Вы находитесь на странице: 1из 121

GSM Radio Network Planning and Optimization

Chapter 5 GSM Radio Network Planning

Confideniality level

Table of Contents
Chapter 5 GSM Radio Network Planning ...................................................................................... 3
5.1 Overview ............................................................................................................................. 3
5.2 Planning Foundation ........................................................................................................... 5
5.2.1 Coverage and Capacity Target Confirmation ............................................................ 5
5.2.2 Performance Target Confirmation............................................................................. 6
5.3 Coverage Analysis .............................................................................................................. 7
5.3.1 Area Division ............................................................................................................ 7
5.3.2 Radio Environment Survey ..................................................................................... 10
5.4 Network Structure Analysis ............................................................................................... 11
5.4.1 Middle-Layer Station ............................................................................................... 11
5.4.2 High-Layer Station .................................................................................................. 12
5.4.3 Low-Layer Station................................................................................................... 13
5.5 Traffic Analysis.................................................................................................................. 14
5.5.1 Traffic Prediction and Cell Splitting ......................................................................... 14
5.5.2 Voice Channel Allocation ........................................................................................ 17
5.5.3 Control Channel Allocation ..................................................................................... 20
5.6 Base Station Number Decision ......................................................................................... 23
5.6.1 Characteristics of 3-sector base stations in urban areas ........................................ 23
5.6.2 References for Design of Base Station Parameters................................................ 25
5.6.3 Uplink and Downlink Balance ................................................................................. 27
5.6.4 Cell Coverage Estimation ....................................................................................... 34
5.6.5 Base Station Address Planning .............................................................................. 37
5.6.6 Coverage Prediction ............................................................................................... 39
5.7 Design of Base Station Address........................................................................................ 39
5.7.1 Address design ....................................................................................................... 39
5.7.2 Project Parameter Decision .................................................................................... 42
5.8 Location Area Design ........................................................................................................ 58
5.8.1 Definition of Location Area...................................................................................... 58
5.8.2 Division of location areas ........................................................................................ 58
5.8.3 Others..................................................................................................................... 63
5.9 Dual-Band Network Design ............................................................................................... 64
5.9.1 Necessity for Constructing Dual-Band Network ...................................................... 64
5.9.2 GSM 1800MHz Coverage Solutions ....................................................................... 65
5.9.3 Location Area Division for Dual-Band Network ................................................... 6867
5.9.4 Traffic Guidance and Control Strategies of Dual-Band Network ............................. 69
5.9.5 Dual-Band Networking Engineering Implementation........................................... 7271
5.10 Design of Indoor Coverage System ................................................................................ 75
5.10.1 Characteristics of Indoor coverage ....................................................................... 75
5.10.2 Indoor Antenna System Design ............................................................................ 76
5.10.3 Capacity Analysis and Design .............................................................................. 83
2006-01-05

All rights reserved

Page 1 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

5.10.4 Frequency Planning .............................................................................................. 85


5.10.5 Traffic Control ................................................................................................... 8685
5.11 Tunnel Coverage ............................................................................................................ 86
5.11.1 Characteristic of Tunnel Coverage ....................................................................... 86
5.11.2 Tunnel Coverage Solution ................................................................................ 8887
5.11.3 Tunnel Coverage Based on Coaxial distributed antenna system .......................... 89
5.11.4 Tunnel Coverage Based on Leaky Cable System................................................. 92
5.11.5 Coverage Solutions to Tunnels in Different Length ............................................... 99
5.12 Repeater Planning ........................................................................................................ 101
5.12.1 Application Background ...................................................................................... 101
5.12.2 Working Principles of Repeater .......................................................................... 106
5.12.3 Repeater Network Planning ................................................................................ 108
5.13 Conclusion .................................................................................................................... 120

2006-01-05

All rights reserved

Page 2 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Chapter 5 GSM Radio Network Planning


5.1 Overview
The design of radio network planning (RNP) is the basis of the construction of a
wireless mobile network. The design level of network planning decides the future
layout of a network.
During network planning, the documents concerning base station distribution,
channel assignment, and cell data must be outputted. And the major tasks
involved are as follows:
1)

Analyze carriers requirements on network coverage, capacity and quality.

2)

Analyze the coverage and capacity features of the candidate mobile


communication systems and bands, and then analyze the investment
feasibility through estimating the network scale.

3)

Decide the network structure and base station type based on further
analysis.
First analyze whether to construct a layering network according to user
distribution, propagation conditions, city development plan and existed
network conditions, and then analyze the sites within this area to decide
whether to use omni antennas or directional antennas to meet the
requirements on coverage and capacity.

4)

Estimate the number of base stations


Before estimating the number of base stations, estimate the coverage
distance of base stations of various types in various coverage areas. The
factors deciding the effective coverage area of a base station include:
Valid transmit power of the base station
Working bands to be used (900 MHz or 1800 MHz)
Antenna type and installation position
Power budget
Radio propagation environment
Carriers indexes on coverage
Then through calculating the coverage distance and dividing the coverage
areas, you can obtain a rough number of base stations for various coverage
areas.

5)

Plan an ideal base station address according to cellular structures.


According to geographic maps or administrative maps and with the help of
on-the-spot surveys, you can have a full understanding of the areas to be
planed, and then mark the area where the number of users is large as a
target address. After that, mark the addresses of other base stations
according to the ideal cellular structure and the result of link budget.

6)

Calculate the number of channels of the cells of each base station

2006-01-05

All rights reserved

Page 3 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Estimate the traffic of a base station according to its ideal location, and then
obtain the number of carriers and channels needed by each base station by
checking Erl table according to the indexes of call loss rate.
Decide the frequency reuse mode according to band width, network quality
requirement, and equipment supportability.
Estimate the maximum base station configuration type according to the
frequency bandwidth and reuse mode provided by the construction carriers.
If the system capacity in some areas cannot be met, you need to add more
base stations or cells to the system according to cell splitting principles and
actual conditions. After that, reselect an ideal base station address on the
map and re-estimate the number of channels required by the base station.
7)

Predict the coverage area and decide the project data, namely, perform the
preliminary emulation. The specific tasks are as follows:
Select the design indexes
Select the minimum received power and the penetration ratio index at the
coverage area edge.
Select the design parameters, which includes:
Antenna height (above the ground), antenna azimuth angle, antenna gain,
antenna tilt angle, base station height above sea level, base station type,
feeder length, antenna feeder system loss, combining and distribution
modes, transmitter output power, receiver sensitivity, base station diversity
reception,

and diversity gains.

Predict the coverage area of each cell according to the propagation models
in different areas, and then give the opinions on adjusting the base station
address, antenna direction, antenna tilt angle, and antenna height in the
areas where dead zones may be present and signals are poor. Finally,
provide the project data.
8)

Select actual base station address and decide base station type:
Perform filed examination according to the ideal base station addresses,
and then record the possible addresses according to various construction
conditions

(including

power

supply,

transmission,

electromagnetic

background, and land taken over). Finally, recommend a suitable address


based on integrated consideration of the deviation from the ideal base
station address, the effect on future cell splitting, economic benefits, and
coverage prediction.
After the base station address is selected, decide the actual base station
type according to the number of base station channels.
After the base station type is decided, you need to make a scheme for
antenna configuration. For moving a network, if you intend to provide a best
combination scheme for the antenna feeders, you must fully investigate the
combination of the antenna feeders of the original carriers, plan the future

2006-01-05

All rights reserved

Page 4 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

expansion of the base station, and design the combination of the antenna
feeders supported by current equipments.
9)

Plan frequency and adjacent cell


Decide the frequency and adjacent planning according to the actual base
station distribution and type.

10) Make cell data


To ensure that the network runs stably, you must design the parameters
relative to performance for each cell. These parameters include system
information parameters, handover parameters, power control algorithm
parameters, and so on.
Note:
For the selection of handover bands, the handover algorithms to be enabled,
and whether to use frequency hopping, power control, and DTX, they must
be decided in coverage prediction and frequency planning, because the
related parameters will be used in emulation.

In addition, sections 5.9 and that later introduce the solutions to the planning of
dual-band network and the planning in special occasions.

5.2 Planning Foundation


5.2.1 Coverage and Capacity Target Confirmation
Before planning a network, you must confirm the network coverage and capacity
target and relative specifications from carriers. They are specified as follows:
Definition of coverage areas
Specific division of the service quality in coverage areas
Grade of service (GoS) at Um interface
Prediction of network capacity and subscriber growth rate
Available bands and restrictions on using bands
Restrictions on base station address and the number of carriers
Penetration loss in cars or indoor environment
Performance and sensitivity of base stations
Rules on base station naming and numbering
Information of the base stations in the existing network
Engineers perform the network planning and guide the subsequent construction
work according to the previous technical specifications. Because any change of
these specifications will affect network construction, you must discuss these
specifications with carriers and get their confirmation.

2006-01-05

All rights reserved

Page 5 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

5.2.2 Performance Target Confirmation


Carriers emphasize much on the future network quality. Therefore, network
planning engineers must judge the indexes concerning network performance
according to construction difficulty and experience, and then cooperate with
carriers to design a reasonable solution.
Generally, the performance of voice services can be judged according to KPI
indexes, which are specified in Table 5-1Table 5-1
Table 5-1 Descriptions of KPI indexes
Number

KPI index

TCH

Meaning

congestion

ratio

Test method

TCH

Reference
value

seizure

failures/attempted

OMC

< 2%

OMC

< 1%

OMC

< 2%

OMC

> 92%

Drive test

< 10s

Drive test

> 90%

Applied to GPRS

Drive test

= 16

Applied to GPRS

Drive test

= 3.2

Applied to GPRS

Drive test

< 20s

TCH seizures 100%


SDCCH seizures and

SDCCH congestion

all

busy

ratio

times/SDCCH seizure
requests 100%
TCH

call

drop

times/TCH

Call drop ratio

occupation

success

times 100%
Handover
4

success

Handover success

times/handover

ratio

attempted

times

100%
5

setup

The percentage of the


received level greater

probability

download

call

times

Coverage

FTP
7

Average

Call setup time

than -90 dBm


average
rate

(kbps)
8

FTP

average

upload rate (kbps)


Forward/reverse
transmission delay

10

Ping success ratio

Applied to GPRS

Drive test

= 90%

11

Ping average delay

Applied to GPRS

Drive test

< 3.5s

2006-01-05

All rights reserved

Page 6 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The voice quality is


12

Mean opinion score

divided

(MOS)

levels from excellent

into

fiver

Drive test

=3

to bad.

Note:
The KPI indexes vary slightly with carriers.

The mean opinion score (MOS) in the previous table is divided into five levels,
which are specified in Table 5-2Table 5-2.
Table 5-2 Mean opinion score (MOS)
Quality level

Quality evaluation standard


5

Excellent

Good

Fair

Poor

Bad

Note:
The call whose quality is above level 3 can access the mobile
communication network.
The call whose quality is above level 4 can access the public network.

5.3 Coverage Analysis


5.3.1 Area Division
I. Types of coverage area
The signal propagation models are applied in accordance with the propagation
environments in areas of different types. The signal propagation models decide
the design principles, network structures, grade of services and frequency reuse
modes for the radio networks in coverage areas. In order to decide the cell
coverage area, you can the radio coverage areas into the following four types:
Big city
Middle-sized city
Small town
Countryside

2006-01-05

All rights reserved

Page 7 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Error! Reference source not found. lists the divisions.


Table 5-3 Coverage area division
Area type

Description
Dense population
Developed economy

Big city

Large traffic
Dense high buildings and mansions distributed in
center areas
Flourishing shopping centers
Relatively dense population
Relatively developed economy

Middle-sized city

Relatively large traffic


Dense buildings distributed in center areas
Active and promising shopping centers
Relative large population
Promising economic development

Small town

Moderate traffic
Relative dense buildings distributed in center areas
A certain scale of shopping centers but with great
potentiality
Scattered population

Countryside

Developing economy
Low traffic

In addition, you must consider the coverage of the areas at the intersections and
various transport arteries, including:
Express way
National high way
Provincial highway
Railway
Sea-route
Roads in mountain areas
Generally, it is recommended to apply omni base stations in the countries plains
and the areas with restricted landforms. In big cities, middle-sized cities, and
along expressways, it is recommended to apply directional base stations.

II. Define the field strength at coverage area edges


When defining the field strength of the uplink edges of a service area, you must
consider the factors listed in Table 5-4Table 5-4.

2006-01-05

All rights reserved

Page 8 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Table 5-4 Typical factors concerning the definition for the filed strength at
coverage area edges
Factor

Value

Mobile station sensitivity

-102 dBm

Fast fading protection

4 dB (3 dB for countryside)

Slow fading protection

8 dB (6 dB for countryside)

Noise

(environmental

noise

and

interfering noise) protection

5 dB

Remark:
To ensure the indoor coverage in big and middle-sized cities, you can
consider 15dB for the average penetration loss between buildings and
consider adding 5dB to the protection margin.
Generally, the propagation loss of GSM 1800MHz signals is 8 dB greater
than that of the GSM 900MHz signals in average.
Radio links have two directions, namely, uplink direction and downlink
direction, and the coverage area is defined by the direction in which the
signals are poor, so you must consider the uplink and downlink balance.
Therefore, if you intend to plan an ideal network, you must make a good
power control budget so that the uplink and downlink can be as balance as
possible.

III. Define coverage probability


The definition of coverage probability varies with the coverage areas, and the
coverage probability is gradually improved along with the construction of the
network.
In China, the coverage probability can be defined according to Table 5-5Table
5-5.
Table 5-5 Definition of coverage probability at different stages and in different
areas
Construction
stage

Areas

Coverage target

Significant national tourism areas,


Early stage

expressways,

national

highways,

Full coverage.

and the areas along busy railways.

2006-01-05

All rights reserved

Page 9 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Other major roads, railways and


sea-routes

Confideniality level

The coverage probability


must be greater than
90%.
With the development of

Key areas, such as government


offices, press centers, airport
Development
stage

lounges, waiting rooms of train


stations, subways, commercial
office buildings of high ranks,
entertainment centers, and large
shopping malls.

the

network

construction, the number


of users grows larger
and

they

require

services of higher grade,


so the quality of indoor
coverage of the areas in
the left column must be
greatly enhanced.

Remarks:
Generally, a call must be ensured to access the network at 90% of the places
and 99% of the time within the coverage area.
For the outdoor environment in big cities, the two ratios must be greater.
For the areas in countryside, the two ratios can be lower.
For transport arteries, different standards are applied, and the coverage
probability can be defined in accordance with the types of the arteries.

5.3.2 Radio Environment Survey


Through surveying radio propagation environments, you can get familiar with the
overall landforms, estimate the rough antenna height, and select the proper radio
propagation model, among which the radio propagation model helps you estimate
the number of base station when predicting the coverage. If necessary, you must
adjust the propagation model.
For GSM 900MHz, the formulas estimating radio path loss in different areas are
simplified in Table 5-6Table 5-6.
Table 5-6 Formulas estimating radio path loss in different areas
Formula

2006-01-05

Application

Propagation model

area

adopted

All rights reserved

Example

Page 10 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning
Walfish-Ikegami

Confideniality level

PLDU = 147.2 +

Densely

If carrier frequency = 925

8d + 40.5lgd

populated

MHz, hBTS < hobstacle, and d <

urban areas

0.5km, hBTS = 25mhobstacle =


30m, street width = 25m,
building width = 50m

PLU = 128.73 +

Common

Walfish-Ikegami

38lgd

urban areas

If

carrier

frequency

925MHz and hBTS > hobstacle,


hBTS = 25m, hobstacle = 20m,
street width = 25m, building
width = 50m

PLSU = 126 +

Suburban

35lgd

areas

Okumura-Hata

PLRU = 116 +

Countryside

35lgd

areas

If

carrier

frequency

925MHz, hBTS = 30m


Okumura-Hata

If

carrier

frequency

925MHz, hBTS = 30m

Note:
The four formulas provided in this section are applicable to simple estimation
during project survey only. For later planning, you must adopt the precise
propagation models. If necessary, you must further adjust the propagation models
through CW measurement.

5.4 Network Structure Analysis


When considering the layout of base stations, you must deeply analyze network
structure. Generally, according to network layers, a network can be divided into
middle-layer, high-layer, and low-layer. The base stations at the middle-layer bear
the greatest traffic in a network

5.4.1 Middle-Layer Station


I. Definition and application
A middle-layer station in big and middle-sized cities is defined as follows:
The antenna is installed on building tops.
The antenna height ranges from 25 to 30 meters, which is greater than the
average height of the buildings.
It covers several blocks.

2006-01-05

All rights reserved

Page 11 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

In small towns and countryside areas, except the high-layer stations are designed
for controlling traffic flow or for landform reasons, most of the base stations are
middle-layer stations.

II. Advantages
Compared with high-layer stations, middle-layer stations can utilize frequency
resources more efficiently. Compared with low-layer stations, middle-layer
stations can absorb traffic more efficiently. Therefore, the middle-layer stations
bear the greatest traffic in a network.

III. Distance between stations


The average distance between most middle-layer stations range from 0.6 to 5 km
except in countryside areas. In big cities, the distance between some
middle-layer stations is shorter than 0.6 km. However, it is suggested that the
distance between middle-layer stations in big cities cannot be shorter than 0.4 km.
If this distance is too short, the buildings will produce strong interference against
the signals of the base stations. In this case, to control the coverage area is quite
demanding.

IV. Challenges
Because no suitable ground objective is available, to ensure the quality of service
of a network is quite demanding. According to the experience on project
construction and maintenance, great challenge is present in the selection of base
station address, station design, project construction, network maintenance, and
network quality.

5.4.2 High-Layer Station


I. Definition and application
A high-layer station in big and middle-sized cities is defined as follows:
The antenna height ranges from 10 to 50 meters, which is far greater than
the average height of the buildings.
Its coverage areas contain the areas covered by multiple middle-layer
stations.
Because the high-layer stations make poor use of the frequency resources, they
are mainly applied to the traffic networks where people move fast in big and
middle-sized cities.
In addition, to control construction cost and meet coverage requirements, you can
install some high-layer stations in suburban areas, highroads, small towns, and
countryside areas.

2006-01-05

All rights reserved

Page 12 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

II. Functions
The high-layer stations must be as fewer as possible but be as effective as
possible. They mainly provide services to the fast-moving subscribers in cities.
Note:
The coverage of high buildings is realized by indoor distribution systems.

5.4.3 Low-Layer Station


I. Definition and application
A low-layer station is defined as follows:
The antenna height is shorter than 20 meters, which is shorter than the
average height of the buildings.
The antenna can be installed on the outer walls of the lower floors of a
building, on the top of lower roofs, or in the rooms of a building.
Generally, at the early stage of the network construction, signal network design is
applied, so most of the base stations are middle-layer stations. After the basic
network is established, you must adjust the base stations and add new base
stations according to traffic and coverage requirements.
For populated commercial areas where the traffic is heavy, you can use low-layer
stations, which are constructed with micro cell layer and distributed antenna
system. In this case, not only the requirements on indoor coverage are met, but
also the interference and difficulties of base station selection caused by short
distance between stations are avoided. With the development of the network, the
low-layer stations will develop into the layering network structure.

II. Other considerations


The coverage area of a low-layer station is small, so it can fully use frequency
resources but cannot absorb the traffic efficiently. As a result, ideal traffic cannot
be ensured if the base station deviates far away from the areas where the traffic is
heavy.
Therefore, when constructing a low-layer station, you must consider whether the
base station is used to make up coverage or solve the problem of heavy traffic,
because the construction purpose is directly related to the selection of the
address and type of the base station.
Note:
A layering network cost much frequency resource, so it is not recommended for
the networks where the frequency resource is inadequate.

2006-01-05

All rights reserved

Page 13 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

5.5 Traffic Analysis


5.5.1 Traffic Prediction and Cell Splitting
I. Traffic prediction
The network construction requires the consideration of economic feasibility and
rationality. Therefore, a reasonable investment decision must be based on the
prediction of the network capacity of the early and late stage.
When predicting network capacity, you must consider the following factors:
Population distribution
Family income
Subscription ratio of fixed telephone
Development of national economy
City construction
Consumption policy
After predicting the total network capacity, you must predict the density of
subscriber distribution. Generally, base stations are constructed in urban areas,
suburban areas, and transport arteries. Therefore, you can use the percentage of
prediction method.
At the early stage of construction, the subscribers in cities account for a larger
percentage of the total predicted subscribers. With the development of the
network construction, the percentage of the subscribers in suburban areas and
transport arteries grows. The traffic of each subscriber is 0.025 Erl in urban areas
and 0.020 Erl in suburban areas.
The formula calculating traffic is:
A = (n T) / 3600
Here,
n is the call times in busy hour
T is the duration of each call, in the unit of second.
In this way, the number of voice channels needed for a base station can be
obtained through predicting the traffic.
Note:
When estimating the number of voice channels needed for a base station in the
future, you must consider the effect caused by cell splitting.

In a GSM system, you can use Erl model to calculate the traffic density that the
network can bear. The call loss can be 2% or 5% depending on actual conditions.

2006-01-05

All rights reserved

Page 14 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Because restrictions on cell coverage area and the width of the available
frequencies are present, you must plan the cell capacity reasonably. If good voice
quality is ensured, you must enhance the channel utilization ratio as much as
possible.
In actual networking, if the network quality is ensured at a certain level, two
capacity solutions are available, namely, a few stations with high-level
configuration and multiple stations with low-level configuration. Both the
advantages and disadvantages of the two solutions are apparent, so which one
should be used depending on the actual conditions of an area.
For network construction, you can expand the capacity either through adding
base stations or through expanding the base station capacity. The expansion
strategies adopted must be in accordance with the traffic density in an area. For
example, the strategies such as adding 1800 MHz base stations, expanding
sector capacity, adding micro cells, or improving indoor coverage can be used to
expand network capacity.

II. Cell splitting


Cell splitting is quite effective for the expansion of network capacity. An omni base
station can split into multiple sectors, and a sector can split into multiple smaller
cells. In other word, you must plan cell radius in accordance with the traffic
density of an area.
Cell splitting means more base station and greater cost are needed. Therefore,
when planning a network, you must consider the following factors:
The rules and diagrams of frequency reuse are repeatable.
The original base stations can still work.
The transition cells must be reduced or avoided.
The cell can split without effect.
Cell splitting is quite important in a network. The followings further describe the
cell splitting based on 1-to-4 splitting.
Cell splitting is used to split a congested cell into multiple smaller cells. Through
setting the new cells whose radiuses are smaller than the original cells and
placing them among the original cells, you can increase the number of channels
in a unit area, thus increasing channel reuse times. In this case, system capacity
is expanded.
Through adjusting the project parameters relative to antenna feeders and
reducing transmitter power, you can narrow the coverage area of a cell. Error!
Reference source not found. shows that a cell splits into four smaller cells by
half of its radius.

