XXX GSM-R Project

Radio Network HLD (High Level Design)

Issue 2.0

Date 10th July, 2013

Huawei Network Design Team

Huawei Technologies Co. Ltd

 Huawei Technologies Co., Ltd. provides customer with comprehensive technical support and service. For
any assistance, please contact our local office or company headquarters.

Huawei Technologies Co., Ltd.

Address: Huawei Industrial Base

Bantian, Longgang

Shenzhen 518129

People's Republic of China

Website: http://www.huawei.com

Email: support@huawei.com

Copyright © Huawei Technologies Co., Ltd. 2007. All rights reserved.

No part of this document may be reproduced or transmitted in any from or by any means without prior written
consent of Huawei Technologies Co., Ltd.

Trademarks and Permissions

And other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.

All other trademarks and trade names mentioned in this document are the property of their respective holders.

Notice

The information in this document is subject to change without notice. E very effort has been made in the
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute the warranty of any kind, express or implied.

Co mmercial in Confidence Page 2 of 33

Update History

Version Description Issue date Prepared by Approved by

1.0 Draft 2013-07-01 Cuiyong Onder

2.0 Modify leaky cable LB 2013-07-08 Cuiyong Qianjin

Co mmercial in Confidence Page 3 of 33

 XXCUS TOMER XX GS M-R Project Radi o Network High Level Design Report

Contents

Contents ............................................................................................................................................ 4
Figures ................................................................................................................................................ 6
Tables .................................................................................................................................................. 7
1 Introduction ................................................................................................................................... 8
1.1 Objectives .........................................................................................................................................................................8
1.2 Scope of Work .................................................................................................................................................................8
1.3 Requirement ....................................................................................................................................................................8
1.4 Assumptions .....................................................................................................................................................................9

2 Coverage Design........................................................................................................................ 10
2.1 Double Layer Design ..................................................................................................................................................10
2.2 RF planning assumptions ..........................................................................................................................................11
2.2.1 Downlink RxLev target .....................................................................................................................................11
2.2.2 Mobile station ......................................................................................................................................................12
2.2.3 Fading margin.....................................................................................................................................................12
2.2.4 Cell Overlap.........................................................................................................................................................12
2.3 BTS selection ................................................................................................................................................................14
2.3.1 DBS3900 Overview ..........................................................................................................................................15
2.3.2 Tunnel scenarios................................................................................................................................................16
2.4 Other design ..................................................................................................................................................................17
2.4.1 Splitter using strategy .......................................................................................................................................17
2.4.2 Antenna selection ..............................................................................................................................................18
2.4.3 Feeder and leaky cable selection ................................................................................................................18
2.5 Propagation model .......................................................................................................................................................19
2.5.1 Propagation model for open areas ..............................................................................................................19
2.5.2 Propagation model for tunnels ......................................................................................................................21

3 Result of Design......................................................................................................................... 25
3.1 BTS Amount ...................................................................................................................................................................25
3.2 Frequency Planning ....................................................................................................................................................25
3.3 Capacity Planning ........................................................................................................................................................26
3.3.1 Point-to-Point Traffic .........................................................................................................................................26

Co mmercial in Confidence Page 4 of 33

 XXCUS TOMER XX GS M-R Project Radi o Network High Level Design Report

3.3.2 Group Call Traffic ...............................................................................................................................................27

3.3.3 Broadcast Call Traffic .......................................................................................................................................28
3.3.4 GPRS Data Service ..........................................................................................................................................29
3.3.5 ETCS Data Service...........................................................................................................................................29
3.3.6 Total Traffic Volume...........................................................................................................................................29
3.4 Coverage Prediction....................................................................................................................................................29

4 Acronyms and Abbreviations ................................................................................................. 32

Customer Acceptance Sign-off Sheet ..................................................................................... 33

Co mmercial in Confidence Page 5 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

Figures

Figure 1-1: EIRENE Protocol Requirement of Coverage .................................................................................................9

