load Rule per instance plus an associated list of allowed target carriers (target
carrier list). The target carrier list includes FDD carriers or GSM or LTE RAT.
The following figure provides an example to the Carrier Selection List. The following conditions apply:
The first instance (#1) of the Carrier Selection List has the highest priority and is processed first (the
conditions stored in the instances (#1, #2#y) of the Carrier Selection List are implicitly linked by an OR
operation = If the load condition does not match or no target carrier has been found, then the next instance of
the Carrier Selection List is processed).
A carrier selection rule can comprise up to four load conditions, which are linked by a logical AND operation,
i.e. if all of conditions of a rule are fulfilled, the carrier(s) in the target carrier list are applied.
Up to 150 Carrier Selection Lists can be defined per RNC (=three in mean per Action List). Each Carrier
Selection List can contain up to 20 rules with up to 4 conditions each.
This function will verify the Load Condition to apply for the associated
target carrier list
This function will verify the Load Condition to apply for the associated target carrier list. Theload condition in the
Carrier Selection List comprises:
A frequency reference to an FDD frequency (F1, F2, F3, F4, F5, F6, F7, F8, F9, F10, F11, F12, Fo); carrier is a
reference to a frequency object defining the carrier (F1 F12) or it references to the originating carrier (Fo).
A load type (Single and Combined load type).
An operator (=, <=, <, >, >=, TRUE); If the "operator" field of the condition is set to TRUE, then the load
condition is always true.
A load value (Green < Yellow< Red < Black).
The RNC supports the following single load types as part of a load condition in the Carrier Selection List: DL POWER
Load; DL OVSF Load; DL IUB Load; DL CEM Load; UL CEM Load; UL RX POWER Load; HSDPA Load;
The RNC supports the following combined load types as part of a load condition in the Carrier Selection List: CEM
Load = max (DL CEM and UL CEM loads); DL RADIO Load = max (DL POWER and DL OVSF loads); DL DCH Load =
max (DL RADIO, DL CEM and DL IUB loads); UL DCH Load = max (UL RX POWER and UL CEM loads); DCH Load =
max (DL DCH and UL DCH loads); HSDPA UL DCH Load = max (HSDPA Load and UL DCH Load); HSPA Load =
HSDPA load; TOTAL Load = max (DCH and HSPA loads); CALL TYPE Load - DCH or HSDPA load.
This function will select a suitable target carrier for redirection among
the list of target carriers fulfilling the load condition
carrier
FDD cells
T
RA
M
GS
Target
LTE RAT
If multiple carrier types are provisioned in the target carrier list, the RNC
tries to selects:
1. first an FDD cell, then
2. the GSM RAT and then
3. the LTE RAT.
If the target carrier list contains two or more applicable FDD cells, the following sorting rules are applied:
1. Less loaded FDD cells shall be ranked highest.
2. Among FDD cells with equal load, the RNC shall rank the origination cell first if it is applicable.
3. Among FDD cells with equal load without originating cell, the RNC shall rank the FDD cells sorted by their
dlFrequencyNumber from low to high (round robin).
The load criterion to be used for load comparison depends on the call type, the UEs HSDPA/HSUPA capability and
the FDD cell capability. The following load criteria are fixed coded:
The 2nd and 3rd criteria are only applied, if the previous selection criterion results in multiple cells with the same
load. The calculated DCH cell load value is derived from the individual cell loads and their weighted load color
FDD2 HSPA
3. LTE Target Carrier:
The RNC checks whether the UE
supports redirection to LTE according FDD 1 DCH
to the IE "Pre-redirection info" traffic
included in the RRC Connection
Request message.
In case of configured GSM target carrier, the RNC cannot check the GSM inter-RAT capability of the UE upon RRC
connection request. Nevertheless, the UE is redirected to GSM, which is performed by the UE via intra-PLMN cell
selection procedure. If the UE does not support the deployed GSM band, it will repeat the RRC connection request.
Each FDD cell refers to two action lists, one is used for iMCRA Service actions and the other is used for iMCRA
CAC actions.
Each action of an action list references to a Carrier Selection List.
Multiple FDD cells can point the same action list / Multiple actions can reference to the same Carrier Selection List.