2006-01-05

All rights reserved

Page 15 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-1 Schematic diagram of cell splitting (1-to-4)


As shown in Figure 5-1Figure 5-1, smaller cells are added without changing the
frequency reuse mode. They are split proportional to the shape of the original cell
clusters.
In this case, the coverage of a service area depends on the smaller cells, which
are 4 times outnumber of the original cells. To be more specifically, you can take a
circle with the radius R as an example, the coverage area of the circle with the
radius R is 4 times that of a circle with the radius R/2.
According to Figure 5-1Figure 5-1, after cell splitting, the number of cell clusters
in the coverage area increases. Thus the number of channels in this coverage
area increases and the system capacity is expanded accordingly.
You can adjust the coverage area of the new cells through reducing the transmit
power. For the transmit power of the new cells whose radiuses are half of that of
the original cell, you can check the power Pr received at the new cell edge and
at the original cell edge, and make them equal. However, you must ensure that
the frequency reuse scheme of the new micro cells is the same as that of the
original cell. As for Figure 5-1Figure 5-1,
Pr [at the edge of the original cell] = Pt1R-n, and,
Pr [at the edge of the new cell] = Pt2 (R/2)-n
Here,
Pt1 and Pt2 are the transmit power of the base stations of the original cell and the
new cell, and n is path fading exponent. If make n = 4, make the received power at
the edge of the new and original cell equal, the following equation can be
obtained:
Pt2 = Pt1/16
2006-01-05

All rights reserved

Page 16 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

That is to say, if the micro cells are used to cover the original coverage area and
the requirement of S/I is met, the transmit power must be reduced by 12 dB.
Not all cells need splitting. In fact, it is quite demanding for carriers to find out a
perfect cell splitting scheme. Therefore, many cells of different scales exist in a
network simultaneously. As a result, the minimum distance among intra-frequency
cells must be maintained, which further complicate frequency allocation.
In addition, you must pay attention to the handover because success handover
ensure the all subscribers to enjoy good quality of service regardless of moving
speed.
As shown in Figure 5-1Figure 5-1, when two layers of cells are present within an
area but their coverage scale is different, according to the formula Pt2 = Pt1/16,
neither all new cells can simply apply the original transmit power, nor all original
cells can simply apply the new transmit power.
If all cells apply great transmit power, the channels used by smaller cells cannot
be separated from the intra-frequency cells. If all cells apply lower transmit power,
however, some big cells will be exclusive from the service areas.
For the previous reason, the channels in the original cells can be divided into two
groups. One group meets the reuse requirement of the smaller cells, and the
other group meets the reuse requirement of the bigger cells. The bigger cells are
applied to the communication of fast-moving subscribers, which requires a fewer
handover times.
The power of the two channel groups decides the progress of cell splitting. At the
early stage of cell splitting, the channels in the low-power group are fewer. As the
requirement grows, more channels are needed in low-power group. The cell
splitting does not stop until all channels within this area are applied in the
low-power group. In this case, all cells in this area have split into multiple smaller
cells, and the radius of each cell is quite small.
Note:
Commonly, you can restrict cell coverage area through adjusting the project
parameters of the base station.

5.5.2 Voice Channel Allocation


I. Voice channel decision
The base station capacity refers to the number of channels that must be
configured for a base station or a cell. The calculation of the base station capacity
is divided into the calculation of the number of radio voice channels and the
calculation of the number of radio control channels.
2006-01-05

All rights reserved

Page 17 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

According to the information of base stations and cells and the density distribution
of subscribers, you can calculate the total number of the subscribers. Then
according to the radio channel call loss ratio and traffic, you can obtain the
number of voice channels that must be configured by checking Erl B table.
Generally, you can decide the number of voice channels as follows:
1)

According to the bandwidth and the reuse mode allowed by current GSM
networks within the areas to be planned, you can obtain the maximum
number of carriers that can be configured for a base station.

2)

Each carrier has 8 channels. You can obtain the maximum number of voice
channel numbers that can be configured for a base station by detracting the
control channels from the 8 channels.

3)

According to the number of voice channels and call loss ratio (generally 2%
dense traffic areas and 5% for other areas), you can obtain the maximum
traffic (Erl number) that the base station can bear through checking Erl B
table.

4)

Through dividing the Erl number by the average busy-hour traffic of


subscribers, you can obtain the maximum number of subscribers that the
base station can accommodate.

5)

According to the data of subscriber density, you can obtain the coverage
area of the base station.

6)

After the areas are specified based on the subscriber density, according to
the area of an area and the actual coverage area of the base station, you
can calculate the number of needed base stations.

7)

For important areas, you must consider back up stations and the cooperation
between carriers. For example, an important county needs at least two base
stations and three important carriers.

8)

For the areas where burst traffic is possible, such as the play ground and
seasonal tourism spots, you must prepare the equipments (such as carriers
and micro cells) and frequency resources for future use.

9)

The dynamic factors, such as roaming ratio, subscriber mobility, service


development, industry competition, charging rate change, one-way charge,
and economic growth, must be considered.

10) To configure a base station, you must consider the transmission at the Abis
interface so that the capacity can be met while saving transmission. For
example, the application and concatenation of the Abis interface 15:1 and
12:1 should be considered.
11) For indoor coverage and capacity, you can use micro cells and distributed
antenna systems. For the coverage in countryside areas and highroads, you
can use economical micro base stations. For the transmission in countryside
areas and highroads, you can use HDSL because it is cost effective.
12) Prepare the some carriers, micro cells, and micro base stations for new
coverage areas and future optimization.
2006-01-05

All rights reserved

Page 18 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

13) In some special areas, you can use the base stations consisting of omni and
directional cells, but you must consider the isolation between omni antennas
and directional antennas. For traffic control, you can use the algorithm in
terms of network layers.
14) For some highroads which require a little traffic by large coverage, you can
use the two networking modes. They are:
(A micro base station with single carrier) + (0.5 + 0.5 cell with two set of
directional antennas)
A micro base station with single carrier + 8-shaped antenna

II. Relationship between carrier number and bearable traffic


Erl traffic model can calculate the traffic that a network can bear. The call loss
ratio can be 2% or 5% according to actual conditions. Table 5-7Table 5-7
describes the relationship between the number of carriers and the traffic that a
network can bear according to Erl B table.
Table 5-7 Relationship between the number of carriers and the traffic that a
network can bear
Number of carriers in

Number of

each cell

TCHs

Traffic (Erl)

2%

5%

2.27

2.96

14

8.2

9.73

21

14.03

16.18

29

21.03

23.82

36

27.33

30.65

44

34.68

38.55

52

42.1

46.53

59

48.7

53.55

67

56.25

61.63

10

75

63.9

69.73

According to this table, the larger the number of carriers and the call loss ratio are,
the greater the traffic that each TCH bear, and the greater the TCH utilization ratio
is (the channel utilization ratio is an important indicator of the quality of network

2006-01-05

All rights reserved

Page 19 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

planning and design). If the number of subscribers of a base station is small, you
can consider delaying the construction.
Because restrictions on the coverage area of a cell and the bandwidth of the
available frequencies, you must plan a reasonable capacity for the cell. If good
voice quality is ensured, you must take measures to enhance the channel
utilization ratio as much as possible.
For the construction of the dual-band network, you can use the frequencies with
wider bands to enhance channel utilization ratio, which is helpful for traffic
sharing.
In actual applications, when the traffic on each TCH accounts for 80-90% of total
given by Erl B table (the call loss ratio is 2%), the congestion ratio in this cell rise
greatly. Therefore, we generally calculate the traffic that a network can bear by
taking the 85% of the traffic given by Erl B table as a reference.

III. Example
The capacity of a local network needs to be expanded. According to the service
development, population growth and mobile popularity, the subscribers in this
area are expected to reach 100,000 in 2 years.
If only the followings are considered:
Roaming factor (according to the development trend of traffic statistics) =
10%.
Mobile factor (the subscriber moves slightly within the local network instead
of roaming) = 10%.
Dynamic factor (with burst traffic considered) = 15%.
The network capacity = 100000 * (1 + 10% + 10% + 15%) = 135,000.
However, because the congestion is present, we generally calculate the traffic
that a network can bear by taking the 85% of the traffic given by Erl B table as a
reference. As a result, the network capacity must be designed as follows:
The network capacity = 135, 000/85% = 158,800, about 160,000.

5.5.3 Control Channel Allocation


I. SDCCH allocation
Stand-alone dedicated channel (SDCCH) is an important channel in a GSM
network. Mobile station activities, such as location update, attach and detach, call
setup and short message, are performed on SDCCH. The SDCCH is used to
transmit signaling and data.
Table 5-8Table 5-8 describes SDCCH configuration.

2006-01-05

All rights reserved

Page 20 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Table 5-8 SDCCH configuration principles


No cell broadcast channel (CBCH)
SDCCH configuration

TRX
number

General cell

Internal cell

Edge cell

SDCCH/4

SDCCH/4

SDCCH/4

SDCCH/8

SDCCH/8

SDCCH/8

SDCCH/4+SDCCH/8

SDCCH/4+SDCCH/8

SDCCH/4+SDCCH/8

2*SDCCH/8

SDCCH/4+SDCCH/8

2*SDCCH/8

2*SDCCH/8

2*SDCCH/8

2*SDCCH/8

SDCCH/4+2*SDCCH/8

2*SDCCH/8

SDCCH/4+2*SDCCH/8

SDCCH/4+2*SDCCH/8

SDCCH/4+2*SDCCH/8

3*SDCCH/8

3*SDCCH/8

3*SDCCH/8

3*SDCCH/8

Cell broadcast channel (CBCH) is present


SDCCH configuration

TRX
number

General cell

Internal cell

Edge cell

SDCCH/8

SDCCH/8

SDCCH/8

SDCCH/8

SDCCH/8

SDCCH/8+SDCCH/4

SDCCH/4+SDCCH/8

SDCCH/4+SDCCH/8

SDCCH/4+SDCCH/8

2*SDCCH/8

2*SDCCH/8

2*SDCCH/8

2*SDCCH/8

2*SDCCH/8

2*SDCCH/8+SDCCH/4

SDCCH/4+2*SDCCH/8

SDCCH/4+2*SDCCH/8

SDCCH/4+2*SDCCH/8

3*SDCCH/8

SDCCH/4+2*SDCCH/8

3*SDCCH/8

3*SDCCH/8

3*SDCCH/8

3*SDCCH/8+SDCCH/4

It is difficult to induce a traffic model for the SDCCH; especially it even becomes
impossible after the large-scale application of layering networks and short
messages. Moreover, the equipments of some carriers support SDCCH dynamic
allocation function. As a result, the traffic model for SDCCH must be adjusted
according to actual conditions.

2006-01-05

All rights reserved

Page 21 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The advantages of the SDCCH dynamic function are as follows:


Adjusting SDCCH capacity dynamically
Reducing SDCCH congestion ratio
Reducing the effect of initial SDCCH configuration against system
performance
Making SDCCH and TCH configuration more adaptive to the characteristics
of cell traffic
Optimizing the performance of the systems under the same carrier
configuration.
In conclusion, the SDCCH dynamic allocation function is divided into two types,
namely,
Dynamic allocation from SDCCH to TCH
Dynamic recovery from SDCCH to TCH

II. CCCH allocation


Common control channels (CCCH) contain access grant channel (AGCH), paging
channel (PCH) and random access channel (RACH). The function of a CCCH is
sending access grant message (immediate assignment message) and paging
message.
All traffic channels in each cell share the CCCH. The CCC can share a physical
channel (a timeslot) with SDCCH, or it can solely occupy a physical channel. The
parameters relative to the CCCH include CCCH Configure, BS AG BLKS PES,
and BS PA MFRMS.
Here,
CCCH Configure designates the type of CCCH configuration, namely,
whether the CCCH shares one physical channel with the SDCCH. If there
are 1 or 2 TRX in a cell, it is recommended that the CCCH occupies a
physical channel and share it with the SDCCH. If there are 3 or 4 TRXs, it is
recommended that the CCCH solely occupies a physical channel. If there
are more than 4 TRX, it is recommended to calculate the capacity of the
paging channels in the CCCH according to actual conditions first, and then
you can perform the configuration.
BS AG BLKS PES indicates that the number of CCCH message blocks
reserved to the AGCH. After CCCH configuration is done, this parameter, in
fact, decides allocates the ratio of AGCH and PCH in CCCH. Some carriers
can set sending priority for the access grant message and paging
message. When the former message set to be prior to the later one, the BS
AG BLKS PES can be set to 0.
BS PA MFRMS indicates the number of multi-frames that can be taken as a
cycle of paging sub-channels. In fact, this parameter decides the number of
paging sub-channels that a cell can be divided into.
2006-01-05

All rights reserved

Page 22 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Note:
In CCCH configuration, the location area planning, paging modes and system
flow control must be considered.

5.6 Base Station Number Decision


After traffic and coverage analysis, according to the selected base station
equipments and parameters, you can obtain the coverage areas of various base
stations through link budget. The coverage area helps you calculate the number
of base stations required by each area. Then you decide the base station
configuration according to traffic distribution. Finally, you must perform emulation
using relative planning software so that coverage, capacity, carrier-to-interference
ratio can be assured and interference can be avoided.

5.6.1 Characteristics of 3-sector base stations in urban areas


Cellular communication is named because the coverage areas of base stations
are extruded through small cellular-shaped blocks. In urban areas, for the
purpose of capacity expansion and radio frequency optimization, mainly 3-sector
base stations are used. This section explains some basic concepts of a 3-sector
base station.
For the concept of the cell radius, see Figure 5-2Figure 5-2.

Figure 5-2 3-sector cellular layout


This is a standard 3-sector cellular layout. According to Figure 5-2Figure 5-2, the
distance between two 3-sector base stations is R + r, here R = 2r. However, R is
mainly used in cell radius estimation because the direction along R is the
2006-01-05

All rights reserved

Page 23 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

direction of the major lobe of the directional antenna. In the design for cellular
layout, however, r indicates the cell radius.
In a cellular cell, if the included angle between a direction and the direction of the
major lobe of the antenna, the coverage distance along this direction is r = R/2,
and the path loss along this direction is about 10dB less than that along the
direction of the major lobe of the antenna (for the deduction, it is introduced in the
following), namely, the equivalent isotropic radiated power (EIRP) along this
direction can be about 10dB less than that along the major lobe.
According to this feature, in the cellular layout of this kind, you can adopt the
directional antenna whose azimuth beam width ranges from 60 to 65 degrees
because their horizontal lobe gain diagram also meets this feature.
If R is the cell radius, the cell area is S = 0.6495 R R. Sometimes the r is
used as cell radius, so the cell area is S = 2 5981rr. Therefore, when
calculating the cell area, you must make clear whether r or R is used.
Figure 5-3Figure 5-3 shows the relationship between R and r.

Figure 5-3 Relationship between R and r


The followings deduce the EIRP required along R direction and r direction.
As shown in Figure 5-3Figure 5-3, the coverage distance along r direction is half
of that along R direction, namely, r = R/2. To keep even coverage, you must
make the field intensity at the edges of the cell equal, namely, RxlvelB =
RxlevelC.
Suppose that the EIPR transmitted from cell A is EIRPR and EIRPr along R
direction and r direction respectively, and the city HATA mode is used for path
loss, the path loss from point A and B is expressed as equation (1) :
EIRPR RXLEVB = 69.55 + 21.66lgf - 13.82lgh1 + (44.9 - 6.55lgh1) lgR

(1)

And the path loss from pint A to point C is expressed as equation (2):
EIRPr- RXLEVc = 69.55 + 21.66lgf - 13.82lgh1 = (44.9 - 6.55lgh1) lgr

(2)

Subtract (2) from (1), the equation (3) is expressed as follows:


EIRPR - EIRPr =(44.9 - 6.55lgh1)(lgR lgr) =(44.9 - 6.55lgh1) lg (R/r)

(3)

Introduce R = 2r, the equation (4) is obtained as follows:


EIRPR - EIRPr = 0.3 (44.9 - 6.55lgh1)
2006-01-05

All rights reserved

(4)
Page 24 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-4Figure 5-4 shows the relationship between antenna height and values
of (EIRPR - EIRPr).

Figure 5-4 Relationship between antenna height and values of (EIRPR - EIRPr)
As shown in Figure 5-4Figure 5-4, when the antenna height h1 increases from
5m to 100m, the values of (EIRPR - EIRPr) decrease from 12 to 9.5, which can be
roughly treated as 10dB.

5.6.2 References for Design of Base Station Parameters


When estimating the number of base stations, you must perform uplink and
downlink budget. Based on the coverage division and propagation environment
survey, you can obtain some project parameters and apply them to link budget.
Table 5-9Table 5-9 lists some recommended base station parameters
Table 5-9 References for base station parameters
Coverage target
Big and middle-sized
cities

Small cities

Highroads

Network type

GSM 900MHz

GSM 900MHz

GSM 900MHz

Antenna gain (dBi)

15

17

18

Small cities

Highroads

GSM 900MHz

GSM 900MHz

GSM 900MHz

25

Coverage target
Big and middle-sized
cities
Network type

Antenna
height

2006-01-05

Densely
populated
urban areas

All rights reserved

Page 25 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning
Other urban
areas
Suburban
areas
Countryside
areas

Confideniality level

30

30

35

35

35

45

45

45

25

20

20

15

15

15

15

10

10

10

Densely
populated
urban areas
Antenna

Other urban

diversity gain

areas

(dB)

Suburban
areas
Countryside
areas
Densely
populated
urban areas

Building

Other urban

penetration

areas

loss (dB)

Suburban
areas
Countryside
areas

Car penetration loss (dB)


Densely
populated
urban areas
Other urban
Slow fading

areas

margin (dB)
Suburban
areas
Countryside
areas
2006-01-05

All rights reserved

Page 26 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Note:
The more densely the base station addresses, the lower the antenna height is.
The building penetration loss in northern cities is greater than that in southern cities.

5.6.3 Uplink and Downlink Balance


After base station parameters are specified, you can perform link budget to
estimate the coverage area of the base station. In addition, you must consider the
sensitivity of the base station equipments at this time.
In a mobile communication system, radio links are divided into two directions,
namely, uplink and downlink. For an excellent system, you must perform a good
power budget so that the balance is present between uplink signals and downlink
signals. Otherwise, the conversation quality is good for one party but bad for the
other party at the edges of the cell. If uplink signals are too bad, the mobile station
cannot start a call even if signals are present.
However, the because the fading for uplink channels and downlink channels is
not totally the same and the other factors such as the difference of the
performances of receivers are present, the calculated uplink and downlink are not
absolute, but the there a fluctuation of 2 to 3 dB.
The measurement report on uplinks and downlinks at the Abis interface can tell
whether the uplink and downlink reach a balance. In addition, dialing tests in
actual network can also tell whether the balance between uplinks and downlinks
are reached. If the conversation quality on downlinks uplinks becomes poor
simultaneously, it means that the downlinks and uplinks are balance.

Note:
Some carriers provide the traffic statistics on uplink and downlink measurement,
which can also tell whether the balance between uplinks and downlinks are
reached.

I. Link budget model


Figure 5-5Figure 5-5 shows the link budget model.

2006-01-05

All rights reserved

Page 27 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-5 Link budget model


When calculating uplink and downlink balance, you must consider the functions of
the tower amplifier first. In a base station receiving system, the thermal movement
of the active parts and radio frequency (RF) conductors cause thermal noise,
which reduces the signal-to-noise ratio of the receiving system. In this case, the
receiving sensitivity of the base station is restricted and the conversation quality is
reduced. To improve the receiving performance of the base station, you can add a
low-noise amplifier under the receiving antenna. And this is the principle of the
tower amplifier.
The contributions of the tower amplifier to uplinks and downlinks are judged
according to the performance of its low-noise amplifier and gain. In fact, it is the
tower amplifier that reduces the noise coefficient of the base station receiving
system. The power amplifier can improve the coefficients for the uplink receiving
system (start from the output end of the receiving antenna). However, if the
functions of the tower amplifier are quantified by this, the uplink improved value
can be represented by the NFDelta (it is the reduced value of the noise coefficient
of the receiving system) after a tower amplifier is added to the system.
(1)

No tower amplifier

When there is no tower amplifier, the sensitivity of the equipments at the duplexer
input interface at the top of the base station cabinet are taken as a reference.
For downlink signals, if,
Mobile station receiver output power = Poutm
Base station diversity received gain = Gdb
Base station receiving level = Pinb
Base station side noise deterioration = Pbn
Antenna receiving gain = antenna transmitting gain (according to reciprocity
theorem)
The following equation can be obtained:

2006-01-05

All rights reserved

Page 28 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Pinb + Mf = Poutm + Gam Ld + Gab + Gdb Lfb Pbn


Generally, Pmn is almost equal to Pbn, so the following equation can be
obtained:
Poutb = Poutm + Gdb + (Pinm Pinb) + Lcb
(2)

With tower amplifier

If a tower amplifier is present, the improved value of the noise coefficients of the
uplink receiving system can be represented by NFDelta, so the equation Poutb =
Poutm + Gdb + (Pinm Pinb) + Lcb can be developed into the following equation:
Poutb = Poutm + Gdb + (Pinm - Pinb) + Lcb + NFDelta
The two equations, Poutb = Poutm + Gdb + (Pinm Pinb) + Lcb and Poutb =
Poutm + Gdb + (Pinm - Pinb) + Lcb + NFDelta are used to calculate base station
transmit power when the uplinks and downlinks are balance. Here,
Pinb is the base station receiving sensitivity
Pinm is the mobile station receiving sensitivity
Gdb (antenna diversity receiving gain) is 3.5dB
According to the requirements in protocols GSM05.05, the mobile station transmit
power and the reference receiving sensitivity of the mobile station and base
station are specified in Table 5-10Table 5-10. At present, however, the
sensitivities in actual systems are greater than the reference values listed in the
following table.
Table 5-10 Base station transmit power and reference receiving sensitivity of
mobile station and base station
Network type

Mobile

station

transmit power

Reference

Reference receiving

receiving sensitivity

sensitivity

of

station (dBm)

mobile

station

of

base

(dBm)
GSM 900MHz

2W (33dBm)

-102

-104

GSM

1W (30dBm)

-100

-104

1800MHz
Note:
From September, 1999 on, the reference receiving sensitivity of mobile station
is -102 dBm as required in GSM protocols. Considering the compatibility of the
previous mobile stations, we adopt -100dBm as the receiving sensibility of the
1800 MHz mobile stations.

2006-01-05

All rights reserved

Page 29 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

II. Bass station sensitivity


This section further introduces the base station sensitivity and the functions of the
tower amplifier.
Receiver sensitivity refers to the minimum signal level needed to by the input end
of the receiver when the certain bit error rate (BER) is met. The receiver
sensitivity detects the performances of the following components:
Receiver analog RF circuit
Intermediate frequency circuit and demodulation
Decoder circuit
Three parameters are used to measure the receiver bit error performance. They
are frame expurgation rate (FER), residual bit error rate (RBER), and bit error rate
(BER). When a fault is detected in a frame, this frame is defined as deleted one.
Here,
FER indicates the ratio of the deleted frames to the total received frames.
For full rate voice channels, the FER is present when the 3-bit cyclic
redundancy check (CRC) detects errors or bad error indication (BFI) is
caused. For signaling channels, the FER is present when the fire code (FIRE)
or other packet codes detect errors. The FER is not defined in data services.
FBER indicates the BER that are not announced as deleted frames, namely,
it is the ratio of the bit errors in the frame detected as good to the total
number of bits transmitted in good frames.
BER indicates the ratio of the received error bits to all transmitted bits.
Because BER occurs at random, the statistical measurement is mainly applied to
measure receiver error rate. That is, sample multiple measuring points on each
channel and when the number of measuring points is certain, if the BER of each
measurement is within the required limit, the BER of this channel meets the BER
as required.
However, the number of sampled measured points and the limit value of the BER
must meet the following conditions:
For each independent sampled measuring point, the times for it to pass a
bad unit must be as fewer as possible, that is, the probability must be
smaller than 2%.
For each independent sampled measuring point, the times for it to pass a
good unit must be as more as possible, that is, the probability must be
greater than 99.7%.
The measurement has vivid statistical features.
The measuring time must be reduced to the minimum.
As a result, you can measure the receiver sensitivity through measuring whether
the receiver BER has reached the requirement while entering sensitivity level to
the receiver.
2006-01-05

All rights reserved

Page 30 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Enter the reference sensitivity level to the receiver according to Table 5-10Table
5-10 in various propagation environments. For the data produced after receiver
demodulation and channel decoding, the indexes for FER, RBER, and BER are
more favorable that that defined in Table 5-11Table 5-11.
Table 5-11 Requirements on static and multi-path reference sensitivity
Propagation condition

Requirement on receiver
sensitivity

Channel type

Static

Parameter

TU50

TU50

RA250

HT100

No

No

Frequency

No

No

frequency

frequency

hopping is

frequency

frequency

hopping

hopping

present

hopping

hopping

FACCH/H

(FER)

0.1%

6.9%

6.9%

5.7%

10.0%

FACCH/F

(FER)

0.1%

8.0%

3.8%

3.4%

6.3%

SDCCH

(FER)

0.1%

13%

8%

8%

12%

RACH

(FER)

0.5%

13%

13%

12%

13%

SCH

(FER)

1%

16%

16%

15%

16%

TCH/F9.6&H4.8

(BER)

105

0.5%

0.4%

0.1%

0.7%

TCH/F4.8

(BER)

104

104

104

104

TCH/F2.4

(BER)

2 104

105

105

105

TCH/H2.4

(BER)

2 104

104

104

104

TCH/FS

(FER)

0.1/a%

6a%

3a%

3a%

7a%

0.4/a%

0.4/a%

0.3/a%

0.2/a%

0.5/a%

2%

8%

8%

7%

9%

Class Ib.
(RBER)
Class II
(RBER)
Note:

The requirements on BCCH, AGCH, PCH, and SACCH are the same as that on SDCCH.
The value of a in this table depends on the channels. It is 1 for base stations, and 1 to
1.6 for mobile stations.

III. Contributions of tower amplifier to base staiton sensitivity


In terms of technical principles, the tower amplifier reduces the noise coefficients
of the base station receiving system, which is helpful for improving the sensitivity
of the base station receiving system.
2006-01-05

All rights reserved

Page 31 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

In an actual system, to improve the receiving performance of the base station,


you can add a low-noise amplifier near the feeder of the receiving antenna.
In a mobile communication system, the receiver sensitivity = noise spectrum
intensity (dBm/Hz) + bandwidth (dBHz) + noise coefficient (dB) + C/I (dB).
Here the noise spectrum intensity, bandwidth, and noise coefficient are system
thermal noise. C/I is the signal-to-noise ratio required at the Um interface. In a
narrow band system, C/I indicates the modulation performance required by the
receiver baseband, and it is a positive number.
In a spreading communication system, because spread spectrum gain is present,
the value of C/I is far beyond the requirement of the modulation performance of
the receiver baseband, and it is a negative number.
When there are n* cascaded receivers, the equivalent noise coefficient is as
follows:

F1

F2 1
G1

F3 1
G1G2

Fn
G1G2

1
Gn

Here,
Gn indicates the receivers gain at each level (including the loss at each
level).
Fn indicates the noise coefficient of the receivers at each level.
The noise coefficient of the passive device is equal to its loss, and the gain of the
passive device is the reciprocal of the loss.
According to the previous equation, the noise coefficient of the cascading system
is determined by the receivers at the first level.
It must be pointed out that the linear values of the parameters must be applied in
the previous equation, so the F is a linear value, which must be converted into a
logarithm. Moreover, according to this equation, the noise the cascaded receivers
are determined by the noise coefficient (F1) of the receivers at the first level.
However, when the tower amplifier stops working, because the loss is present on
duplexer and bypass connectors, about 2dB of redundant loss is introduced on
reverse link.

F1

According to the equation

F2 1
G1

F3 1
G1G2

Fn
G1G2

1
Gn , the

following two assumptions conclude the regularity of the effect of tower amplifier
on the base station system.
(1)

Assumption 1

Hereunder is a series of assumptions:


F1 = 2.5 dB (1.7783), noise coefficient of the tower amplifier
2006-01-05

All rights reserved

Page 32 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

F2 = 4.5 dB (2.8184), noise coefficient of the base station


G = 2 (15.849) dB, tower amplifier gain
Loss of the feeder and other passive devices = 3 dB (2)
Gain of the feeder and other passive devices G0 = 3 dB (1/2)
Noise coefficient of the feeder and other passive devices F0 = 1/G0
When the tower amplifier is not added, the noise coefficient of the base station
receiving system with the antenna output end as reference point is as follows:
F = F0 + (F21)/G0 = 10*log (2 + (2.81841)/0.5) =7.5dB
When the tower amplifier is added, the noise coefficient of the base station
receiving system with the antenna output end as reference point is as follows:
F = F1 + (F0 1)/G + (F2 1)/(G*G0) = 10*log(1.7783 + (2 1)/15.849 + (2.8184
1)/(15.849 0.5) = 3.2dB
At this time, the added tower amplifier improves the noise coefficient, and FDelta is
4.3dB, that is, the uplink is improved by 4.3 dB.
(2)

Assumption 2

Hereunder is a series of assumptions:


F1 = 2.2 dB (1.6596), noise coefficient of the tower amplifier
F2 =2.3 dB (1.6982), noise coefficient of the base station
G = 12 (15.849) dB, tower amplifier gain
Loss of the feeder and other passive devices = 3 dB (2)
Gain of the feeder and other passive devices G0 = 3 dB (1/2)
Noise coefficient of the feeder and other passive devices F0 = 1/G0
When the tower amplifier is not added, the noise coefficient of the base station
receiving system with the antenna output end as reference point is as follows:
F = F0 + (F2 1)/G0 = 10*log (2 + (1.6982 1)/0.5) = 5.3dB
When the tower amplifier is added, the noise coefficient of the base station
receiving system with the antenna output end as reference point is as follows:
F = F1 + (F0 1)/G + (F2 1)/(G*G0) = 10*log(1.6596+(2 1)/15.849 + (1.6982
1)/(15.849 0.5)) = 2.6dB
At this time, the added tower amplifier improves the noise coefficient, and FDelta is
2.7 dB, that is, the uplink is improved by 2.7 dB.
According to the previous calculation, the following conclusions can be obtained:
The tower amplifier improves the noise coefficient of the base station
receiving system, thus improving the receiving sensitivity of the base station.
The tower amplifier improves uplink signals effectively, which is also helpful
for improving the receiving sensitivity of the base station.
The gain of the antenna amplifier reduces the effect of the components
installed behind the tower amplifier against noise coefficient.

2006-01-05

All rights reserved

Page 33 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

When the feeder is long and the loss of the feeder is great, if the tower
amplifier is added, the noise coefficient of the base station receiving system
and the uplink signals will be greatly improved.
The smaller the noise coefficient of the tower amplifier is, if the tower
amplifier is added, the greater the noise coefficient of the base station
receiving system is improved. However, if the noise coefficient of the tower
amplifier is too great, it may cause the noise coefficient of the base station
receiving system to deteriorate.
When the receiving sensitivity of the base station is great and the feeder is
short, the tower amplifier makes a little improvement on the noise coefficient
of the base station.
If the tower amplifier improves the base station sensitivity, the base station is
more sensitive to outside interference.

5.6.4 Cell Coverage Estimation


In actual project planning, the effective coverage area of a base station largely
depends on the following factors:
Effective base station transmit power
Working band (900MHz or 1800MHz) to be used
Antenna type and location
Power budget
Radio propagation environment
Carriers; coverage requirements
Based on the indexes of QoS for the mobile network and the actual applications,
this section introduces the coverage area of the base station in different
environments theoretically.
Table 5-12Table 5-12 lists the assumptions of the minimum received level
required in various environments.
Table 5-12 Assumptions of the minimum received level required in various
environments
Application
environments

2006-01-05

Minimum
received level

Other indexes

(dBm)

All rights reserved

Page 34 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level
Mobile station sensitivity: -102 dBm
Fast fading protection: 3dB

The mobile

Slow fading protection (indoor): 7dB

station works as
the receiver.
The first floor of

(the standard deviation is 7dB for


-70

indoors and 8dB for outdoors, the pass


ratio is 90% in coverage areas)

the high buildings

Penetration loss: 18dB

in big cities

Interference noise: 2dB


Environment noise protection: 2dB

The mobile

Mobile station sensitivity: -102 dBm

station is the

Fast fading protection: 3dB

receiver.

Slow fading protection (indoor): 5dB

In cars.
The first floor of

Penetration loss: 10dB

-80

Interference noise: 2dB

the general

Environment noise protection: 2dB

buildings in urban
areas.
Mobile station sensitivity: -102 dBm
Fast fading protection: 3dB
Outdoors.

-90

Slow fading protection (indoor): 5dB


Interference noise: 2dB
Environment noise protection: 2dB

If the following assumptions are present:


The antenna height of GSM 900MHz and GSM 1800MHz base stations are
30 meters.
The sensitivities of the GSM900 MHz 2W (33 dBm) mobile station and GSM
1800MHz 1W (30 dBm) mobile station are -102 dBm and -100 dBm
respectively.
The mobile station height is 1.5 meters and the gain is 0 dB.
When the combiner and divider unit (CDU) is used, the sensitivities of the
900MHz base station and 1800MHz base station are -110dBm and -108dBm
respectively.
The CDU loss is 5.5dB, and the SCU loss is 6.8dB.
The gain of the 65-degree directional antenna is 13dBd for the 900 MHz
mobile station and 16dBd for the 1800MHz mobile station.
The feeder is 50m in length. For 900MHz signals, the feeder loss is
4.03dBm/100m. For 1800MHz signals, the feeder loss is 5.87dB/100m.
In general cities, select Okumura propagation model.
No tower amplifier and the downlinks are restricted according to the
calculation of the uplink and downlink balance.

2006-01-05

All rights reserved

Page 35 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

According to the previous assumptions, the calculated results are as follows:


(1) Outdoor coverage radius of the 900 MHz base station in urban areas
90 dBm. The
The minimum received level of the mobile station Pmr min
coverage radius is calculated according to the maximum TRX transmit power.
The maximum TRX transmit power for the 900 MHz base station Pbt 40 W (46
dBm).
The EIRP of the base station antenna is:
EIRP Pbt L com L bf Ga b 46 5.5 2.01 13 2.15

53.65 (dBm)

Here,
LCOM indicates the combiner loss
Lbf indicates the feeder loss
Gab indicates the antenna gain of the base station
And the allowed maximum propagation loss is:

Lp

EIRP Pmr min

53.65 ( 90 )

143.65 (dB)

According to the Okumura propagation model introduces earlier,

Lp

69.55 26.16 lg f 13.82 lg h b (44.90 6.55 lg h b ) lg d A h m

Here,

h b indicates the antenna height of the base station.


h m indicates the antenna height of the mobile station.
f = 900 MHz.
A h m (1.1 lg f 0.7 )h m (1.56 lg f 0.8 )

0.01 (dB)

According to the previous known number, the outdoor coverage radius of the 900
MHz base station in urban areas can be obtained, that is, d = 2.8km.
(2) Coverage radius of the 900 MHz base station in urban buildings
The minimum received level of the mobile station Pmr min

Lp

EIRP Pmr min

53.65 ( 70 )

70 (dBm).

123.65 (dB)

Therefore, the coverage radius of the 900 MHz base station in urban buildings
can be obtained, that is, d = 0.75km.
If the previous assumptions are present, this indicates that the 900 MHz base
station can cover the outdoor areas 2.8 km away, but for the subscribers on the
first floor of the buildings 750 m away, the quality of the received signals is not
satisfying.
(3) Coverage radius of the 900 MHz base station in suburban areas
The minimum received level of the mobile station Pmr min

Lp

EIRP Pmr min

53.65 ( 90 )

90 (dBm).

143.65 (dB)

The Okumura propagation model in suburban areas must be modified as follows:


2006-01-05

All rights reserved

Page 36 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Lp

Confideniality level

69.55 26.16 lg f 13.82 lg h b (44.90 6.55 lg h b ) lg d


Ahm

2[lg(f/28 )] 2 5.4

Therefore, the coverage radius of the 900 MHz base station in urban areas can
be obtained, that is, d = 5.4km, so it is obvious that the coverage radius of the
base station with the same configuration is larger in suburban areas that in urban
areas.
(4) Outdoor coverage radius of the 1800 MHz base station in urban areas
The minimum received level of the mobile station Pmr min 90 (dBm). Because
the maximum transmit power of the 1800 MHz TRX is 40W (46dBm), the
coverage radius is calculated based on this maximum transit power.

EIRP Pbt
Lp

L com L bf

EIRP Pmr min

Ga b

46 5.5 2.93 16 2.15

55.73 (dBm)

145.73 (dB)

For the 1800 MHz base station, the Okumura propagation model is:

Lp

46.3 33.9 lg f 13.82 lg h b (44.90 6.55 lg h b ) lg d A h m


(1.1 lg f 0.7 )h m (1.56 lg f 0.8 )

In addition, f = 1800 MHz and A h m

0.04 (dB).

According to the previous known number, the outdoor coverage radius of the
1800 MHz base station in urban areas can be obtained, that is, d = 1.7km.
(5) Coverage radius of the 1800 MHz base stations in urban buildings
The minimum received level of the mobile station P mr min

Lp

EIRP Pmr min

55.73 ( 70 )

70 (dBm).

125.73 (dB)

If the previous assumptions are present, this indicates that the 1800 MHz base
station can cover the outdoor areas 1.7km away, but for the subscribers on the
first floor of the buildings 500m away, the quality of the received signals is not
satisfying.

5.6.5 Base Station Address Planning


I. Overview
When planning base station addresses, first you must estimate the number of the
base stations needed in various coverage areas according to the coverage
distance and the divisions of the coverage areas. For the convenience of
prediction and emulation, you must plan an initial layout the base station
addresses with the help of maps and the estimated results.

II. Planning methods


The base station address can be planned based on standard girds, or it can be
planned from a specific area.
(1)

Plan base station address based on standard grids

2006-01-05

All rights reserved

Page 37 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

First you set the base stations in the coverage areas according to the distance of
the standard grids, and then adjust the address layout and project parameters
according to the estimated coverage results to meet the coverage requirement.
After that, continue the planning according to the following instructions:
If a satisfying address layout is obtained, you must analyze the capacity of
the base stations to be planned according to this layout, and determine the
reasonable number of base stations. When designing the capacity, you must
calculate the number of TRXs needs to be configured for each base station,
and then analyze and adjust the configuration of the base station according
to the number of the configured TRXs.
The adjustment of the configuration of the base station is determined by
subscriber distribution. If the number of base stations in some areas does
not meet capacity requirement, another base stations must be added.
(2)

Plan base station address based on a specific area

According to this method, you are required to start the planning from the areas
where the subscribers are most densely distributed or the planning work is quite
hard to be performed. As a result, you must fully survey the subscriber distribution,
landforms, and ground objectives within the coverage area to position the key
coverage area where the center base stations should be planned. And these
center base stations function as ensuring the coverage and capacity in important
areas.
After the layout of these center base stations is determined, you can plan other
base station addresses according to coverage and capacity target. And this is
how the final layout of the base station addresses come from. After the overall
solution is determined, the subsequent steps are performed according to the first
planning method.

Note:
The difference of the traffic intensity and the abnormality of the landforms
and ground objectives result in irregularity of the radio coverage. Therefore,
the distance between base stations varies. Generally, this distance is smaller
in the areas where traffic intensity is great. In some hot areas, you can
ensure the system capacity by using micro cells and distributed antennas to
provide multi-layer coverage.
For restrictions from frequency resources are present, you must consider
avoiding interference while ensuring system capacity.
There is no standard available for the layout of the base station addresses. A
good planning solution is selected based on the integrated performance of
the network.

2006-01-05

All rights reserved

Page 38 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

5.6.6 Coverage Prediction


The coverage prediction is to predict the coverage of the network to be
constructed according to the selected base station addresses, designed base
station types, suitable electronic maps, and network planning tools to judge
whether the coverage meet the requirements of the subscribers.
The coverage of a base station is determined by the following factors:
Indexes of QoS
Output power of transmitters
Available sensitivity of receivers
Direction and gain of antennas
Working bands
Propagation environment (such as landforms, city constructions)
Application of diversity reception
If the predicted results of the network coverage fail to meet the requirements, you
can take the following adjusting measures:
When there are subscribers distributing beyond the cell coverage area, but it
is not economical for you to install a base station, you can use a repeater to
ensure the requirement of those subscriber.
When the signals are weak or blind zones are present within the coverage
area, you can consider whether to use micro cells according to actual
conditions.
If a large blank area is present between neighbor cells, you can increase the
antenna height and add base stations according to the principles of cell
splitting.
When the cell coverage area fails to meet the co-channel interference index,
you can adjust the frequency configuration of the cell, adjust base station
addresses, or adjust design of the parameters, such as antenna
specification, antenna height, azimuth angle, tilt angle, and transmit power.
Note:
When taking these adjusting measures, you must consider the mutual effect
between base stations.

5.7 Design of Base Station Address


5.7.1 Address design
Generally, in GSM radio network planning, the base station address is designed
according to the following requirements:
2006-01-05

All rights reserved

Page 39 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The address must serve to the reasonable cell structure.


Based on the comprehensive analysis of the electronic maps and paper
maps, you can select several candidate addresses from the perspective of
coverage, anti-interference, and traffic balance.
In actual conditions, carriers are required to discuss the selected addresses
with owners. Generally, the addresses must be located within the area 1/4
radius of the cellular base station.
During the early construction stage when only a few base stations are
installed, the base stations must be located in the center of the areas where
subscribers are densely populated.
For the selection of the base station addresses, the priority must be given to
the important areas, such as government offices, airports, train stations,
news center, and great hotels so that good conversation quality can be
assured. Furthermore, overlapped coverage must be avoided in these
areas.
For other coverage areas, the base station addresses are designed
according to standard cellular structures. For the suburban areas, highroads,
and countryside areas, the design of base station addresses has little
relation with cellular structures.
Without affecting the layout of base stations, you can select the
telecommunication buildings and post offices as the base station addresses
so that the facilities, such as the equipment room, power supplier, and iron
tower can be fully utilized.
The direction of antenna major lobe must be in accordance with the area
where the traffic intensity is great. In this case, the signal strength of the area
can be enhanced, so does the conversation quality. Meanwhile, the direction
of the antenna major lobe must be deviated from intra-frequency cells so that
the interference can be controlled efficiently.
In urban areas, it is recommended that the overlapped depth of the antennas
in adjacent sectors cannot excel 10%. In suburban areas and small towns,
the overlapped depth between coverage areas cannot be too great, and the
included angle between sectors must be equal to or higher than 90.
In addition, for actual design, you must consider the mapping relationship
between carrier number and cells. Generally, more carriers are configured for
the cells with high intensity.
The azimuth angle must be designed according to not only the traffic
distribution in the areas around the base stations, but also the performance
of the overall network.
Generally, it is recommended to adopt the same azimuth angle for the
3-sector base stations in urban areas so that the complicated network
planning can be avoided after cell splitting in the future. Moreover, the
2006-01-05

All rights reserved

Page 40 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

antenna major lobe cannot directly point to the straight streets in populated
urban areas, because it can cause cross-coverage.
In the areas connecting urban and suburban areas, and along transport
arteries, you must adjust the azimuth angle according to coverage target.
Generally, the base station address is not considered on the high mountains
in urban and suburban areas. To be more specifically, the high mountains
are those over 200 to 300 meters higher than above the sea-level).
Otherwise, not only strong interference and weak signals may be present
within the coverage area, but also the base stations are hard to be installed
and maintained on high mountains.
New base stations must be installed at the spots where the traffic is
convenient, the power supply is available, and the environment is secure. In
contrast, new base stations must not be installed at the spots near the radio
transmit stations with high power, radar stations, and other equipments
which produces great interference, because the interference-field intensity
cannot be greater than that defined by the base station.
The base station addresses must be far away from forests or woods to keep
the receiving signals from fading.
The transmission between base station controllers must be considered in the
design of the base station address.
When selecting a base station address from high buildings in urban areas,
you can divide the network into several layers with the help of the building
height. The antenna height of major base stations must be a little higher than
the average height of buildings. Generally, the antenna height of the base
stations in populated urban areas ranges from 25 to 30 meters. In suburban
areas (or the antenna points to suburban areas), the antenna height ranges
from 40 to 50 meters.
Along highroads or in mountain areas, the base station address is selected
based on full survey of the landforms. For example, the address can be
determined in an open area or at the turns of the highroads.
When selecting a base station address from the cities characterized by
mountains and hills and from the areas where high buildings are constructed
with metals, you must consider the effect of time dispersion. In this case, the
base station address must near reflected objectives. When the base station
is far away from reflected objectives, you must adjust the directional antenna
to the reverse direction of the reflected objectives.

2006-01-05

All rights reserved

Page 41 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Caution:
Time dispersion mainly refers to the intra-frequency interference arising from the
time difference between the master signal and other multipath signal arriving at
the receiver in terms of space transmission. According to the requirements in
GSM protocols, the equalizer of the receiver must carry the time window with
16s (equivalent to 4.8 km). The multipath signal with time difference greater than
16 s is regarded as intra-interference signal. In this case, you must consider
whether the level difference between the master signal and multipath signal meet
the carrier-to-interference ratio (C/I), namely, the master signal is 12 dB greater
than the multipath signal at least.

5.7.2 Project Parameter Decision


After finishing designing a base station address, you must decide the project
parameters needed for the base station installation. These parameters include:
Latitude and longitude of the location of base station antenna
Antenna height
Directions of the antenna
Antenna gain
Azimuth angle
Tilt angle
Feeder specifications
Transmit power for each cell of the base station
And the previous parameters are decided through field survey.
Before beginning field survey, you must familiarize yourself with the overall
project and collect the materials and tools relative to the project. They are:
All types of project documents
Background information
Information about the existing network
Local map
Configuration lists required in contracts
Relative tools (including digital camera, GPS, compass, ruler, and laptop
computer)
Note:
Make sure that all the materials and tools are usable before setting out.

The following items must be emphasized before field survey:


2006-01-05

All rights reserved

Page 42 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The GPS must be placed in an open land to position the latitude and
longitude of a base station
Make a detailed record of the surroundings around the base station, such as
the distribution of the buildings, facilities with strong interference, and the
equipments sharing the same base station address.
It is better to record the previous information with a camera.
Prevent the compass from magnetizing, because the magnetization will
cause great deviation during the measurement.
Field survey determines the layout of the base station addresses ultimately. The
field survey for the base station includes optical measurement, spectrum
measurement, and base station address survey. They are specified as follows:
Optical measurement
Measure if a barrier that may reflect electrical waves around the base station,
such as high buildings.
Spectrum measurement
Check if the electromagnetic environments around the base stations are
normal at present or in recent days.
Base station address survey
Check the installation conditions of antenna and equipments, power supply,
and natural environment.
The following sections introduce the design for antenna installation.