Node RadioAccessService
B
FddCel Imcra ImcraImcta FddCarrie DedicatedConf
l
isImcraActivated
r
dlFrequencyNumber
imcraServiceActionListId
imcraCacActionListId ImcraActionList CarrierSelectionList ulFrequencyNumber CacConfClass
twinCellList fddUnsupportedTargetCarrierList
fddRejectCellList CarrierSelectionRule
isRedirectionBasedOnEstablishmentCause HspaCellLoadParameters
isEmergencyCallRedirectionOnCac csrTargetCarrierList green2YellowHSPAPwrDlLoadThreshold
ecNoThresholdForTwinCellSelection ImcraAction yellow2GreenHSPAPwrDlLoadThreshold
isEcNoCheckFor3GTwinCellAllowed imcraCallType CarrierSelectionCondition yellow2RedHSPAPwrDlLoadThreshold
isEcNoCheckForLteTwinCellAllowed redirectionPercentage csrCarrier red2YellowHSPAPwrDlLoadThreshold
carrierSelectionListId csrLoadType red2BlackHSPAPwrDlLoadThreshold
NodeBConfClass csrOperator Black2RedHSPAPwrDlLoadThreshold
csrLoadValue green2YellowHSPACodesDlLoadThreshold
yellow2GreenHSPACodesDlLoadThreshold
yellow2RedHSPACodesDlLoadThreshold
UlIrmCEMParameters IrmIubTransportLoadParameters red2YellowHSPACodesDlLoadThreshold
red2BlackUlCEMThreshold red2BlackHSPACodesDlLoadThreshold
red2BlackDlTLCThreshold Black2RedHSPACodesDlLoadThreshold
black2RedUlCEMThreshold black2RedDlTLCThreshold
DlIrmCEMParameters IrmOnCellColourParameters
red2BlackDlCEMThreshold
red2BlackCLCThreshold
black2RedDlCEMThreshold
black2RedCLCThreshold
New UA08 Object red2BlackPLCThreshold
black2RedPLCThreshold
Load balancing policy with 2 Carriers (F1, F2) & 1 GSM Layer
Question: Give the suitable parameters configuration of the Action List and
Carrier Selection List given next slide
iMCRA Service Carrier Selection Lists iMCRA CAC Carrier Selection Lists
iMCRA Carrier Selection List #1 (Service R99) iMCRA Carrier Selection List #3 (CAC MO Voice)
# Carri Load Opera Load Target # Carr Load Operat Load Target
er Type tor Value Carrier List ier Type or Value Carrier List
1 1 <
2
iMCRA Carrier Selection List #2 (Service HSPA)
# Carri Load Oper Load Target iMCRA Carrier Selection List #4 (CAC All)
er Type ator Value Carrier List
# Carri Load Opera Load Target
1 er Type tor Value Carrier List
1
UTRAN
UA08 9300 W-CDMA R99 Algorithms Description
TMO18255_V2.0-SG Edition 1
Document History
#
16-QAM Switch to
16 Quadrature notes view!
Amplitude D
Modulation DCCH Dedicated Control CHannel
1xEV-DO 1x EVolution Data Only DCH Dedicated CHannel
1xEV-DV 1x EVolution Data and Voice DL Downlink
1xRTT 1 times 1.25MHz Radio Transmission DPCCH Dedicated Physical Control CHannel
Technology DPCH Dedicated Physical CHannel
3GPP 3rd Generation Partnership Project DPDCH Dedicated Physical Data CHannel
3xEV-DV 3x Evolution Data and Voice D-RNC Drift-Radio Network Controller
DS Delay Sensitive
A DS-CDMA Direct Sequence-Code Division
AAL2 ATM Adaptation Layer type 2 Multiple Access
AAL5 ATM Adaptation Layer type 5 DSCH Downlink Shared CHannel
ACK ACKnowledgment DTCH Dedicated Traffic CHannel
AICH Acquisition Indicator CHannel DTX Discontinuous Transmission
AM Acknowledged Mode
AMC Adaptive Modulation and Coding E
AMD Acknowledged Mode Data E1 Standard European PCM link (2.048
AMR Adaptive Multi-Rate Mbps)
ARQ Automatic Repeat Query EDGE Enhanced Data for Global Evolution
AS Access Stratum EGPRS EDGE GPRS
ASC Access Service Class
ATM Asynchronous Transfer Mode F
FACH Forward Access CHannel
B FBI FeedBack Information
BCCH1 12 3Broadcast Control CHannel FDD Frequency Division Duplex
BCH W-CDMA R99Broadcast CHannel
Algorithms Description Abbreviations and Acronyms FDMA Frequency Division Multiple Access
COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
RNSAP Radio Network Subsystem Application UARFCN UMTS Absolute Radio Frequency
Part Switch to notes view! Channel Number
RNTI Radio Network Temporary Identity UDP User Datagram Protocol
RRC Radio Resource Control UE User Equipment
RRM Radio Resource Management UM Unacknowledged Mode
RTT Radio Transmission Technology UMTS Universal Mobile Telecommunication
RV Redundancy Version System
RX Receiver / Reception UP User Plane
URA UTRAN Registration Area
S U-RNTI UTRAN-Radio Network Temporary
SA Service Area Identity
SAP Service Access Point UTRAN Universal Terrestrial Radio Access
SAW Stop And Wait Network
S-CCPCH Secondary-Common Control Uu the radio interface between UTRAN
Physical CHannel and UE
SCH Synchronization CHannel
SCR Sustainable Cell Rate V
SDU Service Data Unit VCC Virtual Channel Connection
SF Spreading Factor VoIP Voice over IP
SFN System Frame Number
SHO Soft HandOver W
SIM Subscriber Identity Module W-CDMA Wideband-CDMA
SIR Signal to Interference Ratio
SM Session Management
SNR Signal to Noise Ratio
SPI Scheduling Priority Indicator (CmCH-
1 12 5
PI)
W-CDMA R99 Algorithms Description Abbreviations and Acronyms
COPYRIGHT ALCATEL-LUCENT 2012. ALL RIGHTS RESERVED.
T
TAF That's All Folks!
TB Transport Block
TBS Transport Block Size
TCP Transmission Control Protocol
TDD Time Division Duplex
TDM Time Division Multiplexing
TDMA Time Division Multiple Access
TF Transport Format
TFC Transport Format Combination
TFCI Transport Format Combination
Indicator
TFI Transport Format Indicator
TFO Tandem Free Operation
TFRC Transport Format and Resource
Combination
TFRI Transport Format and Resource
Indicator
TFS Transport Format Set
TPC Transmit Power Control
TrCH Transport CHannel
TrFO Transcoder Free Operation
TS Time Slot
TTI Transmission Time Interval
TX Transmitter / Transmission
Congratulations
You have finished the training
training.feedback@alcatel-lucent.com
Please include the training reference in your email (see cover page)
Thank you!