I. Environment for antenna installation


The environment for antenna installation can be divided into the environment
near the antenna and the base station. For the environment near the antenna,
you must consider the isolation between antennas and the effect of iron tower
and buildings against the antenna. For the environment near the base station,
you must consider the effect the high buildings within 500 meters against the
base station. However, if the height of the buildings is properly used, you can
obtain the intended coverage area.
If a directional antenna is installed on the wall, the radiation direction of the
antenna is perfectly perpendicular to the wall. If its azimuth angle must be
adjusted, the included angle between the radiation direction and the wall is
required to be greater than 75. In this case, if the front-to-back ratio of the
antenna is greater than 20 dB, the effect of the signals reflected by the wall in
reverse direction against the signals in the radiation direction is quite slight, as
shown in Figure 5-6Figure 5-6.

2006-01-05

All rights reserved

Page 43 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-6 Included angle between antenna and wall


When installing an antenna, you must consider whether large shadows will be
present within the coverage area of the antenna. The shadows are produced
mainly because the base station is surrounded by some huge barriers, such as
high buildings and great mountains. Therefore, the antenna must be installed in
the areas with no such barriers.
When a directional antenna is installed on building roofs, you must prevent the
building edges from barring the radiation of antenna beams. Therefore, to reduce
or ease the shadow, you can install the antenna near building edges.
Because the building roofs are diversified and complicated, if an antenna must be
installed far away from building edges, the antenna must be installed higher than
the roof. In this case, the wind load of the antenna must be considered.
Table 5-13Table 5-13 and Table 5-14Table 5-14 lists the recommended height
between antenna and roof for GSM 900 MHz and GSM 1800 MHz without the
consideration of the effect of the antenna tilt angle.
Table 5-13 Recommended height between antenna and roof for GSM 900 MHz
Distance between antenna and building
edge (m)

Height between antenna and roof (m)

01

0.5

1 10

10 30

> 30

3.5

Table 5-14 Recommended height between antenna and roof for GSM 1800 MHz
Distance between antenna and building
edge (m)

02
2006-01-05

Height between antenna and roof (m)

0.5
All rights reserved

Page 44 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

2 10

> 10

II. Antenna isolation in GSM system


To avoid inter-modulation interference, you must leave certain isolation between
the receiver and transmitter of the GSM base station, namely, Tx - Rx: 30 dB and
Tx -Tx: 30 dB. They are applicable to the situation that a GSM 900MHz base
station and a GSM 1800MHz base station share the same address.
The antenna isolation depends on the radiation diagram, space distance, and
gain of the antenna. Generally, the attenuation introduced by the voltage standing
wave ratio (VSWR) is not considered. The antenna isolation is calculated as
follows:
For vertical arrangement, Lv = 28 + 40lg (k/ ) (dB)
For horizontal arrangement, Lv =22 + 20lg (d/ ) (G1+G2) (S1 + S2) (dB)
Here,
Lv indicates the required isolation.
indicates the length of carrier waves.
k indicates the vertical isolation distance.
d indicates the horizontal isolation distance.
G1 indicates the gains of the transmitter antenna in the maximum radiation
direction, in the unit of dBi.
G2 indicates the gains of the receiver antenna in the maximum radiation
direction, in the unit of dBi.
S1 indicates the levels of the side lobes of the transmitter antenna in the 90
direction, in the unit of dBp, and it is a negative value relative to the main
beam.
S2 indicates the levels of the side lobes of the receiver antenna in the 90
direction, in the unit of dBp, and it is a negative value relative to the main
beam.
Table 5-15Table 5-15 lists the values of S for antennas of different types.
Table 5-15 Values of S for various antennas
Antenna type

Value of S (dBp)

65fan-beam antenna

-18

90fan-beam antenna

-9

120fan-beam antenna

-7

2006-01-05

All rights reserved

Page 45 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Note:
The value of S can be determined according to the directional diagram of the
antenna. When the omni antenna is used, the S is 0.

The followings introduce the requirements on the antenna mount in GSM


900MHz and GSM 1800MHz.
(1)

Directional antenna

In one system, the following requirements must be met in terms of isolation:


The horizontal distance between two antennas in the same sector must be
equal to or greater than 0.4m.
The horizontal distance between two antennas in different sectors must be
equal to or greater than 0.5m.
In different systems, the following requirements must be met when two antennas
are in the same sector and direction:
The horizontal distance between the two antennas must be equal to or
greater than 1m.
The vertical distance between the two antennas must be equal to or greater
than 0.5m.
The distance between the bottom of the antennas and the enclosing wall of
building roof must be equal to or greater than 0.5m.
The included angle between the line connecting the bottom of the antenna to
the antenna-facing roof and the horizontal direction must be greater than
15.
In addition, the included angle between the line connecting the two antenna
mounts and the antenna direction are required in Table 5-16Table 5-16.
Table 5-16 Range of the included angle between the line connecting two antenna
mounts and antenna direction
Antenna lobe width in horizontal plane (degree)
Included angle between mount-connecting line
and antenna direction (degree)

(2)

60-70

90

90

> 40-50

> 55

> 70

Omni antenna

The antenna separation for omni antennas is required as follows:


The horizontal distance between antennas must be equal to or greater than
10m or the vertical distance between antennas.
The vertical distance between the antennas must be equal to or greater than
0.5m.
2006-01-05

All rights reserved

Page 46 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The distance between the bottom of the antennas and the enclosing wall of
building roof must be equal to or greater than 0.5m.

III. Anrenna isolation of GSM and CDMA base station


The analysis of the interference in CDMA and GSM systems must be based on
the frequency relationship and the transmitting and receiving features of the two
systems. Three types of interference are present, namely, spurious interference,
congestion interference and cross-modulation interference, among which the
spurious interference affects the system most. Therefore, the spurious
interference is a key concern in network design. Compared with the spurious
interference, the cross-modulation interference and congestion interference has
little effect on the network, so they are not introduced hereunder.
The followings describe the spurious interference of the CDMA2000 1X against
the GSM 900MHz.
Currently, bands of China Unicoms CDMA2000 1X and that of the GSM 900MHz
are listed in Table 5-17Table 5-17
Table 5-17 Bands of China Unicoms CDMA2000 1X and that of GSM 900MHz
System

BTS transmitting band (MHz)

BTS receiving band (MHz)

GSM 900MHz

945-960

890-915

CDMA

870-880

835-835

As listed in Table 5-17Table 5-17, the bands of the two systems are close to each
other, the interference against each other will easily occur. Mostly, the
transmission of CDMA2000 1X base station will interfere with the reception of
GSM 900MHz base station.
The disclosure signals of the CDMA band falling into the channels of the GSM
base station receivers will enhance the noise level of the GSM receivers. In this
case, the GSM uplinks become weak, which will reduce the coverage area of the
base station and worsen the quality of the network.
If there is not enough isolation between base stations or the transmitting filter
interfering base stations does not provide enough out-of-band attenuation, the
signals falling into the band of the interfered base station receiver may strong,
which will increase the noise level of the receiver.
The deterioration of the system performance is closely related to the strength of
interference signals, and the strength of interference signals is determined by the
factors, such as the performance of the transmitting elements of the interfering
base stations, the performance of the receiving elements of the interfered base
stations, the distance between bands, and the distance between antennas.
2006-01-05

All rights reserved

Page 47 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-7Figure 5-7 shows an interference model.

Figure 5-7 Interference model diagram


According to Figure 5-7Figure 5-7, the signal from the amplifier of the interfering
base station is first sent to the transmitting filter, and then it attenuate due to the
isolation between the two base stations. Finally, it is received by the receiver of
the interfered base station. The power of the spurious interference arriving at the
antenna end of the interfered base station can be expressed by the following
equation:

I b = PTX

AMP -

Pattenuation - I isolation + 10 lg

WB int erf ered


WB int erf ering

Here,
Ib indicates the interference level received at the antenna receiving end of
the interfered base station, in the unit of dBm.
PTX-AMP indicates the output power at the amplifier of the interfering base
station, in the unit of dBm.
Pattenuation indicates

the out-of-band

suppression

attenuation

at the

transmitting filer.
Iisolation indicates the isolation between the antennas of the two base stations,
in the unit of dB.
WBinterfered indicates the bandwidth of the signals at the interfered base
station.
WBinterfering indicates the measurable bandwidth of the interfering signals, or it
can be understood as the bandwidth defined by spurious radiation.
Regulate the previous equation and the following equation can be obtained:

I isolation = PTX

AMP -

Pattenuation - I b + 10 lg

WB int erf ered


WB int erf ering

Suppose the transmit channel number of CDMA2000 1X is the last one on its
working band, that is, 878.49MHz, the spurious signal level on the band of
890-915MHz must be equal to or lower than -13dBm/100kHz. If you intend to put
2006-01-05

All rights reserved

Page 48 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

this assumption into practice, you can filter and combine each transmitted
channel number by using band-limited filter with a bandwidth of only 1.23MHz. The
band-limited filter of this type has great out-of-band attenuation, which can reach 56 dB at

890 MHz and 80 dB at 909 MHz. Here you must consider the worst situation, that
is, the frequencies at the highest end of the CDMA system interfere with the
frequencies at the lowest end of the GSM system.
In this case, Iisolation = (-13dBm/100kHz) - 56 - Ib + 10lg (200kHz/100kHz)

Here Ib indicates the highest interference level (dBm) allowed by the receiving end
of the interfered base station. If the receiving sensitivity of the interfered base
station is ensured, the outside interference level are required to be 10 dB lower
than the back noise of the receiver. In this case, the sensitivity affected only
accounts to about 0.5 dB.
The back noise of the GSM receiver is the sum of the noise intensity, bandwidth,
and noise coefficient. If the noise coefficient is 8 dB, the back noise is -174+noise
coefficient+10lg (200000) = -174+8+53 = -113 (dBm). Therefore, the maximum
spurious interference allowed is -113-10 = -123 (dBm/200kHz).
As a result, the spurious interferences from other systems falling at the GSM
receivers are required to be smaller than -123 (dBm/200kHz); otherwise, the
spurious interferences will seriously affect the GSM system.
Therefore, Iisolation = (-13dBm/100kHz) 56 - Ib + 10lg (200kHz/100kHz) = -13- 56(-123dBm/200kHz) + 10lg (200kHz/100kHz) = 57 dBm/200kHz.
That is, according to the assumption, the isolation between a CDMA antenna and
GSM 900MHz antenna must be at least 57dB regardless whether they share the
address or not.
Many ways can be used to reduce the interference. For example, you can adopt
the following ways:
Design enough distance between antennas
Filter the out-of-band interference of the transmitter
Add different equipments to the filter, such as receiver, duplexer, and divider.
According to the requirements in TIA/EIA-97 protocols, the spurious interference
from the CDMA antenna interface falling within the GSM 900MHz receiving bands
must be less than -13 dBm/100kHz. Therefore, the problems, such as mutual
interference and co-address construction must be considered in the initial design.
To be specific, you can filter and combine each transmitted channel number using
a limited-band filter with the bandwidth of only 1.23 MHz. The band-limited filter of
this type has great out-of-band attenuation, thus the space distance between the
antennas of the CDMA system and GSM system must be shortened.
In addition, to minimize the interference, you must keep suitable isolation
between the antennas of the CDMA system and GSM system.

2006-01-05

All rights reserved

Page 49 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The antenna isolation is calculated according to the following two formulas, which
has been introduced earlier:
For vertical arrangement, Lv = 28 + 40lg (k/ ) (dB)
For horizontal arrangement, Lv =22 + 20lg (d/ ) (G1+G2) (S1 + S2) (dB)
According to the two formulas, the requirements on the isolation between the
antennas of CDMA system and GSM 900 MHz system are specified in the
following three circumstances.
The antennas of the CDAM system and GSM 900MHz system do not share
the same address, with the antennas horizontally opposite to each other, or
the antennas of the two systems share the same address, with the antenna
type of omni antenna.
Suppose the effective gains of the antennas of the two systems in the
maximum radiation direction are 10 dBi (with the feeder loss considered),
and the interference signals are 890MHz, according to previous analysis, the
isolation between the CDMA system and GSM system is required at least
57dB.
Therefore, the following equation can be obtained according to the previous
formula:
57 = 22 + 20lg (Dh/ ) (10 + 10)
And the horizontal distance between the two antennas is d = 190m.
Table 5-18Table 5-18 lists the isolation requirements between omni
antennas of the two systems.
Table 5-18 Isolation requirements between omni antennas of CDMA system
and GSM 900MHz system
Effective antenna gain in

Antenna

isolation

Antenna

distance

radiation direction (dBi)

requirement (Db)

requirement (m)

10

57

190

10

57

599

The antennas of the CDMA and GSM 900 MHz system share the same
address (the antennas are installed on the same platform and horizontally
separated), with the antenna type of directional antenna.
Suppose that the two antennas are horizontally placed, and their tilt angle is
65, and that the effective gains of the two antennas in the radiation direction
are 15dBi.
And if the side lobe of the 65antenna is -18dB in the horizontal plane, the
effective gain of the antenna in this direction is (15 18) dBi = -3 dBi.
2006-01-05

All rights reserved

Page 50 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Therefore, 57=222+0lg (Dh/ ) - {(15+15) + [(-18) + (-18)]}.


According to the previous equation, the horizontal distance between the two
antennas are d = 9.5m.
Table 5-19Table 5-19 lists the isolation requirements between directional
antennas of CDMA and GSM 900 MHz systems.
Table 5-19 Isolation requirements between directional antennas of CDMA
and GSM 900 MHz systems
Effective antenna gain in

Antenna

isolation

Antenna

distance

radiation direction (dBi)

requirement (Db)

requirement (m)

10

57

190

15

57

599

The antennas of the CDMA and GSM 900 MHz antennas share the same
address (the antennas are not installed on the same platforms of the iron
tower and vertically separated), with the antenna types of directional
antenna and omni antenna.
In this case, the equation 57=28 + 40 lg (k/ ) is present.
According to this equation, the vertical distance between the two antennas is
d = 1.8m.
Note:
The previous descriptions are just theoretical detections. In actual networking,
other types of antennas may be installed at the same address. In this case, some
equipment indexes must be considered, among which the important ones are
spurious radiation, the interference power of the interfering signals to interfered
signals, and the antenna isolation.

IV. Installation distance between antennas


Diversity technology is the most anti-fading effective. When two signals are
irrelevant to each other, the horizontal distance between the diversity antennas
must be 0.11 times that of the valid antenna height. The higher place the antenna
is installed, the larger the horizontal distance between diversity antennas is.
When the distance between diversity antennas is equal to or greater than 6m,
however, the antenna is hard to be installed on an iron tower.
In addition, the distance required by vertical diversity antennas is 5 to 6 times that
of the horizontal diversity antennas when the same coverage is ensured.
Therefore, the vertical diversity antenna is seldom used in actual projects, but

2006-01-05

All rights reserved

Page 51 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

antennas are often vertically installed to meet isolation requirement, especially


omni antennas are vertically installed.
In addition, for highroad coverage, the line connecting two receiving antennas
must be perpendicular to the highroad. If space diversity is used, the diversity
distance is the perpendicular, as shown in Figure 5-8Figure 5-8.

Figure 5-8 Space diversity distance of directional antennas


Table 5-20Table 5-20 and Table 5-21Table 5-21 lists the required distance
between GSM antennas (suppose no barrier is present between the antennas)
Table 5-20 Required diversity distance between omni antennas
Isolation requirement: Tx-Tx, Tx - Rx: 30 dB
Vertical distance
(recommended)
GSM 900MHz:
Tx-Tx, Tx - Rx

Horizontal distance

Distance between
= 0.5m

Gain = 10 dBi: 8m

antenna and tower:


2m

GSM

Distance between

1800MHz:

= 0.25m

Gain = 10 dBi: 4m

Tx-Tx, Tx - Rx

GSM

1800MHZ:

antenna and tower:


2m

GSM 900MHz
+

Remarks

Distance between
= 0.5m

Gain = 10 dBi: 1m

antenna and tower:


2m

Tx-Tx, Tx - Rx
Diversity requirement
GSM 900MHz:
Tx-Tx, Tx - Rx

2006-01-05

= 4m (recommend
6m)

All rights reserved

Distance between
antenna and tower:
2m

Page 52 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

GSM
1800MHz:

Confideniality level

= 2m (recommend

3m)

Tx-Tx, Tx - Rx

Distance between
antenna and tower:
2m

Table 5-21 Required diversity distance between directional antennas


Isolation requirement: Tx-Tx, Tx - Rx: 30 dB
Same

sector

antennas

Vertical distance
(recommended)

GSM 900MHz:
Tx-Tx, Tx - Rx

Horizontal distance

For 65
= 0.5m

antenna gain: 15
dBi: 0.4m

GSM

For 65

1800MHz:

antennas,

Tx-Tx, Tx - Rx

No

effect

from

tower structure in
forward

direction

of the antenna
antennas,

antenna gain: 15

= 0.25m

Remarks

dBi: 0.2m

No

effect

from

tower structure in
forward

direction

of the antenna

Adjacent-sector
antennas
the

(on
same

Vertical distance

Horizontal distance

Remarks

= 0.5m

= 0.5m

platform)
GSM 900MHz:
Tx-Tx, Tx - Rx
GSM
1800MHz:
Tx-Tx, Tx - Rx
Diversity requirement
No
GSM 900MHz:
Tx-Tx, Tx - Rx

effect

from

= 4m (recommend

tower structure in

6m)

forward

direction

of the antenna
No

GSM
1800MHz:
Tx-Tx, Tx - Rx

2006-01-05

effect

from

= 2m (recommend

tower structure in

3m)

forward

direction

of the antenna

All rights reserved

Page 53 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Note:
The installation for GSM 900MHz and GSM 1800MHz antennas is flexible, but no
matter what specifications are used, they must meet the requirements on
isolation and distance. In addition, in actual projects, barriers are present
between antennas. For example, a tower is always present between two omni
antennas, so you must shorten the horizontal distance between them.

V. Design of base station parameters in residential areas


A large number of residential areas are distributed in urban areas, so this section
introduces the design of base station parameters in these areas.
(1)

Features of residential areas

Residential areas in urban areas are characterized by regular arrangement of


buildings, and they can be divided into the following types as listed in Table
5-22Table 5-22.
Table 5-22 Division of residential areas
In terms of

Description
High-building residential areas: above 10 floors (30 meters)

Building height

Multi-floor residential areas: 5 to 8 floors (15 -30 meters)


Villas and low residential areas: less than 4 floors (12 meters)
Great-intensity residential areas: the distance between
buildings is within 10 meters.

Building

Middle-intensity residential areas: the distance between

intensity

buildings ranges from 10 to 20 meters.


Low-intensity

residential

areas:

the

distance

between

buildings is larger than 20 meters.


The walls of the residential areas are constructed with
concretes.
The walls of the residential areas are constructed with bricks
and concretes.
The walls of the residential areas are constructed with hollow
Construction
material

blocks.
Notes:
The thickness of the buildings varies with the regions and
climates. Three specifications are available, namely, 24m,
47m, and 49m. Generally, the walls are thicker in southern
parts and thinner in northern parts.

2006-01-05

All rights reserved

Page 54 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

(2)

Confideniality level

Antenna installation in residential areas

The address where the antenna should be installed in residential areas is hard to
be determined. Generally, when adopting micro cells, you can install the antenna
within a residential area near to the target coverage area.
In this case, the antenna can be installed in the following spots:
On outer walls (not roofs) of a building
On pillars
Install a micro cell in underground garages
For the residential areas large in size and regular in arrangement, the antenna
can be installed as shown in Figure 5-9Figure 5-9.

Figure 5-9 Installed position of micro cell antennas


To realize better stereo coverage, you can install the antenna at wall corners with
enough height, namely, 1/2H building H antenna 3/4H building . In this case, the
azimuth angle of the antenna must be designed as shown in Figure 5-10Figure
5-10.

Figure 5-10 Installed position of micro cell antennas


If the antenna is installed at a wall corner, the major lobe of the antenna can
radiate the space between buildings. Generally, the major lobe of the antenna
cannot face the walls of the buildings nearby directly.
If frequencies are reusable among these micro cells, the directions of antennas
must be consistent with each other. In addition, you can also use the cell splitter
to enable a cell to coverage the areas in two directions, as shown in Figure
5-11Figure 5-11. In this case, however, the frequency utilization ratio may
decrease and extra power splitter will introduce loss of 3 dB.

2006-01-05

All rights reserved

Page 55 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

For the residential areas with regular arrangement, the directional antennas
whose horizontal beam width is 90 to 120 and vertical beam width is greater
than 30 are recommended.
Under certain conditions, the micro cell antenna can be installed on the pillars
within a residential area, as shown in Figure 5-11Figure 5-11.

Figure 5-11 Installed position of micro cell antennas


For the residential areas with irregular arrangement, the antenna can be installed
on the walls of a building, so the reflected waves can coverage the walls of
opposite buildings, as shown in Figure 5-12Figure 5-12. In this case, the
antennas whose horizontal beam width is greater than 120and vertical beam
width is greater than 30are recommended.

Figure 5-12 Installed position of micro cell antennas


(3)

Antenna selection

When the walls of a building is selected as an installed position, you can use the
build-in antenna of the micro cell directly, or other antennas with small size.
According to coverage features of residential areas, when selecting the
specifications for the micro cell antennas to be used, you must consider the
following factors:
Antenna gain
2006-01-05

All rights reserved

Page 56 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Horizontal beam width


Vertical beam width
Polarization mode
Visual effect (antenna size, shape, and weight)
The antenna gain is recommended less than 9 dBi for micro cell antennas.
Because the coverage area of a micro cell antenna is small and the installed
position is near to the coverage area, the antenna gain can be adjusted to a
smaller value, especially if the gain of an antenna is greater than 10dBi, its size is
large, which may cause opposition from residents.
The selection of the horizontal and vertical beam width for an antenna is related
to radio environment. If a micro cell antenna is installed on a wall, the antenna
height is lower than the average height of surrounded buildings. In this case, if
both the indoor coverage of lower floors and higher floors can be assured, you
must select the antennas with greater vertical beam width. According to the
height of buildings, you can select the directional antennas whose vertical beam
width ranges from 35to 80.
The selection of the horizontal beam width of the micro cell antenna and the
installed position of the antenna are related to coverage target. In this case, you
can select the directional antennas whose beam width ranges from 60 to 150,
or you can choose omni antennas or bi-directional antennas (8-shaped
antennas).
Both vertical polarization antennas and dual polarization antennas can be
selected for a micro cell. The coverage area of a micro cell in urban areas is small,
so the diversity reception is unnecessary. In this case, a vertical polarization
antenna can meet the coverage requirements in residential areas. As for the dual
polarization antenna, however, it is expensive and large in size, so it is not
recommended.
The visual effect must be emphasized for the micro cell antennas installed in
residential areas. They must be small and moderate. In addition, they must be
light for installation convenience. If the contract between the color of the antenna
and that of the surrounded buildings is great, you must color the antenna with the
same color of the buildings.

Note:
In some cases, you should consider adopting dual-band antennas. When
selecting a small-sized antenna, you should consider whether its maximum
output power can bear the micro cell output power. When adopting short jumpers
instead of 7/8 feeders, you should consider whether the antenna connector
(N-shaped male/female, 7/16 DIN header) matches the jumper connector.

2006-01-05

All rights reserved

Page 57 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

5.8 Location Area Design


5.8.1 Definition of Location Area
In GSM protocols, a mobile communication network is divided into multiple
service areas according to the codes of location areas. Thus the network pages a
mobile subscriber through paging its location area.
Location area is the basic unit of paging areas in a GSM system. That is, the
paging message of a subscriber is sent in all cells of a location area. A location
area contains one or more BSCs, but it belongs to one MSC only.
Figure 5-13Figure 5-13 shows the division of service areas.

Figure 5-13 Division of service areas

5.8.2 Division of location areas


The coverage area of each GSM PLMN is divided into multiple location areas, in
which an MS is positioned. The size of a location area, namely, the area covered
by a location area code (LAC), plays a key role in a GSM system. Therefore, this
section mainly introduces the principle for planning location areas.

I. Dividing the location area according to the distribution and behaviour of


mobile subscribers
The distribution of location areas in cities and suburbs is different. Generally,
suburban areas or counties occupy independent location areas. In cities, the
distribution of location areas is similar to a concentric circle. (The areas in the
internal circle can be divided into several location areas due to the requirements
on capacity. The concentric circle can be divided into several fragments.)

2006-01-05

All rights reserved

Page 58 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Practice has proved that if the location areas are divided according to the
previous methods, as shown in Figure 5-14Figure 5-14, both coverage and
call-connected ration can be improved.

Figure 5-14 Division of LAs


In addition, if two or more location areas are present simultaneously in a big city
of great traffic, the landforms, such as mountains and rivers within this city can be
used as edges of the location areas. In this case, the overlapped depth between
the cells of the two location areas can be reduced. If no such landforms available
within this city, the areas (such as streets and shopping centers) with great traffic
cannot be used as edges of the location areas.
Generally, the edge of a location area is oblique instead of parallel or
perpendicular to streets. In the intersected areas of urban areas and suburban
areas, to avoid frequent location update, you must design the edges of location
areas near the outer base stations instead of the base stations just installed at the
intersections.

II. Calculating coverage area and capacity of a location area


If the coverage area of a location area is too small, the mobile station will perform
frequent location update. In this case, the signaling flow in the system will
increase. If the coverage of a location area is too larger, however, the network will
send a paging message in multiple cells until the mobile station is paged. In this
case, the PCH will be overloaded and the signaling flow at the Abis interface will
increase.
The calculation of location areas varies with the paging strategies designed by
different carriers. During early network construction stage, the traffic is not great,
so a location area can accommodate more TRXs. However, it is still necessary for
you to monitor the PCH load and traffic growth. When the traffic grows great, you
can enhance the PCH capacity by adding a BCCH to the system, but the number
of voice channels can be added is reduced by one accordingly.
Generally, the capacity of a location area is calculated as follows:
2006-01-05

All rights reserved

Page 59 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The number of paging blocks sent in each second the number of paging
messages sent in each paging block = the maximum paging times in each second.
As a result, the number of paging times in each hour, the traffic allowed in each
location area, and the number of carriers supported in each location area can be
deducted.
The followings introduce the items present in the previous paragraph respectively.
(1)

The number of paging blocks sent in each second

1 frame = 4.61ms, 1 multiframe = 51 frames = 0.2354s; suppose the number of


access grant blocks is AGB, the number of blocks, the number of paging blocks
sent in each second is calculated by the following formulas:
For non-combined BCCH, the number of paging blocks sent in each second
= (9 AGB)/0.2345 (paging block/second).
For combined BCCH, the number of paging blocks sent in each second = (3
AGB)/0.2345 (paging block/second).
For non-combined BCCH, the AGB is 2 according to Huawei BSC. Therefore, the
number of paging blocks sent in each second is 29.7 (paging block/second);
when AGB is 0, it is 38.2 (paging block/second).
For combined-BCCH, the AGB is 1, so the number of paging blocks sent in each
second is 8.5 (paging blocks/second); when the AGB is 0, it is 12.7 (paging
block/second).
According to the previous analysis, the larger the number of AGB, the smaller the
number of the paging blocks sent in each second and the smaller the paging
capacity is. Moreover, the paging capacity of the combined BCCH is far less than
that of the non-combined BCCH.

Note:
Generally, a combined-BCCH cell and a non-combined-BCCH cell are not
configured simultaneously within a LAC, and the number of AGB must be
consistent with a location area; otherwise the paging capacity of the location area
will decrease (now the paging capacity of the cell with the least paging capacity is
the paging capacity of the location area).
However, if the capacity of a location area is small and the LAC resource is
scarce, you can configure the combined-BCCH cell and non-combined-BCCH
cell within a LAC to enlarge the number of traffic channels for O1 and S111 base
stations.

(2)

The number of paging messages sent in each paging block (X)

2006-01-05

All rights reserved

Page 60 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

According to section 9.1.22 of GSM0408 protocols, each paging block has 23


bytes, and can send 2 IMSI pages, or 2 TMSI and 1 IMSI pages, or 4 TMSI
pages.
According to the paging strategies of Huawei MSC, if the IMSI paging mechanism
is adopted, the number of paging messages sent in each paging blocks is 2
(paging times/paging block); if the TMSI paging mechanism is adopted, it is 4
(paging times/paging block)
(3)

The maximum paging times in each second (P)

The maximum paging times in each second is calculated by the following two
formulas:
For non-combined BCCH, P = (9 AGB)/0.2345 (paging block/second)
(paging times/paging block).
For combined BCCH, P = (3 AGB)/0.2345 (paging block/second) (paging
times/paging blocks).
If the IMSI paging mechanism is adopted, for non-combined BCCH, when AGB =
2, P = 59.47 (paging times/second); when AGB = 0, P = 76.47 (paging
times/second). For combined-BCCH, when AGB = 1, P = 16.99 (paging
times/second); when AGB = 0, P = 25.49 (paging times/second).
If the TMSI paging mechanism is adopted, for combined BCCH, when AGB = 2, P
= 118.95 (paging times/second); when AGB = 0, P = 152.93 (paging
times/second). For combined BCCH, when AGB = 1, P = 33.98 (paging
times/second); when AGB = 0, P = 50.98 (paging times/second).
According to the previous analysis, the paging capacity under IMSI paging
mechanism is half of that under TMSI paging mechanism.
(4)

The traffic allowed in each location area (T)

When designing the capacity for a location area, you must be attention that the
paging capacity of a location area cannot break its limit. For network expansion,
you can collect the times of the busy-hour paging orders delivered by BSC from
OMC, and then convert the times into the number of paging orders sent in each
second.
If no traffic measurement data is available, such as in the case of new network
construction, you can calculate the traffic allowed in each location area by
assuming a traffic model.
For example, if the average conversation duration is 60s and the ratio of the times
for the mobile station to be successfully paged to the times of total pages is 30%,
the 60s of conversation duration matches 1/60 calls (in the unit of second. Erl),
and 30% of calls is generated by the called parties. Therefore, the successful
calls of the 30% mobile stations are 0.05 times (that is, 1/60*30% = 0.005), in the
unit of second. Erl.

2006-01-05

All rights reserved

Page 61 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

If the 75% of the mobile stations respond to the first page and 25% respond to the
second page, the mobile stations responding to the third page can be neglected.
(It is just an assumption, which may be different from actual conditions.).
Therefore, 1.25 pages are needed if a mobile station is successfully called each
time (25% of the pages must be resent). In this case, the following equation is
present:
Y = 0.005*(1+25%) = 0.00625 paging times/(second. Erl)
Suppose the congestion on paging channels will occur when the paging capacity
is 50% greater than maximum theoretical paging capacity, the original paging
messages are still present even the paging queue is full in the BTS. In this case,
the paging capacity in one second is P*50%.
Therefore, the traffic allowed in each location area can be calculated according to
the formula T = P*50%/Y, and the specific values are listed in Table 5-23Table
5-23.
Table 5-23 Traffic allowed in each location area
Paging
mechanism

IMSI

TMSI

(5)

BCCH

T (Erl)

AGB (block)

4757.86

6117.25

1359.39

Combined BCCH

2039.08

Combined BCCH

9515.72

12234 .49

2718.78

Combined BCCH

4078.16

Combined BCCH

combination mode
Non-combined
BCCH
Non-combined
BCCH

Non-combined
BCCH
Non-combined
BCCH

The number of carriers supported by each location area (NTRX)

Each TRX had 7.2 TCHs in average, so the maximum traffic of each TRX in each
hour is 7.2.
Therefore, the number of carriers supported in each location area can be
calculated according to NTRX = T/7.2 and the specific values are listed in
2006-01-05

All rights reserved

Page 62 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Table 5-24 Number of carriers supported in each location area (TRX/LA)


Paging
mechanism

IMSI

TMSI

BCCH

NTRX (TRX/LA)

AGB (block)

660

849

188

Combined BCCH

283

Combined BCCH

1321

1699

377

Combined BCCH

566

Combined BCCH

combination mode
Non-combined
BCCH
Non-combined
BCCH

Non-combined
BCCH
Non-combined
BCCH

Note:
All the previous assumptions do not include the effect of the point-to-point short
messages against on paging capacity. If the conversation times of a subscriber
are equal to the number of the short messages to be sent, and if the sent ratio
and received ratio are consistent with each other, the paging times/second. Erl
will double in busy hour and the capacity of the location area will reduce by half.
Therefore, some common short messages must be sent on CBCH.

5.8.3 Others
This section introduces some other information about location area design.
The capacity of a location area is closely related to paging mechanism, and
is directly related to the combinations of AGB and BCCH. When the
combinations of AGB and BCCH are inconsistent with each other in a
location area, the capacity of the location area is determined by the cell with
the smallest capacity. Therefore, the combinations of AGB and BCCH must
be designed to be consistent in location area planning.

2006-01-05

All rights reserved

Page 63 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

If the number of point-to-point messages grows large immediately, the


number of paging messages will increase, but the number of supported
subscribers will decrease. In this case, you must control and protect the
flows in the system.
Because the traffic density varies with location areas, it is recommended that
the combined-BCCH cells, non-combined-BCCH cells, and multi-BCCH cells
form a location area respectively. When a cell with BCCH/SDCCH
combination, the location area can be as large as possible when the paging
capacity of the BTS does not reach the limit. However, because all paging
messages will be broadcasted in all cells within a location area, the cell with
BCCH/SDCCH combination is the bottleneck of the location area.
The LAC is a kind of number resource. Therefore, you must cooperate with
carries to plan location areas.

5.9 Dual-Band Network Design


5.9.1 Necessity for Constructing Dual-Band Network
The earlier GSM mobile communication network is constructed on the 900 MHz
band. With rapid growth of subscribers, the network capacity also grows rapidly.
Therefore, the lack of frequency resources and radio channels is a major concern
for mobile telecommunications.
Many methods can be used to expand the capacity of a GSM system, including:
Adding macro cell base stations to the system
Reducing distance between base stations
Adopting aggressive frequency reuse technologies (such as MRP and 13)
Adding micro cells to the system
Applying half rate to the system
However, all these methods cannot thoroughly solve the problems concerning
network capacity. As a result, the GSM 1800MHz network is introduced (uplink:
18051880 MHz; downlink: 17101785 MHz). And the network integrating GSM
900MHz and GSM 1800MHz can meet the growth of network capacity.
The application of GSM 1800MHz can bring the following advantages:
It does not occupy the bands of GSM 900MHz and has a communication
bandwidth of 75M. Therefore, it breaks the bottleneck of GSM 900MHz in
terms of frequency resources.
The system networking, project implementation, network planning, and
network maintenance of a GSM 1800MHz network are almost the same with
that of a GSM 900MHz network.
The GSM 1800MHz and GSM 900 MHz can share a base station, so a GSM
1800MHz network can be finished in a short time, which is quite helpful for
network expansion.
2006-01-05

All rights reserved

Page 64 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Dual-band mobile phones now accounts for a major part of the total, so a
GSM 1800MHz network can provide services to the dual-band subscribers.
In this case, the capacity pressure on GSM 900MHz can be greatly eased.

5.9.2 GSM 1800MHz Coverage Solutions


I. Propagaiton features of GSM 1800MHz
The propagation features of the electromagnetic waves of 900 MHz and 1800
MHz are different in the following aspects:
The propagation loss in free space
The propagation loss of the 1800 MHz signals is 6 dB greater than that of the
900 MHz signals in free space.
Penetration loss
The penetration loss of the 900 MHz signals is greater than that of the 1800
MHz signals, but their difference is slight.
Diffraction loss
The longer the waves, the smaller the diffraction loss is. The diffraction ability
of the 1800 MHz signals is poorer than that of the 900 MHz signals.

II. Dual-Band Networking Mode


There are three dual-band networking modes, namely, independent MSC
networking, co-MSC/independent BSC networking, and co-BSC networking,
among which the former two are called independent networking, and the later is
called hybrid networking.

III. Coverage requirements on GSM 1800 MHz


Outdoor coverage
The outdoor coverage can be easily realized when the distance between
base stations are not large. In necessary cases, you can add a GSM
1800MHz base station at the address of the original GSM base station. And
in some places, you should consider add a new base station.
Indoor coverage
To ensure that the indoor coverage of GSM 1800MHz is good, you must
control the distance between the base stations installed in urban areas within
1000 meters. In China, however, the buildings in most cities are constructed
by concretes and metals, so the penetration loss is great. As result, the
distance between base stations in urban areas of China ranges from 500 to
800 meters.

IV. Coverage mode of GSM 1800MHz


(1)

Scattered coverage in hotspot areas

2006-01-05

All rights reserved

Page 65 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

At the early network construction stage, the GSM 1800MHz base stations are
scattered in hotspot areas. When the capacity configured for a GSM 1800 MHz
base station is small, you must solve the problems, such as SDCCH congestion,
TCH congestion, and frequent update between GSM 1800MHz and GSM
900MHz. The cost in early construction stage is small.
Figure 5-15Figure 5-15 shows the scattered coverage of GSM 1800MHz in
hotspot areas.

Figure 5-15 Scattered coverage of GSM 1800MHz in hotspot areas


The coverage of the dual-band network of this mode is based on the original GSM
900MHz network. The GSM 1800MHz base station is constructed in some
hotspot areas, so the seamless coverage of GSM 1800MHz is not available in
this case.
If a dual-band mobile phone starts conversation in an area covered by GSM
1800MHz, after leaving this coverage area, it hands over to the GSM 900MHz cell
where it originally was. And the handover of this type is called the inter-band
handover caused by coverage.
If a dual-band mobile phone starts the conversation in an area covered by GSM
900MHz, but because the traffic in this area is great, the mobile phone will hand
over to an area covered by GSM 1800MHz. And the handover of this type is
called the inter-band handover caused by capacity.
The scattered coverage in hotspot areas only relieves capacity problems in a
short term. Moreover, frequent inter-band frequency handover increase the
signaling load, which results in the loss of system capacity.
(2)

Seamless coverage in hotspot areas

If the coverage of this mode is available; the GSM 1800 MHz network can share
greater traffic for GSM 900MHz network and expand the system capacity. In
addition, it is cost-effective.
(3)

Perfect seamless coverage

2006-01-05

All rights reserved

Page 66 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

If a GSM 1800MHz network adopts the coverage of this type, the advantages are
as follows:
The seamless coverage area within a city can be realized.
The GSM 1800MHz network can share the traffic load for GSM 900MHz
network as much as possible.
The system capacity can be greatly expanded.
The ratio of the handover between layers is small.
The quality of the network is quite satisfying.
The frequencies can be planned by patch.
The carriers can be expanded step by step.
However, there are still disadvantages. They are as follows:
The number of base stations is large.
The work load of network planning and optimization is huge.
The investment is large.
The base station addresses cannot be decided once.
Figure 5-16Figure 5-16 shows the perfect seamless coverage of GSM 1800MHz
in hotspot areas.

Figure 5-16 Perfect coverage of GSM 1800MHz in hotspot areas


If a GSM 1800MHz network adopts this coverage mode, it can be easily
expanded to meet future coverage.
Compared with the scattered coverage in hotspot areas, the perfect seamless
coverage is characterized by great intensity and large area. Therefore, the ratio of
inter-band handover under this coverage mode is far smaller than that under
scattered coverage mode. As a result, the signaling load is reduced greatly.
Therefore, this coverage mode is an ideal coverage solution. If a GSM 1800MHz
network adopts this coverage mode, it does not necessarily attach to the GSM
900MHz network, instead, it can form an independent network.

2006-01-05

All rights reserved

Page 67 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

5.9.3 Location Area Division for Dual-Band Network


The location area division for dual-band network is suggested as follows:
If 1800 MHz cells and 900 MHz cells are under the control of two MSCs
respectively, their location areas are different. Therefore, you must set
related parameters to maintain the mobile stations stay in the 1800 MHz
cells where the traffic is absorbed. In this case, the times for the mobile
station to handover between the two bands and reselect cells will decrease.
Meanwhile, when designing signaling channels, you must fully consider the
load resulted from location update.
If 1800 MHz cells and 900 MHz cells share a MSC, at the early network
construction stage, they are suggested to use the same location area without
affecting the network capacity. If the restriction on paging capacity is present,
two location areas must be divided for them either in terms of band or
geographic location, as shown in Figure 5-17Figure 5-17 and Figure
5-18Figure 5-18.

Figure 5-17 Location area division based on band

Figure 5-18 Location area division based on geographic location


If the location area is divided in terms of band, because frequent location updates
are resulted from inter-band handover and cell reselection, you must set related
parameters to maintain the mobile stations stay in the 1800 MHz cells where the
2006-01-05

All rights reserved

Page 68 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

traffic is absorbed. In this case, the times for the mobile station to handover
between the two bands and reselect cells will decrease. Meanwhile, when
designing signaling channels, you must fully consider the load resulted from
location update.
If the location is divided in terms of geographic location, the frequent location
updates resulted from inter-band handover and cell reselection can be avoided.
However, you need to modify the related data of the original 900 MHz network. In
addition, at the edges of the location areas, because the location updates caused
by intra-band and inter-band handover and cell reselection is present
simultaneously, the signaling flow is huge at these edges. As a result, you must
carefully design the edges of the location areas.

5.9.4 Traffic Guidance and Control Strategies of Dual-Band Network


I. Traffic guide of Dual-Band Network
At early construction stage of a dual-band network, traffic control concerns how to
use the new GSM 1800MHz network to share the traffic flow for the GSM
900MHz network. According to the original intension of the GSM 1800MHz
network, the traffic can be guided according to the following principles:
1)

At the early construction stage of a dual-band network, the GSM 1800MHz


network is mainly applied to absorb the traffic of the dual-band subscribers
so that the load of the GSM 900MHz network can be eased.

2)

When the number of dual-band subscriber grows large, each band must
share the traffic so that the inter-band handover times can be reduced.

Figure 5-19Figure 5-19 shows the process of traffic guide and control strategies.

Cell selection

Idle mode

Directed retry

Call setup

Cell
reselection

Inter-layer and interlevel handover

Call
status

Traffic load handover

Co-BSC/MSC
handover control

Figure 5-19 Process of traffic guide and control strategies


The various traffic control strategies can be realized through adjusting parameter
settings as follows:
1)

In idle mode, when the mobile station is selecting cells after it is switched on
and reselecting cells when it is in standby state, you can set higher priorities

2006-01-05

All rights reserved

Page 69 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

for the 1800 MHz cells by designing the system parameters, including CBQ,
CBA, CRO, TO, and PT. In this case, subscribers are more likely to stay in
the 1800 MHz cells. As a result, their calls are established on the 1800 MHz
cells.
2)

If traffic congestion is present in the service cell when a mobile station is


setting up a call, the system applies directed retry function to assign the
mobile station to a TCH in the neighbor cells of the service cell and adjust
the traffic allocation.

3)

In conversation state, the traffic must be guided to the 1800 MHz cells in
lower layers and levels according to the hierarchy cell structure. In addition,
you can use Huawei dual-band handover algorithms so that the traffic load
can be allocated more properly.

II. Hierarchical Cell Structure


According to the hierarchy cell structure of the dual-band network, a GSM system
covering an area can be divided into four layers, as listed in Table 5-25Table 5-25.
Table 5-25 Four layers of a GSM system
Layer

Description
This layer consists of umbrella cells, namely, GSM 900MHz macro

cells. They provide upper coverage and connect the fasting moving
mobile stations.

This layer consists of GSM 900 micro cells, and most of the
subscribers are gathering in this layer.
This layer consists of the GSM 1800MHz cells whose coverage area is

similar to that of the GSM 900 MHz cell. It is designed in case of


frequency resource emergency. In the future, most of the dual-band
subscribers will gather in this layer.
This layer consists of the mini cells of GSM 1800 MHz and GSM

900MHz. They are designed for covering hotspot areas and dead
zones.

Figure 5-20Figure 5-20 shows the hierarchy cell structure of a dual-band network.

2006-01-05

All rights reserved

Page 70 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-20 Hierarchical cell structure of a dual-band network


To enable the network to develop smoothly and flexibly, you can divide each of
the four layers into multiple levels, and then you can set multiple priority classes
(for example, 16 classes) for the levels in each layer. This method is not only
helpful for adjusting the traffic load in part of the areas. Therefore, the hierarchical
cell structure enhances the cooperation of the current network equipments and
meets the devolvement of the future network.
In terms of traffic priority, the cells in lower layers and levels has higher priorities,
namely, the cells in lower layers has the priority to absorb the traffic.
The handover algorithms of a dual-band network are listed in Table 5-26Table
5-26.
Table 5-26 Handover algorithms of a dual-band network
Name

Description

Inter-layer and inter-level handover

The network support to control the cells

algorithm

according to their priority classes.

Emergency handover algorithm

Load sharing handover algorithm

It manages the emergent cases, such as


TA, BQ, and fast level fall.
It enables cells to share load with each
other.

Handover algorithms based on

It helps fast-moving mobile phones to

speed and sensibility

perform handover flexibly.

Interference handover algorithm

Edge handover algorithm

2006-01-05

It helps reduce interference and enhance


voice quality.
It provides mobile phones with better
levels and service qualities.

All rights reserved

Page 71 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

PBGT handover algorithm based

It supports mobile phones to hand over to

on path loss

a better cell.

Co-MSC/BSC

handover

control

algorithms

5.9.5 Dual-Band Networking Engineering Implementation


During network construction and optimization, a dual-band network is debugged
and commissioned step by step, which facilitates debugging the new GSM 1800
MHz networks and the original GSM 900MHz networks that has been expanded
respectively. After each signal network is perfectly adjusted, you must debug
each base station in the dual-band network. And you cannot stop the debugging
until the whole dual-band network is finished.
The construction of a whole dual-band network can be divided into three stages,
namely, deployment preparation, signal 1800 MHz network debugging, and
900/1800 MHz dual-band network debugging.

I. Deployment perparation
The coordination of dual-band technologies and network planning must be
finished in this stage. The coordination of dual-band network technologies is a
prerequisite for the cooperation of different carriers networks. Network planning
is the first step in network construction and involves many tasks, including base
station address survey, channel number planning, electromagnetic background
test, coverage test, and so on.
The followings must be emphasized in dual-band cooperation:
The customers, the third party (the designing institute or the original
equipment supplier), and the new equipment supplier must be cooperate
with each other well.
If one party meets a tough problem during the debugging of the dual-band
network, the engineers from a third party must be present in site and help
position the problem.
The 900 MHz BSC and 1800 MHz BSC must synchronize their clocks with
the same source clock. Meanwhile, the clock of each base station in the
existing GSM 900 MHz network can lock the clock of the BSC, and the clock
of the BSC can lock the clock of the MSC.
When modifying the parameters related to dual-band handover (such as
modifying the parameters at the BSC side or MSC side), you must notify that
to other two parties.

2006-01-05

All rights reserved

Page 72 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

If the some problems concerning the cooperation of dual-band network arise,


a meeting must be organized, in which each party discuss with each other on
how to solve the problems.
Both the designing institute or the original equipment supplier and the new
equipment supplier must provide the project implementation plan, cutover
plan, and precise cell information.

II. Signal 1800 MHz network debugging


At this stage, you need not modify any data of the original GSM 900 MHz network,
but it is still the GSM 900MHz network provides services to subscribers. The
GSM 1800MHz network does not absorb traffic.
When debugging the GSM 1800MHz network, you must adjust the following
parameter so that the existing subscribers can be least affected.
In the system message data list, set the parameter CBA to NO to prevent
general subscribers from selecting and reselecting the 1800 MHz network.
Theoretically, general subscribers can hand over to the 1800 MHz network, but in
fact, the handover relationship is not configured with the dual-band network, so
the general subscribers cannot enter the 1800 MHz network.
After that, you use the testing mobile phone which can access the network by
force to perform dialing test in each cell. If all goes normal, you can test coverage,
handover, power control, interference, downlink and uplink balance, power
adjustment, the coverage of the GSM 900MHz network, and the coverage of the
GSM 1800MHz network.
Through these tests, you can not only discover the problems present in the
networks, but also adjust the channel number, power, tilt angle, and parameter
setting and optimize the parameter configuration for the GSM 1800MHz cell. In
this case, the coverage and operation of the single GSM 1800MHz network can
be ensured.

III. 900/1800 MHz dual-band network debugging


After finishing the single GSM 1800MHz network debugging, you must change
back the parameter CBA to YES and configure the data for dual-band
handover. The tests involved into the dual-band network debugging include:
Cell reselection and location update
Traffic load control
Continuous conversation mode
Automatic dialing and scan
Dual-band network handover
Calls and handovers initiated on major streets
Calls and handovers initiated on edge areas
Dialing tests in poor coverage areas and indoor environment
2006-01-05

All rights reserved

Page 73 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Dialing tests in outdoor and indoor environments in key areas


Table 5-27Table 5-27 lists the test items during cell selection and handover
Table 5-27 Test items during cell reselection and handover
Test state

Test time

When a mobile phone

When the mobile phone performs cell selection from

performs cell selection in

a 900MHz cell to an 1800MHz cell.

idle mode (CRO, C1, and

When the mobile phone performs cell selection from

C2.)

an 1800MHz cell to a 900MHz cell.


The mobile phone first establishes the conversation
in a 900MHz cell, and then hands over to an
1800MHz cell.
The mobile phone first establishes the conversation
in an 1800MHz cell, and then hands over to a
900MHz cell.

When a mobile phone


performs

handover

conversation state

in

The mobile phone first establishes the conversation


in a 900MHz cell, and then hands over to an
1800MHz cell, and finally hands over back to a 900
MHz cell controlled by the same MSC.
The mobile phone first establishes the conversation
in an 1800MHz cell, and then hands over to a
900MHz cell, and finally hands over back to an 1800
MHz cell controlled by the same MSC.

At this stage, you must configure the dual-band data for the GSM 900MHz and
1800MHz cells. The data includes neighbor cell relationship, layer and level
setting, handover type, and handover threshold. In this case, when a mobile
phone is in idle mode, it can reselect an 1800MHz cell, the GSM 1800MHz
network can absorb the traffic of dual-band subscribers, and the subscribers can
perform handover between 1800MHz cells and 900MHz cells.
At the beginning, you can control the GSM 1800MHz network to absorb only a
small part of the traffic of subscribers through adjusting the setting of CRO and
handover threshold. When good cell reselection and dual-band handover are
ensured, you can take measures to enable the GSM 1800MHz network to absorb
more traffic, with the prerequisites that no congestion is present among cells and
the network quality is ensured.
At this stage, the following parameters must be configured:

2006-01-05

All rights reserved

Page 74 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The parameters related to cell selection and reselection, including CBA,


CBQ, ACCMIN, CRH, and CRO.
The parameters related to neighbor cell relationship, layer and level setting,
and handover.
The configuration of the previous parameters must be based on the prerequisite
that the cooperation of the GSM 1800MHz cells and GSM 900MHz cells is
normal.
After the GSM 900MHz and 1800MHz dual-band network is enabled, you must do
the followings:
1)

Find out the problems present in the network through multiple means, such
as drive test.

2)

Adjust and optimize the network according to the problems so that the
dual-band network can run stably.

3)

Check if the dual-band network runs stably, analyze all the traffic statistic
data, and check the network operation indexes.

4)

Make sure the problems and take effective measures according to the
analysis of the drive test and traffic statistics.

5)

Adjust the related parameters and retest the network till the network indexes
meet the design requirements.

Thus, a dual-band network is constructed and optimized according to the three


stages as introduced in this section.

5.10 Design of Indoor Coverage System


5.10.1 Characteristics of Indoor coverage
With the rapid development of economy, hotels, commercial centers, large-scale
flats, underground railways, and underground parking areas are arising by batch.
As a result, mobile stations are more frequently used in indoor environment. Thus,
they require better indoor mobile communication services.
Generally, the following problems are present in indoor mobile communication
systems:
From the perspective of coverage, the complex indoor structure and the
shielding and absorbing effect of the buildings cause great radio wave
transmission loss. As a result, the signals in some areas may be weak,
especially the signals in the first and second floors in the underground are
quite weak, or even there are dead zones. In this case, mobile stations
cannot necessarily access the network, there is no paging response, or
subscribers are not in service areas.
From the perspective of network quality, the factors interfering radio
frequencies are probably present in upper floors of high buildings. In this

2006-01-05

All rights reserved

Page 75 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

case, the signals in service areas are not stable, so ping pong effect may
occur and conversation quality cannot be ensured.
From the perspective of network capacity, if mobile stations are frequently
used in buildings, such as large-scale shopping centers, conference halls,
some areas in the network cannot meet the requirements of subscribers. In
this case, congestion may occur on radio channels.
If the indoor coverage is realized by a repeater, an outdoor high-power base
station, or a great-height outdoor antenna, the following problems may arise:
The penetration loss is great, so the indoor coverage is not satisfying. In this
case, a large number of dead zones are present, so subscribers cannot keep
conversation.
If a repeater is adopted, the level of original signals must be high. In addition,
the cross-modulation and intra-frequency interference is great, so the
conversation quality is weak and call drop ratio is high.
The network capacity is limited and the call connected ratio is low.
The frequency planning is hard to be performed for the network and the
network capacity is hard to be expanded.
The detached island effect is great.
The value-added services are restricted for group subscribers due to
network quality and capacity.
To enhance the grade of service, we must improve indoor coverage immediately.
When designing an indoor coverage system, we must make the following
considerations:
A new indoor coverage system cannot affect the existing network.
Enough capacity of an indoor system must be ensured.
An indoor system must support new services and functions.
The chapter analyzes the design of indoor coverage system from the following
aspects:
Indoor Antenna System Design
Capacity Analysis and Design
Frequency Planning
Traffic Control

5.10.2 Indoor Antenna System Design


I. RF design
(1)

Link budget

In an indoor coverage system, the link budget formula is as follows:

Pant

MS sens

RFm arg IFm arg BL LNFm arg L path G ant

Here,
2006-01-05

All rights reserved

Page 76 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Pant = antenna input interface power


RFmarg = Raleigh fading margin
IFmarg = access margin (depends on environment)
LNFmarg = design margin (generally, it is 5 dB)
BL = body loss (900MHz: 5 dB; 1800/1900MHz: 3 dB)
MSsens = mobile station sensitivity
Lpath = path loss
Here, Lpath = 20logd (m) + 30logf (MHz) - 28 dB + a. When there no barrier loss, Lp
= 20logd (m) + 30logf (MHz) - 28 dB. The a indicates the loss caused by other
bariers.
Table 5-28Table 5-28 lists the penetration loss caused by some typical barriers.
Table 5-28 Penetration loss caused by some typical barriers
Barrier

Penetration loss (dB)

Partition wall

520 dB

Floor

> 20 dB

Furniture

215 dB

Thick glass

610 dB

Train carriage

1530 dB

Elevator

Around 30 dB

Tunnel curve

1040 dB/km (for the signals from fixed sigal sources)

Rectangle tunnel

1015 dB/km

Cylindrical tunnel

3540 dB/km

Note:
Because the penetration in cylindrical tunnels is great, leaky cables are applied
in cylindrical tunnels.

When performing link budget, you must consider the followings:


In an indoor multi-antenna system, the link budget for test points must be in
accordance with the link with the minimum loss.
Under the same converge area, the EIRP at each antenna interface must be
consistent, and the error must be controlled within 10 dB.
The uplink signal must be designed to a high value, so antenna diversity is
unnecessary.

2006-01-05

All rights reserved

Page 77 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

To reduce uplink interference, you must properly set the maximum transit
power of the mobile station and enable the power control function of the
mobile station.
A certain margin must be leaved for error correction and future system
expansion.
The estimation and design for interference margin vary with the distance
from the outer wall. The smaller the distance, the larger the interference
margin is designed.
(2)

Service quality design (interference degree)

Table 5-29Table 5-29 describes the interference degrees of indoor cells in


different situations.
Table 5-29 Interference degree of indoor cells
Situations

Interference degree

The indoor cell and the surrounded buildings are


of the same height, and frequency reuse degree
is greater than 12.
The outdoor coverage system has little effect on

Little interference

the indoor cells.


The indoor system has dedicated frequencies
and frequency reuse seldom exists.
The situation and frequency reuse are intervenient
between the two.

General interference

The buildings within the indoor cell are higher


than surrounded buildings.

Great interference

The frequency reuse degree is smaller than 9.


Note:
The actual interference level changes with network layout and frequency
re-planning, and it can be tested according to actual situations.

(3)

Service quality design (interference margin design)

The greater the interference in an area, the greater the interference margin (IFmarg)
is designed, and the higher the level the mobile station needs to receive, as listed
in Table 5-30Table 5-30.
Table 5-30 Relationship between interference and mobile station receiving level
Actual level interference degree

2006-01-05

Mobile station receiving level (dBm)

All rights reserved

Page 78 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Great

-65

General

-75

Little

-85

Note:
When a dual-band system is adopted in the indoor environment, the indexes of
mobile station receiving level are designed according to the 1800 MHz system
standard.

II. Antenna system design


When designing an indoor distribution system, you must first survey the building
type, structure, interference environment, customers, and then analyze the path
loss. Finally, decide the antenna type, number, and installation location according
to the requirements of an area.
This section introduces the antenna design guidelines in some typical cases.
(1)

Single cell

If the indoor coverage is realized by a signal cell, each antenna must be designed
to ensure that signals are evenly distributed in the coverage area. Generally, it is
recommended to install the antenna in a zigzag way, as shown in Figure
5-21Figure 5-21.

Figure 5-21 Antenna design guideline in signal cell


(2)

Multi-cells

If the indoor coverage is realized by multiple cells, a certain distance must be


leaved between intra-frequency reuse cells. Each antenna must also be designed
to ensure that signals are evenly distributed in the coverage area of each cell. If
the frequencies are reused frequently, it is recommended to install the antennas
on different layers at the same position of the layer, as shown in Figure
5-22Figure 5-22.
2006-01-05

All rights reserved

Page 79 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-22 Antenna design guidelines in multi-cells


(3)

Closed building

A closed building has the characteristics, such as thick outer wall, great signal
attenuation, and little leakage. In addition, it is little affected by outdoor
intra-frequency cells. Therefore, the frequency between floors is easily to be
planned. For the antenna design guideline in a closed environment, see Figure
5-23Figure 5-23.

Figure 5-23 Antenna design guideline in closed environment


(4)

Half-open environment

For a half-open building, the outer wall is made of glasses, so the signal
attenuation is small. Within the building are the open conference halls, which are
greatly affected by outdoor intra-frequency cells, so you must plan dedicated
frequencies or adopt the multi-antenna system with low output power to limit the
edges of the indoor cells within the building, as shown in Figure 5-24Figure 5-24.

Figure 5-24 Antenna design guideline in half-closed environment


2006-01-05

All rights reserved

Page 80 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

(5)

Confideniality level

Frame-structure building

For a frame-structure building, the number of internal walls is large and they are
thick. Therefore, if the antenna is installed at the corridors, the antenna output
power must be high so that good coverage can be ensured. In this case, signals
will leak at the windows near the corridor, so you must plan dedicated frequencies
for the building. The distance of the intra-frequency cells between floors is larger
than that in other environments. For the antenna design guideline in
frame-structure building, see Figure 5-25Figure 5-25.

Figure 5-25 Antenna design guideline in frame-structure building


(6)

Office building

The indoor environment of office buildings requires high grade of services, so its
coverage is realized by several directional and omni antennas. You can control
the coverage area easily through properly designing the effective radiation power
in the cell. For design guideline, see Figure 5-26Figure 5-26.

Figure 5-26 Antenna design guideline in office building


(7)

Parking area

2006-01-05

All rights reserved

Page 81 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Parking area has no special requirement on capacity and mobile station receiving
level (-90 dBm). For a parking area, the elevator, escalator, entrance and exit are
key coverage areas, as shown in Figure 5-27Figure 5-27.

Figure 5-27 Antenna design guideline in parking area


(8)

Supermarket

Supermarkets have certain requirements on coverage and capacity. The


antennas can be designed according to actual structure of the buildings. For the
antenna design guideline, see Figure 5-28Figure 5-28.

Figure 5-28 Antenna design guideline in supermarket

III. Survey
The antenna design and installation is finally decided according to the survey,
which includes the following aspects:
Detailed coverage area and signal quality and converge requirements
Distribution of the signals in coverage areas
Composition of buildings in coverage areas
2006-01-05

All rights reserved

Page 82 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Signal access location and mode


Installation position
According to the survey, you must output the final topological structure diagram,
antenna cabling scheme, and list of materials. Generally, the omni antenna is
installed at the ceiling center. The small directional antenna is hung on the inner
side of the outer wall, with the radiation directed to indoor part. In this case, the
effect of the antenna against the outdoor system can be reduced to the minimum,
so the C/I requirement of the outdoor system can be met.
If possible, you can test the coverage and adjust the antenna design according to
the test result, or re-plan the frequency to ensure the voice quality. Generally, if
the radiation power at the antenna interface is 10 dBm, the 2 dBi small indoor
omni antenna is used. In this case, if the walls are densely distributed in the areas
within 30 meters from the antenna, the coverage level can reach -70 dBm.

5.10.3 Capacity Analysis and Design


Before analyzing the capacity, you must define the type of the indoor service area.
For details, see Table 5-31Table 5-31.
Table 5-31 Definition of indoor service area type
Indoor
service area

Characteristic

Example

type
The traffic is hard to be predicted.
The population number varies with day
and night.
Public
service area

The capacity characteristics, such as

Airport, shopping

uneven distribution and bursting must

center, and play

be considered.

ground.

The grade of service and the traffic of


each subscriber are similar to that for
outdoor cells.

2006-01-05

All rights reserved

Page 83 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

The

existed

fixed

Confideniality level

networks

are

frequently used.
The traffic is relatively fixed and easy to

Office

Commercial

be calculated.

and commercial

service area

High service quality is required.

hotels

Generally, the grade of service (GoS) is

building
of

high

ranks.

1%, the traffic of each subscriber can


reach 0.1 Erl.

For the cell organization mode of distributed antenna system, see Figure
5-29Figure 5-29.

Figure 5-29 Cell organization mode of distributed antenna system


As shown in Figure 5-29Figure 5-29, there are two cell organization modes of
distributed antenna system, namely, single cell and multiple vertical split cells.
The single cell is applied to the indoor environment which requires smell
coverage area. The multiple vertical split cells are applied to the indoor
environment with dense traffic.
Likewise, a single cell will split when the capacity does not meet the requirement,
with vertical splitting the splitting mode. Generally, a cell will vertically split into at
least three cells so that frequency reuse can be ensured. Four layers must be
present between two intra-frequency cells, as shown in Figure 5-30Figure 5-30.
To avoid interference between frequencies, you must take measures to prevent a
cell from horizontally splitting.

2006-01-05

All rights reserved

Page 84 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-30 Vertical cell splitting

5.10.4 Frequency Planning


If the dedicated frequency is adopted in indoors, the frequency planning is
relatively simple. Generally, the frequency reuse mode in business service areas
is almost the same as that in pubic service areas. If the frequency resource is
adequate, you must try best to use dedicated band for indoor coverage. If not,
you can search the available channel numbers with relatively small interference
through scanning the channel numbers. If the frequency resources of the 900
MHz cannot meet requirements, you can introduce the 1800 MHz frequency;
namely, use a dual-band system.
If you steal frequency resource for indoor system due to no available dedicated
frequency, you must pay attention to the followings:
Do not select the frequencies of the neighbor cells.
Ensure that the BCCH frequencies are not interfered.
The interference on the TCH frequencies can be reduced with the help of
radio frequency hopping.
Search the available uplink frequencies through using BTS equipments to
scanning the uplink channel numbers.
Search the available downlink frequencies through using drive test
equipment to scanning the downlink channel numbers.
If the hierarchical cell structure is not used, the cell with the strongest signal
level is the service cell, and the interference from neighbor frequencies can
be neglected.
If the hierarchical cell structure is used, the cell with the strongest signal level
cannot necessarily be the service cell, so you must take measures to reduce
the interference from neighbor frequencies.
Because the environment is urban areas is quite complicated, especially the
effect of the antenna back lobe is present, the service areas for high buildings are
greatly interfered, so you must carefully plan the frequencies for the indoor
coverage of high buildings. Generally, for the lower floors, you can plan the
2006-01-05

All rights reserved

Page 85 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

frequencies according to general method. For the higher floors where the
interference is strong, you can use dedicated channel numbers. However, the
final frequency planning must be based on practical tests.

5.10.5 Traffic Control


The indoor coverage system for high buildings can be taken as a system
independent of outdoor systems if the coverage of the indoor system is good.
Theoretically, you can only consider the cell selection and reselection, handover
relationship, and the compact on outdoor networks at the entrances and exits of
the building.
However, the actual conditions are quite complicated. For example, the signals
outside of the building may be strong. In this case, if a mobile station is powered
off, it may camp on an outside cell. Therefore, when optimizing the network, you
must set the one-way adjacent cell and two-way adjacent cell according to actual
conditions and set the parameters, such as CRO and TO to a proper value
according to the regularity of cell selection and reselection. In addition, you can
set the indoor cells to a high priority so as to reserve more traffic. And the
inter-layer handover threshold and hysteresis are defined and adjusted according
to actual conditions.

5.11 Tunnel Coverage


5.11.1 Characteristic of Tunnel Coverage
At present, most of the tunnels are dead zones, so you must make out special
solutions for tunnel coverage. The tunnel types include railway tunnel, highroad
tunnel, and underground railway tunnel. Each tunnel has its characteristics, and
they are specified as follows.
For the highroad tunnel, it is wide. The coverage in the highroad tunnels is
relatively stable. When there are vehicles passing by, you can select the
antennas with a larger size to obtain a higher gain, so the coverage distance is
larger.
For the railway tunnel, it is narrow, especially when there is a train passing by;
only a little room is left in the tunnel, so the radio propagation is greatly affected.
Moreover, the train has great effect on radio signals. Since the antenna
installation room is quite limited, the antenna size and gain are greatly restricted.
In addition, because general cars cannot be driven to such tunnels, the tunnel
coverage is hard to be tested. Therefore, the planning for highroad coverage is
different from that of the railway coverage.
The length of tunnels ranges from several hundred meters to several kilometers.
For short tunnels, you can adopt flexible and economical means to realize the

2006-01-05

All rights reserved

Page 86 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

coverage. For example, you can install a general antenna near one end of the
tunnel, with the radiation directed to the inside. For long tunnels, however, you
must adopt other means. Actually, the coverage solution varies with tunnels, so it
is designed according to actual conditions.
Figure 5-31Figure 5-31 and Figure 5-32Figure 5-32 show the cross section of the
single-track railway tunnel and multi-track railway tunnel. The smaller the area of
the cross section, the greater the loss when a train passes through the tunnel.
The related calculation and analysis are based on the multi-track railway tunnels
and highroad tunnels. For the calculation and analysis for single-track tunnels,
the protection margin can be 5 dB greater than that of multi-track railway tunnels.

Figure 5-31 Cross section of single-track railway

Figure 5-32 Cross section of multi-track railways


Before planning tunnel coverage, you must prepare for the following data:
Length of the tunnel
Width of the tunnel
Number of tunnel holes (1 or 2)
Needed coverage probability (50%, 90%, 98% or 99%)
Structure of the tunnel (it is constructed with metals or concretes)
Number of needed carriers (130)
Minimum receiving level in the tunnel (generally, it ranges from -85 dBm to
-102 dBm)
Distance between tunnel holes
Whether AC/DC is available
2006-01-05

All rights reserved

Page 87 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Whether the hole can be punched in the tunnel wall


Signal level at the tunnel entrance
Existed signal level in the tunnel

5.11.2 Tunnel Coverage Solution


I. Link budget
Indoor radio link loss is mainly decided by path loss medium value and shadow
fading. A tunnel can be taken as a tube. The signals are transmitted through the
reflection of walls and straight transmission, with straight transmission the major
form. ITU-R suggests an indoor propagation model on page 1238, which is also
effective for tunnel coverage. The formula is as follows:
Lpath = 20 lg f + 30 lg d + Lf (n) - 28 dB
Here,
f indicates frequency (MHz)
d indicates distance (m)
Lf indicates penetration loss factors between floors (dB)
n indicates the number of floors lying between the mobile station and
antenna.
The Lf (n) can be neglected in tunnel coverage, so the following equation can be
applied in the calculation of the radio propagation in tunnels. That is:
Lpath = 20 lgf + 30 lg d - 28 dB
Table 5-32Table 5-32 lists the path loss in different tunnels.
Table 5-32 Path loss in different tunnels
Distance (m)

GSM 900MHz (dB)

GSM 1800MHz (dB)

50

82.0

88.1

100

91.0

97.0

150

86.3

102.1

200

100.1

106.1

300

105.3

111.4

II. GSM signal source selection


A GSM signal source and a set of distributed antenna system are a must for
tunnel coverage. For tunnel coverage, the GSM signal source is selected
according to the radio coverage, transmission, traffic, and the existing network

2006-01-05

All rights reserved

Page 88 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

equipments near the tunnel. A macro cell base station, a micro cell base station,
or a repeater can work as a GSM signal source for the tunnel coverage.
For the coverage of railway tunnels and highroad tunnels, the indoor macro cell
base station is seldom used as signal source, but it can be used for an
underground railway which requires the coverage of platforms and entrances. In
this case, the capacity of the signal source must be great. In most cases, however,
the tunnel coverage is realized by micro cell signals.
For the areas to be covered, if the nearby network capacity is adequate, the
capacity expansion is unnecessary. And if there are good GSM signals available,
namely, the donor signal level meets the requirements of a repeater (for example,
-70 dBm); a repeater can work as the signal source for the tunnel coverage. With
the increase of traffic, however, you must use GSM base stations to replace the
repeaters.
Adequate isolation must left between donor antenna and retransmission antenna,
though it will cause difficulty in antenna installation. Generally, the log-periodical
antenna with great front-to-back ratio is used as the retransmission antenna.
The general antenna (wireless repeater), coaxial cable, and optical fiber (optical
repeater) can connect a repeater to a donor cell.
For tunnel coverage, the installation space and auxiliary equipments are quite
limited, so micro cell base stations and repeaters instead of macro cell base
stations are often applied in tunnel coverage.
In mountain areas, repeaters are more likely used because strong signal level
often exists at the mountain tops near the tunnel. In this case, the antenna
isolation requirement can be easily met. If the signal level of the existed network
near the tunnel is not strong enough, you can use a micro cell for the tunnel
coverage.

III. Antenna feeder system selection


After deciding the GSM signal source, you must configure the antenna feeder
system for the tunnel coverage according to actual conditions. Three types of
configuration are available, namely, coaxial feeder passive distributed antenna,
optical fiber feeder active distributed antenna, and leaky cable. Hereunder
introduces the tunnel coverage based on coaxial feeder passive distributed
antenna and leaky cable.

5.11.3 Tunnel Coverage Based on Coaxial distributed antenna system


In a coaxial distributed antenna system, the following RF components are used:
Feeder (3/8", 1/2", or 7/8") and jumper
Power splitter
Power splitter
2006-01-05

All rights reserved

Page 89 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Antenna
This section introduces three tunnel coverage solutions based on the coaxial
distributed antenna system.

I. Solution 1
Figure 5-33Figure 5-33 shows the tunnel coverage solution based on the
bi-directional passive distributed antenna system.

Figure 5-33 Tunnel coverage solution based on bi-directional passive distributed


antenna system
According to this solution, if the needed minimum signal level is -85dBm (the
location probability is 50%), you must add a margin of 8 dB if the want to enhance
the location probability to 90%.
If the gain of the bi-directional antenna is 5 dBi, the loss of the equal probability
power splitter and the jumper is 2 dB, and the feeder with the specification of 7/8"
is used, the path loss in 100 meters is 4 dB and the output power of the
equipment is 39 dBm.
Suppose that the level of the signals transmitted by the first bi-directional antenna
is -85 dBm at the tunnel entrance, you can calculate the distance between the
antenna and the tunnel entrance using the following equation:
Pout- Lpath (d) Lcable (d) Ljumper + Gant = -85dBm + 8dB90%_loc.Prob
Here,
Pout indicates the output power (39dBm).
Lpath (d) indicates the path loss from the first bi-directional antenna to the
tunnel entrance.
Lcable (d) indicates the cable loss.
Ljumper indicates the jumper loss (2 2 dB).
Gant indicates the antenna gain (5 dBi).
If introducing the previous data to the equation, you can obtain the sum of the
Lpath (d) and Lcable (d), that is, 117 dB.
For the relationship between distance d and Lpath (d) and Lcable (d),
see Figure 5-34Figure 5-34, in which the curve indicates Lpath (d) and the slant
line indicates Lcable (d).
2006-01-05

All rights reserved

Page 90 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-34 Relationship of path loss and 7/8 cable loss


According to Figure 5-34Figure 5-34, you can obtain that d = 301m through
estimation.
If a power splitter is adopted for the first antenna, a loss of 3dB must be added. In
this case, the sum of Lpath (d) and Lcable (d) is 114 dB.
According to Figure 5-34Figure 5-34, you can also obtain that d = 261m through
estimation.
For railway tunnels, train filling will affect signal propagation, so a protection
margin of 5dB must be considered when the antenna is installed in the tunnel. In
this case, d = 240m. That is, if a bi-directional antenna is installed in the tunnel, it
can coverage a distance of 480m.
If a power splitter is adopted for the second antenna, the coverage distance
between the first antenna and the second antenna will be shortened unless an
amplifier is used.
The followings analyze the coverage when no amplifier is adopted for the second
antenna.
The total power output by the first power splitter (it is installed at the first antenna)
Pout1 is expressed as follows:
Pout1 = Pout Lcable (d) - Ljumper - Lsplitter = 39dBm Lcable (261m) - 2dB 3dB= 23.56 dBm. (The cable loss in 261m is about 10.44 dB, jumper loss is 2 dB,
and the power splitter intersection loss is 3dB).
Suppose the overlapping level between the two antennas is -85 dBm, the
distance between the second antenna and the first antenna is: d2 = d + x. Here,
d indicates the coverage distance of the first antenna (261m), and x indicates
the coverage distance of the second antenna in the single direction.
According to the previous analysis, the following two equations can be obtained:
2006-01-05

All rights reserved

Page 91 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Pout1 Lcable (261m) Lcable (x) Ljumper + Gant Lpath (x) = - 85dBm
+ 8dB90%_loc.Prob
Lpath (x) + Lcable (x) = 108.56dB
Plus the two equations, you can obtain the value of x, that is, 100m. This means
that when no amplifier is adopted, two antennas can coverage a tunnel distance
of 722m, namely, 2*(261 + 100) m = 722m.
If you adopt cascaded antennas, the transmit power is relative low due to the
coaxial cable loss. In this case, you can use the amplifier to amplify the power.

II. Solution 2
If a tunnel is not long, you can adopt a simpler coverage mode, as shown
in Figure 5-35Figure 5-35.

Figure 5-35 Tunnel coverage solution based on a single antenna


According to this solution, a directional antenna is installed at the tunnel entrance,
with the radiation directed to the inside. The following analyze this coverage
solution.
In this solution, Pout = 39 dBm (suppose that the output power of the GSM signal
source is 8W).
If the Lpath (d) indicates propagation loss, the sum of Lcable (d) and Ljumper is
5dB, the antenna gain Gant is 8 dBi, and the needed received level is -77dBm,
the Lpath (d) is expressed as follows:
Lpath (d) = 39dBm - 5dB + 8dBi (-77dBm) = 119 dB
According to the equation Lpath (d) = 20 lg10f + 30 lg10d - 28 dB, the value of d
can be obtained, that is, 858m.
The previous analysis is applicable to highroad tunnels. For railway tunnels, you
can consider a margin of 10 dB due to the effect of train filling, but the coverage
distance of the antenna in railway tunnels is calculated the same as that in
highroad tunnels. According to the calculation, d = 398m.

5.11.4 Tunnel Coverage Based on Leaky Cable System


If adopting leaky cables to realize the tunnel coverage, you must find the
specifications of the leaky cables and complete the leaky cable design according
to the following steps:
2006-01-05

All rights reserved

Page 92 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

1)

Decide coverage factor

2)

Calculate the gain of the bi-directional amplifier

3)

Estimate the length of the leaky cable between the feeder source and the
first amplifier

4)

Estimate the length of the leaky cable between the amplifiers

5)

Decide the number of needed amplifiers

The followings describe these steps in details.

I. Decide coverage factor


The following information is needed for deciding the coverage factor:
Coupler loss
Number of carriers
Coverage probability
Coverage factor indicates the loss in the areas 2 meters beyond the leaky cable
(along the vertical direction). This loss includes the coupler loss of the leaky cable
and protection margin required by the coverage probability. If 90% of coverage
probability is required, you must add 8dB to the medium level. Some leaky cables
specify the relationship between the coverage probability and coupler loss.
The coverage factor is determined by the parameters, such as coupler loss, RF
carrier number, coverage probability, and tunnel type. For the decision of
coverage factor in concreter tunnels, see Figure 5-36Figure 5-36. For the
decision of coverage factor in metal tunnels, see Figure 5-37Figure 5-37. When
deciding the coverage factor, you can fix a point in the graph and mark a
horizontal line through this point, and this line intersects required coverage
probability. This intersection point is the coverage factor.

Figure 5-36 Coverage factor in concrete tunnels


2006-01-05

All rights reserved

Page 93 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-37 Coverage factor in metal tunnels


For example, if the leaky cable with a coupler loss of 71 (900 MHz) is used, the
RF carrier number is 18, and the coverage probability is 90, the coverage factor in
a concrete tunnel is -77 according to Figure 5-36Figure 5-36.

II. Decide cable length between GSM signal source and the first amplifier
Before deciding cable length between GSM source and the first amplifier, you
must obtain the following information:
Transmit power of the signal source (dBm)
Jumper loss: 1 dB
Connector loss: 1 dB
Leaky cable loss: 2 dB
Transmit power at the feeder source (dBm)
When calculating the power at a point of the feeder, you must subtract the feeder
propagation loss from the GSM signal source. If a wireless repeater with an
output power of 18 dBm (18 carriers) is used as the GSM signal source, and the
attenuation from the jumper to feeder, and from the feeder to the leaky cable is 7
dB (That is, the power from the repeater is transmitted from a jumper to a feeder,
and then from the jumper to a leaky cable, so four connectors are needed.
Generally, the attenuation is 2 dB for each jumper, 1 dB for each feeder, and 0.5
dB for each connector, so the total attenuation is 7 dB.), the transmit power at this
point is 11 dB. For the connection of leaky cable, see Figure 5-38Figure 5-38.

2006-01-05

All rights reserved

Page 94 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-38 Connection scheme of leaky cable


Suppose the needed signal level in a tunnel is -85 dBm, the signal level at the first
amplifier must be equal to or greater than -85 dBm. The coupler loss and
longitudinal propagation loss of the leaky cable are present between the signal
feeder point and the first amplifier. They are calculated according to the following
equation:
LossLong = 11dBm (-85dBm) + Losscoup. Here, Losscoup indicates the coverage
factor, and it is -77dB when 90% coverage is ensured. Therefore, the LossLong is
19 dB (that is, 11dBm + 85dBm -77dB = 19dB).
The cable length between the signal feeder source and the first amplifier can be
obtained according to Figure 5-39Figure 5-39 and Figure 5-40Figure 5-40. For
example, suppose that the attenuation is 4.3dB/100 for the leaky cable, you can
mark a plumb line at the point indicating 4.3dB. This plumb line will intersect the
curve indicating 19 dB at a point, and then you mark a horizontal line starting from
this point. The horizontal line will intersect the right vertical axis at a point. And
this point shows the cable length. According to this example, the distance
between the signal source and the first amplifier is 440m (that is, 19/4.3 = 440m).

2006-01-05

All rights reserved

Page 95 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-39 Cable length between amplifiers in metal tunnels

Figure 5-40 Cable length between amplifiers in concrete cables


According to the previous figures, the left vertical axis indicates Required
RADIAMP Gain, which can be replaced by the radial loss of the leaky cable, but
it makes no difference.

III. Needed amplifier gain


Before calculating the maximum amplifier gain, you must collect the following
information:
The minimum acceptable signal level (dBm)
Coverage factor (dB)
The maximum output loss allowed by a single carrier (dBm)
2006-01-05

All rights reserved

Page 96 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

If the amplifier is not added, the signal level output by the leaky cable for the
longest transmission distance is equal to the difference of the minimum
acceptable signal level and the coverage factor.
The signal level at the leaky cable beyond the longest transmission distance may
be lower the minimum acceptable level, so an amplifier must be added to amplify
the signals to the maximum output power allowed by a single carrier. The
amplification of this power is related to the specifications of the amplifier and the
number of carriers. If the maximum output power allowed by a single carrier is
known, the amplifier gain can be calculated as follows:
Needed amplifier gain = the maximum output power allowed by a single carrier (it
depends on the number of carriers) (the minimum acceptable signal level
coverage factor)
Along the leaky cable, the maximum output power allowed by each carrier of a
bi-directional amplifier is related to the number of carriers that have been
amplified. This is considered mainly for the intermodulation interference is
present, because the intermodulation interference will increase with the total
number of carriers that have been amplified, as shown in Figure 5-41Figure 5-41.

Figure 5-41 Relationship between the maximum output power allowed by a


single carrier and the number of carriers that have been amplified
Needed amplifier gain = the minimum acceptable signal level coverage factor +
the maximum output power allowed by a single carrier.
According to the previous equation, if the minimum acceptable signal level is -85
dBm, the coverage factor is -77, and the maximum output power allowed by a
single carrier is 5 dBm, the needed amplifier gain is 13 dB.

2006-01-05

All rights reserved

Page 97 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

IV. Decide cable length between amplifiers


Before deciding the cable length between amplifiers, you must know the needed
amplifier gain and the cable loss (dB/100m). Figure 5-39Figure 5-39 and Figure
5-40Figure 5-40 help you decide the cable length between amplifiers. For
example, in a concrete tunnel, if the amplifier gain is 13 dB and the cable
attenuation is 4.3dB/100m, the cable length between two amplifiers is 300m.

V. Decide the number of needed amplifiers


Before deciding the number of needed amplifiers, you must know the following
information:
The cable length between the feeder source and the first amplifier
The cable length between amplifiers
The tunnel length
If the previous information is known, the following formula can be used to
calculate the number of needed amplifiers. That is:
The number of amplifiers = (the tunnel length the cable length between the
feeder source and the first amplifier)/(the cable length between amplifiers),
rounding up to the nearest integer.
According to the formula, if the tunnel length is 1000m, the cable length between
amplifiers is 300m, and the cable length between the feeder source and the first
amplifier is 420m, 2 amplifies are needed. That is, (1000 420)/300 = 1.93, so
the nearest integer is 2.
After deciding the number of needed amplifiers, you can optimize the distance
between amplifiers. That is, you can obtain the distance between the two
amplifiers by dividing the remaining distance by the number of needed amplifier.
According to the previous example, it is 580/2 = 290m, namely, the distance
between the two amplifiers is 290m, as shown in Figure 5-42Figure 5-42.

Figure 5-42 Distance between amplifiers

VI. Remarks on leaky cable installation


The leaky cable must not touch any metal. Generally, a leaky cable must be
installed at a spot 5m away from concrete walls and at least 10m away from metal
walls. In addition, a leaky cable must be installed near to the coverage area. You

2006-01-05

All rights reserved

Page 98 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

cannot necessarily consider the line-of-sight propagation, because the signals


leaking from the cable will fill the space nearby.

5.11.5 Coverage Solutions to Tunnels in Different Length


This section introduces the coverage solutions to tunnels in different length. In
actual networking, the following coverage solutions may be used:
Micro base station (or repeater) + a single antenna
Micro base station (or repeater) + distributed antenna system
Micro base station (or repeater) + leaky cable
Before deciding which coverage solution should be adopted, you must consider
the followings:
Is the GSM signal near the tunnel entrance strong enough?
Is there any available transmission link near the tunnel?
Generally, if the existed signal level near the tunnel entrance (including nearby
mountains) is lower than -80 dBm, the micro base station is recommended. If it is
greater than -80 dBm, the micro base station or the repeater is recommended. If
problems concerning transmission are present, the repeater is recommended.
When using the repeater, you must consider that certain isolation is required
between repeaters.

I. Coverage solution to short tunnels


Generally, the tunnels shorter than 100m are defined as short tunnels. When
planning the coverage for these tunnels, you must consider the coverage areas
near the tunnels. If several tunnels are close to each other, you can install a base
station or a repeater between the tunnels. If adopting a micro base station, you
must adopt the bi-directional antenna. If the antenna gain is 5 dBi, you should
install the antenna at the tunnel entrance so as to ensure coverage.
When designing tunnel coverage solutions, you must fully consider that fact that
cars and trains move at a high speed, so how to ensure normal handover after
the cars or trains steering into the tunnels is of vital importance.
If the repeater is used as the GSM signal source and the signals outside the
tunnel and the signals within the tunnel belong to the same cell, no handover
problem will occur. If the micro cell is used as the GSM signal source and the
signals outside the tunnels and the signals within the tunnel belong to different
cells, the signals in the outside cell will drop dramatically when the train steers
into the tunnel. In this case, handover failure may occur and call drop will be
resulted in.
To solve this problem, you can consider adopting the following methods:
Adopt the bi-directional antenna for the tunnel coverage, because it can
provide enough overlapping area for handover.
2006-01-05

All rights reserved

Page 99 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Enable special handover algorithms, such as fast level fall handover


algorithm. In this case, a mobile station can hand over to another cell when
the signal level falls fast.
Select the directional antenna with small front-to-back ratio.

II. Coverage solution to middle-length tunnels


This section introduces several typical coverage solutions to railway tunnels.
The followings are a series of assumptions:
The Huawei BTS3001C (the maximum output power is 8W) is used as the
GSM signal source.
The repeater with 1 amplified carrier and a maximum output power of 2W is
considered.
The lowest receiving level is designed to -85 dBm, and the coverage
probability is 90% (with a protection margin of 8 dB).
For railway tunnel coverage, because the train will affect signal transmission,
if the antenna is installed at the tunnel entrance, the protection margin must
be increased by 10 dB. If the antenna is installed in the tunnel, the protection
margin must be increased by 5dB.
The dedicated directional antenna with the specification of DB771S50NSY,
the horizontal half power angle of 60, and the antenna gain of 8 dBi is used
at the tunnel entrance.
The bi-directional antenna with the specification of K738446 and antenna
gain of 5 dBi is used within the tunnel.
According to these assumptions, if a micro base station (39 dBm) is used as the
GSM signal source, the coverage distance is 400m when the antenna with a gain
of 8 dBi is installed at the tunnel entrance, and the coverage distance is 480m
when the bi-directional antenna with a gain of 5 dBi is installed in the tunnel.
If a repeater (33 dBm) is used as the GSM signal source, the coverage distance
is 250m when the antenna with a gain of 8 dBi is installed at the tunnel entrance,
and the coverage distance is 360m when the bi-directional antenna with a gain of
5 dBi is installed in the tunnel.
Therefore, for the tunnels shorter than 500m, you can use the combination of a
micro base station and a single antenna (or a repeater) for the tunnel coverage.
For curve tunnels, you can install a bi-directional antaean in the tunnel.
According to on-site survey on the cross-section, the available antenna size, and
the tunnel length, you can use the antenna with a higher gain to coverage the
tunnels a little longer than 500m.

III. Coverage solution to long tunnels


For the tunnels longer than 500m, you need to use the distributed antenna
system or the leaky cable for the coverage. The followings introduce the coverage
2006-01-05

All rights reserved

Page 100 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

realized by the combination of a micro base station and a leaky cable (or a
repeater).
Hereunder is a series of assumptions:
The Huawei BTS3001C (the maximum output power is 8W) is used as the
GSM signal source.
The repeater with 1 amplified carrier and a maximum output power of 2W is
considered.
The lowest receiving level is designed to -85 dBm, and the coverage
probability is 90% (with a protection margin of 8 dB).
The leaky cable with the specification of SLWY- 50- 22 and the radial loss of
5dB/100 m is used.
The coupler loss may be 77 dB when the 90% of signals are received.
According to these assumptions, if a micro base station (39 dBm) is used as the
GSM signal source, the coverage distance is 800m when only the leaky cable but
no amplifier is used. If a repeater (33 dBm) is used as the GSM signal source, the
coverage distance is 680m when only the leaky cable but no amplifier is used.
The coverage distance will be larger if leaky cables with smaller loss are used.
For the coverage of still longer tunnels, you must use amplifiers to amplify signals.
That is, you can use either the distributed antenna system or the leaky cable for
the coverage solution. In terms of technical indexes and installation space,
coverage solution based on leaky cable is recommended. In terms of cost, you
must select a suitable coverage solution base on actual conditions.

5.12 Repeater Planning


5.12.1 Application Background
With rapid development of mobile communication networks, people have higher
requirements on service quality. They hope to enjoy mobile services anywhere
and anytime. As for telecommunication carriers, they cannot enable a base
station in some dead zones due to the reasons such as cost and transmission
conditions. In this case, a repeater can provide an auxiliary and economical
means to coverage the dead zones.

I. Repeater types
For the division of repeater types, see Table 5-33Table 5-33.

2006-01-05

All rights reserved

Page 101 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Table 5-33 Division of repeater types


Division
standard

Type

Remark

A wireless repeater adopts a set of donor


antenna to receive the signals from the base
station. After amplifying the signals, it adopts a
Wireless repeater

set of retransmission antenna to forward the


signals in another direction. Generally, a

Transmission

wireless repeater has only one receiving path,

mode

so the diversity antenna is unnecessary.


An optical repeater transmits signals using
Optical repeater

optical fibers, so the repeater side and base


station side must have the optical transmission
capability.

Bandwidth

selection

repeater

A bandwidth selection repeater is also called


wideband repeater, and it can select a
frequency (for example, the frequency with a
bandwidth of 6M, 19M, or 25M) and amplify it.

Channel
A channel selection repeater is also called

bandwidth
Channel

selection

repeater

narrow band repeater or frequency selection


repeater. It amplifies the selected channel
numbers only. It is a narrow band repeater and
amplifies a limited channel numbers.

New style

Solar

energy

repeater

Wireless

A solar energy repeater is of the wideband


type. It is similar to a general wideband
repeater except that its power is solar energy.

frequency

selection repeater

Product type

Optical

frequency

selection repeater

Currently, the types of the repeaters listed in

Wireless

the left column are in commercial use.

wideband

repeater
Optical

wideband

repeater

2006-01-05

All rights reserved

Page 102 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

II. Comparison between repeater and micro cell


For the comparison between the repeater and micro cell, see Table 5-34Table
5-34.
Table 5-34 Comparison between repeater and micro cell
Micro cell

Repeater

Many equipments and a long period


are needed for constructing a micro
cell.

A repeater is installed in a flexible way


and the base station equipments and
transmission

equipments

are

unnecessary.

A micro cell can expand the system


capacity. When the cells near a base
station are busy, a micro cell can be
used to ease the congestion.

A repeater can absorb traffic. When a


cell is idle, it brings the traffic to this
cell, thus enhancing the utilization ratio
of the equipments. A repeater does not
expand the capacity for a system.

The system needs to allocate channel

The system does not need to allocate

numbers to a micro cell, but this is

channel numbers to a repeater, but it

hard to be realized in the areas where

must

the frequency resource is scarce.

interfering with other cells.

prevent

the

repeater

from

Note:
The filter of an intra-frequency repeater will produce a delay of about 5s.
Theoretically, the maximum effective coverage distance of a GSM cell will be
smaller than 35km in this case.

A GSM system must enable the dynamic power control function, which is
transparent to a repeater. Generally, you must adopt the automatic level control
technologies (ALC) for a repeater.

Note:
When the ALC technology is applied to a repeater, if a mobile station is too near
to the repeater, the repeater will reduce the gains for all the mobile stations within
its service area. In this case, the conversation quality of some mobile stations will
become poor, or even call drop may occur; especially the mobile stations far
away from the repeater are greatly affected.

2006-01-05

All rights reserved

Page 103 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

III. Application characteristics


Repeaters are mainly used to cover the dead zones in vast open land, and they
are the extension of the base stations. A repeater improves the coverage but
does add up to the traffic capacity of a network. However, because it enlarges the
coverage of the base station, the total traffic volume increases.
A wireless repeater applies the radio transmission mode, with short construction
period and effective cost. An optical repeater adopts optical fiber as transmission
medium, so the transmission loss is small and transmission distance is large, but
construction cost is greater than that of the wireless repeater.
The application advantage of the wireless repeater lies in low transmission
requirement. If you plant the optical fiber, there is no price advantage against the
construction of a micro cell base station. In this case, considering the network
quality, you are recommended to select the micro cell base station.
Compared with wideband repeater, a narrow band repeater has better
performance and provides better signal quality. However, the following problems
are still present in application:
The carriers of a narrow band repeater must outnumber the carriers
configured for the source base station; otherwise the repeater cannot
capture a channel.
The number of paths of many repeaters is set to 4, so the base stations
outnumber 4 carriers cannot work as the signal source.
For the base stations with radio frequency and frequency hopping, if the
frequencies in the frequency hopping set outnumber the paths selected by
the repeater, the conversation cannot be maintained.
When the channel number of the donor cell of the repeater changes, you
must adjust the channel number, otherwise the problems such as channel
assignment failure, call drop, and interference will occur.
The wideband repeater allows the base station to adopt frequency hopping, and
you do not have to adjust the channel number of the repeater after the channel
number of the donor cell changes if the channel number is within the bandwidth of
the repeater. However, the wideband repeater will amplify all the signals within
the band, so it causes great interference against other cells.
No matter whether the optical fiber or wireless repeater is applied, the sum of the
radius of the service area of the repeater and the distance between the repeater
and base station cannot break the TA limitation. For general base stations, the
distance between a repeater and the base station must be shorter than 35
kilometers.
The optical repeater can be used in the areas where the GSM radio signals
cannot reach and no space is left for a repeater. Because the transmission loss of

2006-01-05

All rights reserved

Page 104 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

optical fiber is small and its bandwidth is wide, the optical repeater is quite helpful
for transmitting RF signals.
Either an omni antenna or a directional antenna can be selected for an optical
repeater according to the actual landforms. For an optical repeater, its
transmission does not have to be isolated from the reception. In addition, the
address of an optical repeater is easy to be decided. Generally, an optical
repeater is applied in the dead zones within countryside, highroads, touring areas,
factories, and urban areas.
In remote mountain areas and along highroads, you can also consider using a
solar energy repeater.
In conclusion, the repeater is used for the following purposes:
Enlarge coverage area and eliminate dead zones.
Strength the field strength and enlarge converge of the base stations in
urban areas.
Ensure the coverage along the highroads and tunnels.
Realize indoor coverage.
For the application of the wireless repeater and optical repeater, see Figure
5-43Figure 5-43 and Figure 5-44Figure 5-44.

Figure 5-43 Application of wireless repeater

2006-01-05

All rights reserved

Page 105 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-44 Application of optical repeater

5.12.2 Working Principles of Repeater


I. Wireless frequency selection repeater
Figure 5-45Figure 5-45 shows the working principles of a wireless frequency
selection repeater. The repeater receives the RF signals from the selected base
station (donor antenna) and amplifies and forwards the signals. The antenna
receiving the signals from the base station is called donor antenna, the other
antenna is called retransmission antenna.

Figure 5-45 Working principle of wireless frequency selection repeater

2006-01-05

All rights reserved

Page 106 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

According to Figure 5-45Figure 5-45, the working principles of a wireless


frequency selection repeater are as follows:
1)

The low-noise power amplifier processes the signals (received by the donor
antenna) from downlink carriers.

2)

The signals (900 MHz RF signals) are down converted into 71 MHz
intermediate frequency (IF) signals.

3)

The IF filter (with a bandwidth of 200 KHz) amplifies the 71 MHz IF signals
and up converts the signals into the 900 MHz RF signals.

4)

The retransmission antenna (service antenna) transmits the signals to the


coverage areas.

The uplink signals are also processed according to the previous procedures.

II. Wireless wideband repeater


Figure 5-46Figure 5-46 shows the working principles of a wireless wideband
repeater.

Figure 5-46 Working principle of wireless wideband repeater


The wireless wideband repeater works as the same way as the wireless
frequency selection repeater except the filter part. The bandwidth of the filter of
the wireless wideband repeater is fixed. Generally, it is 6M, 19M, or 25M.

III. Optical repeater


Figure 5-47Figure 5-47 shows the working principles of an optical repeater.

2006-01-05

All rights reserved

Page 107 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-47 Working principle of optical repeater


The difference between the optical frequency selection repeater and the optical
wideband repeater lies in the coverage end. The former adopts the frequency
selection components, but the later adopts the variable bandwidth options.
Compared with the wireless repeater, the optical repeater does require isolation
between donor antenna and retransmission antenna.

5.12.3 Repeater Network Planning


I. Repeater address selection
There is no special requirement on the repeater address selection except the
following items:
A repeater address must lie between the donor base station and the dead
zone, and the azimuth angle between the donor antenna and the
retransmission antenna cannot be smaller than 90, as shown in the
following figure.

If the service antenna is a directional antenna, the repeater must be installed


about 200 to 500 meters beyond the dead zone. If the repeater is installed
within the dead zone, the coverage quality cannot reach the best, as shown
in the following figure.

2006-01-05

All rights reserved

Page 108 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

When the repeater is used to coverage the dense residential areas at the
edges of the urban area, it cannot face the buildings, because great
penetration loss will be caused. In this case, the repeater must be installed at
the one side of the building, as shown in the following figure.

The areas to be covered must meet the requirement of line-of-sight


transmission.
The repeater address must ensure the received signal level required by the
repeater. Generally, the received signal level ranges from -50 dBm to -80
dBm.
No strong carrier whose channel number is the same as that of the donor
base station is present at near the repeater address.
The landforms, buildings, or towers where the donor antenna and
retransmission antenna can be installed. (The donor antenna must be
directed to the base station and the retransmission antenna must be directed
to the service area of the repeater. In addition, the isolation between the two
antennas must be greater than 170 dBc.)

II. Antenna selection


When selecting the antenna for a repeater, you must consider the followings:
Select the proper antenna gain according to the signals and coverage
condition
Do not adopt the omni antenna because the wireless repeater is affiliated to
the intra-frequency relay system, otherwise the system will perform
self-excitation.
2006-01-05

All rights reserved

Page 109 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

The communication between the donor antenna and the donor base station
antenna is point-to-point communication, so you must select the antenna
with high gain or narrow horizontal beam width. For example, to reduce
interference, you can select the reflector antenna or the log-periodical
antenna.
Select retransmission antenna according to the characteristics of a coverage
area. For a large coverage area, you can select the general directional
antenna with high gain. For tunnel coverage, you can select the Yagi
antenna or the spiral antenna. For indoor coverage, you must select the
antenna specially designed for indoor use. No matter in what occasions, you
must control the transmit direction of the retransmission antenna to prevent
the retransmitted signals from feeding in the donor antenna.
The front-to-back ratio of the antenna must be as great as possible (it is
better to be greater than 30 dB) so that a better isolation between the donor
antenna and retransmission antenna can be ensured.

III. Requirements on antenna isolation


The isolation between repeater antennas depends on the host gain, but the host
gain cannot excel the isolation coefficient for self-excitation. According to the
requirements in GSM protocols 03.30, the isolation must be at least 15 dB greater
than the host gain. In actual project design, you can judge whether the installation
position meets the requirements on antenna isolation according to on-site
measurement.
According to the formulas calculating the antenna horizontal isolation introduced
in 5.7.2

II. 5.7.2 II. , the following formula can be deducted:

AH = 31.6 + 20 lgd (Gt + Gr) dB (900 MHz)


AH = 37.6 + 20 lgd (Gt + Gr) dB

(1800 MHz)

Here, d indicates the distance between the donor antenna and retransmission
antenna, in the unit of meter. Gt and Gr indicate the antenna gain relative to the
major lobe in the direction of the two antennas. If the two antennas are
back-to-back installed, Gt and Gr indicate the front-to-back ratio of the antenna,
as shown in Figure 5-48Figure 5-48.

2006-01-05

All rights reserved

Page 110 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-48 Horizontal isolation of repeater antennas


The formula calculating the vertical isolation of repeater antennas is as follows:
Av = 47.3 + 40 logd dB

(900 MHz)

Av = 59.3 + 40 logd dB

(1800 MHz)

Figure 5-49Figure 5-49 shows the vertical isolation.

Figure 5-49 Vertical isolation of repeater antennas


If the horizontal isolation and vertical isolation are present simultaneously, the
total isolation can be calculated by the following formula:
AS = (AV - AH) a/90 + AH, here AV indicates the vertical isolation; AH indicates the
horizontal isolation; and a indicates the antenna included angle.
Figure 5-50Figure 5-50 shows the combination of vertical isolation and horizontal
isolation.

2006-01-05

All rights reserved

Page 111 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Figure 5-50 Combination of vertical isolation and horizontal isolation


Figure 5-51Figure 5-51 shows the antenna isolation when the donor antenna and
retransmission antenna are horizontally installed.

Figure 5-51 Antenna isolation when donor antenna and retransmission antenna
are horizontally installed
As shown in Figure 5-51Figure 5-51, the donor antenna and retransmission
antenna are installed on the top of the building. Suppose the host gain is 100 dB,
the isolation between the two antennas can be 120 dB. If the front-to-back ratio of
the donor antenna and the retransmission antenna is 30 dB, when no barriers are
present between the two antennas, the requirement on the isolation can be met.
If the space loss of the signals between the two antennas is 60 dB, the horizontal
isolation distance can be obtained, that is, d = 26m.
During project implementation, you must select the antenna installation position
according to on-site measurement. You can use a signal source and a receiver for
2006-01-05

All rights reserved

Page 112 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

the repeater. If the signal attenuation between the signal source and the receiver
reaches 60 dB, it means that the antenna installation position meets the
requirement on antenna isolation.
When installing the antenna for a repeater, you must pay attention to the following
items:
If the antennas are horizontally installed, the host of the repeater must be
installed between the donor antenna and the retransmission antenna (it must
be nearer to the donor antenna.)
A good isolation must be ensured regardless that the antennas are
horizontally or vertically installed. When they are horizontally installed, it is
better that there are some barriers lying between the donor antenna and the
retransmission antenna, because you do not have to particularly design a
large installation space to ensure antenna isolation in this case.

IV. Uplink and downlink balancce calculation


For a GSM repeater, the link balance is realized by four links, namely, the uplink
and downlink between the donor base station and repeater, and the uplink and
downlink between the repeater and mobile station.
This section employs the wireless repeater applied in outdoors as an example to
calculate the link balance. To simplify the calculation, we introduce the effective
donor path loss (EDoPL), which includes all the loss and gain from the output end
of the base station combiner or the input end of the multi-path coupler to the input
end of the repeater, as shown in Figure 5-52Figure 5-52.

Figure 5-52 Link balance of repeater


The link balance is calculated according to the following two formulas:
For downlinks, Pbout - EDoPL + GRD - LRF + GRA - Lpass - Pmn = Pmin.
For uplinks, Pmout - Lpass + GRA - LRF + GRU - EDoPL - Pbn = Pbin.
Here,
Pbout indicates the output power of the base station.
2006-01-05

All rights reserved

Page 113 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Pmout indicates the output power of the mobile station.


GRD indicates the downlink gain of the repeater.
GRU indicates the uplink gain of the repeater.
LRF indicates the feeder loss of the retransmission antenna.
GRA indicates the gain of the retransmission antenna.
Lpass indicates the path loss the mobile stations from the repeater to the
service area.
Pbn indicates the attenuation margin of the mobile station.
Pbin indicates the receiving level of the base station.
Pmin indicates the receiving level of the mobile station.
BTSsens indicates the base station sensitivity.
MSsens indicates the mobile station sensitivity.
If the uplink EDoPL and downlink EDoPL are equal to the uplink path loss and the
downlink path loss from the repeater and mobile station, the attenuation margin of
the base station is equal to that of the mobile station. Therefore, if you subtract
the formula calculating uplink balance from the formula calculating downlink
balance, you can get Pbout - Pmout + GRD - GRU = Pmin - Pbin.
If the links are balance, the equation Pmin - Pbin = Dsens = MSsens- BTSsens is present.
In this case, the formula calculating link balance is Pbout - Pmout + GRD - GRU = Dsens.
Therefore, the Dsens is fixed after the base station equipments are selected.
Moreover, the output power of the base station and mobile station may be
decided in GSM system planning. As a result, to achieve the balance of the whole
links, you need to adjust the uplink gain and downlink gain of the repeater only.
The followings employ the repeater system installed in outdoors as an example to
calculate the whole link balance.
For downlink budget of the outdoor repeater , output power of the transmitter
(+43dBm) loss of the combiner (4dB) EdoPL (90dB) = input power of the
repeater (-51dBm) + downlink gain of the repeater (80dB) = downlink output
power of the repeater (+29dBm) feeder loss of the retransmission antenna (3dB)
+ gain of the retransmission antenna (18dBi) path loss of the repeater in the
coverage area (127dB) = input level of the mobile station (-83dBm) attenuation
margin (20dBm) = the mobile station sensitivity (-103dBm).
Note:
To obtain the value of EDoPL, you can measure the input level of the donor
repeater and output level of the base station combiner first, and then obtain the
difference between the two, and the difference is the value of EDoPL. In addition,
the gain of the mobile antenna must be converted to 0 dBi.

For uplink budget of the outdoor repeater, output power of the mobile station
transmitter (+33dBm) path loss of the repeater in the coverage area (127dB) +
2006-01-05

All rights reserved

Page 114 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

gain of the retransmission antenna (18dBi) feeder loss of the retransmission


antenna (3dB) = input power of the repeater (-79dBm) + uplink gain of the
repeater (80dB) = output power of the repeater (+1dBm) EdoPL (90dB) = input
level of the base station (-89dBm) attenuation margin (20dBm) = base station
sensitivity (-109dBm).
Note:
Because you do not have to consider the diversity function, the attenuation
margin on uplinks is the same as that on downlinks. According to the previous link
budget, the downlinks are restricted by the output power of the repeater, the
uplinks are restricted by the output power of the mobile station, and the noise
restricts the maximum gain (EDoPL-10 dB), so the link balance is present.
However, this is the most common situation. Actually, you must calculate the
margin for all links when installing or optimizing the repeater system. The latest
repeater supports the uplink gain and downlink gain to be set respectively.

Hereunder is an example.
There is a base station covering parts of a highroad. Its coverage radius is
about 20 km.
The measured signal strength at the edges of the base station cells is
-93dBm.
The microwave link tower on the top of the hill near the base station is
selected as the address of the repeater.
In the areas (including mountains) 350m below the top of the tower, the
received level of the mobile station is -71 dBm.
The log-periodical antenna with a gain of 18dBi and an azimuth angle of
35is used as the donor antenna.
The antenna is installed at 15 meters under the tower top and faces the base
station.
If the previous conditions are present, the signals output by the repeater are -54
dBm. If a plane antenna with a gain of 17 dBi and a horizontal azimuth angle of 60
degrees is installed at the top of the tower and the antenna radiates to the reverse
direction of the donor antenna, the requirements on antenna isolation can be met
even if the gain of the repeater reaches 85 dB. In this case, the output power of
the repeater is 30 dBm. And the level of the signals in the areas along the
highroad which are 20 km beyond the tower can reach -90 dBm. Therefore, the
radius of the cell along the highroad is enlarged by 50%.
Note:
If a retransmission antenna is installed at the top of the tower, you must ensure
that the received signal level in the zero point filling areas near the tower.

2006-01-05

All rights reserved

Page 115 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

V. Repeater output power control


When adopting a repeater, you must pay special attention to the effect of the
intermodulation products against the system. The intermodulation products of the
repeater depend on the number of the amplified carriers, the output power of
each carrier, and the linearity of the amplifier. For the linearity of the amplifier,
see Figure 5-53Figure 5-53.

Figure 5-53 Linearity of the amplifier


Third order intermodulation will increase with output power due to the nonlinearity
of the amplifier. Therefore, you must control the output to a certain degree to
ensure that that the indexes on third order intermodulation meet the requirements.
The following formula shows the relationship between the output power of each
carrier of the repeater and the requirements on third order intermodulation.
Po = IP3 + (PIMP/2) +10 lg (N/2)
Here,
Po indicates the output power of each carrier (dBm)
IP3 indicates the third order section of the amplifier (dBm)
PIMP indicates the level of the third order intermodulation (dBc)
N indicates the number of carriers
If the third order section of the amplifier of a typical repeater is 50 dBm, and the
intermodulation level must be lower than -45 dBc according to the requirement of
the wireless communication institutes in Britain, the relationship of the output
power of each carrier and the number of carriers is listed in Table 5-35Table 5-35.

2006-01-05

All rights reserved

Page 116 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Table 5-35 Relationship of output power of each carrier and the number of
carriers of a repeater
The number of carriers

The output power of each carrier (dBm)

+24.5

+21.5

10

+20.5

20

+17.5

VI. Repeater gain setting


The gain of the early repeaters must be set manually, but the latest gain of the
latest repeaters can be automatically set. For the repeaters whose gain is set
manually, the sum of the repeater gain and the protection margin must be equal
to or smaller than the repeater isolation; otherwise, the self-excitation of the
repeater will be caused. Here the repeater isolation indicates the isolation
between the donor antenna and the retransmission antenna of the repeater.
Generally, the protection margin ranges from 10 dB to 15 dB.

VII. Repeater adjacent cell planning


The coverage areas of a repeater may overlap other donor cells, so you must
configure the corresponding adjacent cell relationship for the repeater to ensure
normal handover. In addition, you must pay attention that the frequencies in the
coverage areas of the repeater and that in the donor cells cannot be the same
frequency and neighbor frequency.

VIII. Effect of delay processing against repeater planning


If only one repeater cannot fully cover an area (such as a narrow and long tunnel),
you can use several cascaded repeaters to provide the coverage. The selection
of the address and antenna for the repeater of each level is the same as that for a
single repeater.
However, the repeater will amplify the same frequency and it takes some time for
the repeater to process the signal, so there is a delay for each signal segment. If
the delay is greater than the time for the GSM system to identify the time window,
the intra-frequency interference will occur. Therefore, you must consider the
effect of the delay when adopting cascaded repeaters, because the delay will also
accelerate the time dispersion and shorten the coverage distance.
If adopting the optical repeater, you must consider that the transmission speed of
the signals in optical fibers is 2/3 that of in free space, namely, if the extension cell
technology is not used, the maximum transmission distance of the signals in
2006-01-05

All rights reserved

Page 117 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

optical fiber is 35 km multiplies 2/3 (about 23.3 km) due to the restriction on
transmission delay.
In addition, if one of three synchronous cells adopting the optical repeater, the TA
of two cells will be different due to the difference of transmission mode and rate.
In this case, the synchronous handover failure will occur. Therefore, you must
adopt the asynchronous handover to obtain the TA of a new cell, which works as
the handover target cell.
The delay processing varies with repeater types. Some take 2 to 3 s and some
takes 5 to 6s. In a GSM system, the delay of two signals cannot be greater than
16s. For the effect of repeater delay processing against time dispersion,
see Figure 5-54Figure 5-54.

Figure 5-54 Effect of repeater delay processing against time dispersion


In Figure 5-54Figure 5-54, the distance between point A and the repeater d is
2.1km. The delay for the mobile station at point A to receive the signals from the
repeater and the cell is as follows:
(2.1km + 2.1km)/c (light speed) + 3s = 14s + 3s = 17s > 16s.
In this case, the intra-frequency interference may be present. If the difference of
the levels of the two signals is equal to or lower than 12 dB, the conversation
quality will be affected.
The time dispersion will cause intra-frequency interference, and the time
dispersion is caused by the overlap of the signal source cell and the repeater
coverage area. Therefore, you must select the signal of the secondary cells in the
coverage areas of the repeater instead of the signals of the major service cell as
the source signal of the repeater. In this case, the time dispersion caused by
overlap can be avoided.

IX. Effect of background noise against repeater planning


Suppose that the maximum received noise level allowed by the base station is DN,
if the uplink background noise level of the repeater host is too great, the base
station channels will be congested when the noise level at the base station is
greater than DN. However, how to set the repeater without affecting the base
station? They are introduces as follows.
2006-01-05

All rights reserved

Page 118 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

If the following assumptions are present:


The transmitted signal strength of the base station is Tb.
The received signal strength of the base station is Rb.
The received downlink signal strength of the base station host is Dr.
The transmitted uplink signal strength of the base station host is Ut.
In this case, the path loss between the base station and the repeater is Tb-Dr, so
Rb = Ut (Tb-Dr). As a result, if the repeater does not affect the base station, Rb <
DN, so the following two inequities are present:
Ut ( Tb - Dr) < DN
Ut < Tb-Dr + DN
According to the previous analysis, the repeater does not affect the base station if
the uplink background noise level output by the repeater host is lower than
(Tb-Dr+DN). From this perspective of review, the background noise must be
particularly emphasized in repeater planning because it is easier to bring
interference than other types of base stations.

X. Specifications of wireless repeaters


For the specifications of some repeaters in commercial use, see Table 5-36Table
5-36.
Table 5-36 Specifications of parts of repeaters in commercial use
Source &

Type/band

Power (W)

Gain (dB)

Noise

3rd order

No.

Reference

/channel

Downlink

Downlink

Figure

intercept

channels

(mm)

(dB)

(dBm)

Downlink

Downlink

(Uplink)

(Uplink)

4.5 (4.5)

40,

47,

N/A

460x

95 (30, 50,

50,

54

80, 95)

(40, 47,

selective

(Uplink)

(Uplink)

of

Size

AFL
GSM

Band

900MHz

1,

5,

10,

20, 25

30, 50, 80,

550x
220

50, 54)

(10W)
GSM
1800MHz

Band

1,

5,

20,25

10,

30, 50, 80

<

(30, 50, 80)

(<4.5)

4.5

40,

47,

50,

54

N/A

(40, 47,
50, 54)

460x
550x
220
(10W)

2006-01-05

All rights reserved

Page 119 of 121

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

GSM

Channel

1,

5,

10,

20,25

Confideniality level

30, 50, 80

6.0 (6.0)

40,

47,

95, (30, 50,

50,

54

80, 95)

(40, 47,

1, 2, 4, 8

460x
550x
220

50, 54)

(10W)
GSM

Channel

1800MHz

1,

5,

10,

20,25

30, 50, 80

< 6.0

40,

47,

95, (30, 50,

(< 6.0)

50,

54

80, 95)

1, 2, 4, 8

460x
550x

(40, 47,

220

50, 54)

(10W)
Allgon
AR 1200

Band

50-90

52

N/A

440(W)

GSM

530(H)

900MHz

174(D)

AR 120

Channel

GSM

24 dBm

40-60

< 6.0

(20) dBm

285(H)

900MHz
AR 2100

230(W)

120(D)
Channel

33 dBm

50-90

1-4, 5-8

440(W)

GSM

(2-channel)

530(H)

1800MHz

30 dBm

174 or

(4-channel)

240(D)

Mikom
MR 340

Channel

32 dBm
(2-channel)

85-89
(2-channel)

28 dBm

82-86

(4-channel)

(4-channel)

6-8

2-channel

425(W)

modules

255(H)

per
cabinet
[2]

5.13 Conclusion
Network planning is the foundation of a mobile communication network,
especially the wireless parts in a mobile communication network costs great and
is of vital importance to network quality, so you must make a good planning at
earlier stage, which is helpful for network expansion and service update in the
future.

2006-01-05

All rights reserved

Page 120 of 121

110(D)

GSM Radio Network Planning and Optimization


Chapter 5 GSM Radio Network Planning

Confideniality level

Network planning requires engineers to analyze coverage, decide network layers,


and analyze traffic based on relative technologies and parameters, and finally
output the results of RF planning, including base station layout and scale.
RF planning, as well as the application of cell parameters, determines the cell
coverage. The cell coverage must be properly designed so that the mobile station
can always enjoy the best service at the best cells. In addition, the cell coverage
must be designed in a way conducive to network capacity expansion.
This chapter also introduces the solutions to dual-band network, indoor coverage,
tunnel coverage, and so on. Last, this chapter introduces the repeater application.

2006-01-05

All rights reserved

Page 121 of 121