Figure 2-1: The railway line .................................................................................................................................................10

Figure 2-2 Double layer design............................................................................................................................................11

Figure 2-3 Inter-cell handover time .....................................................................................................................................13

Figure 2-4 DBS3900 Overview ...........................................................................................................................................15

Figure 2-5 DBS3900 Mult i-RRU co-cell feature ..............................................................................................................16

Figure 2-6 DBS for coverage of tunnels .............................................................................................................................17

Figure 2-7 handover area design in the site .......................................................................................................................17

Figure 2-8 Typical site ...........................................................................................................................................................22

Figure 2-9 Coverage Scheme for the Tunnels ...................................................................................................................23

Figure 2-10 Coverage Model for long Tunnels .................................................................................................................23

Co mmercial in Confidence Page 6 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

Tables

Table 2-1:DL minimu m levels (d Bm) in norma l conditions ...........................................................................................12

Table 2-2 Antenna specifications .........................................................................................................................................18

Table 2-3 leaky cable specifications ....................................................................................................................................18

Table 2-4 XXX GSM-R Project Po wer Budget for BTS3012 ........................................................................................19

Table 2-5 Link Budget results for Tunnels using Yagi antenna ......................................................................................23

Table 3-1: BTS count .............................................................................................................................................................25

Table 3-2: Frequency groups ................................................................................................................................................26

Table 3-3 Erlang B table ........................................................................................................................................................29

Co mmercial in Confidence Page 7 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

1 Introduction

1.1 Objectives
According to the network equipment and experience, the High Level Design (HLD) is to design
an excellent RF network, which meets the requirement of network dimension, with high security
& availability, reasonable resource distribution, and good coverage and quality for GSM-R
services.

1.2 Scope of Work

The HLD report covers the design of RF network design for antenna, cell and BTS, which
doesn’t includes the SGSN, core network design such as MSC, MGW, HLR, and the BSS
network design.

The High Level Design includes

 Coverage Design assumptions

 BTS selection
 Other design
 Propagation model
 Result of design

1.3 Requirement
Customer intends to implement a complete ERTMS Level-2 system along an existing railway
section XXX and a GSM-R system on a commuter section XXX

Co mmercial in Confidence Page 8 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

The total section is XX – XX - XX - XX section that consists of approximately 51km of track

including one tunnel.

Radio planning was conducted based on double coverage (co-located) and high redundancy
level.

1.4 Assumptions
The HLD design is provided based on the following general assumptions/facts.

 GSM-R Radio Network Proposal for XXXproject.

 EIRENE Protocol Requirement of Coverage as the following:

Figure 1-1: EIRE NE Protoc ol Requirement of Coverage

Co mmercial in Confidence Page 9 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

2 Coverage Design

As required by customer, the scheme of co-site double-layer coverage network is proposed

for the XXX line.

XXX Railway

Figure 2-1: The railway line

2.1 Double Layer Design

This scheme consists in radio coverage redundancy in each geographic location. The

Co mmercial in Confidence Page 10 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

double coverage network therefore operates on two BTS layers named Layer A (master)
and Layer B (backup). Both layers are connected to one single BSC. The radio network
parameters are configured in each cell to force the mobile stations to automatically camp
on Layer A initially. The MSs immediately initiate switchover or reselection to the
corresponding backup BTS site of layer B when the master BTS site in layer A is broken
down.
The principle is explained in the figures below:

Figure 2-2 Double layer design

2.2 RF planning assumptions

2.2.1 Downlink RxLev target
Huawei considered following design specifications for the purpose of this project:

The coverage area probability shall be 95%.

Non-tunnel areas:

At height of 4 m from rails, coverage levels shall be measured via the antenna at the top of the
train and shall be at least -82 dBm at station sites and at least -85 dBm along the route. The
coverage area probability shall be 95%.

For handsets, the measured coverage levels shall be at least -92 dBm for outdoor areas
excepting shunting areas. It shall be -75 dBm for shunting areas. The coverage area probability
shall be 95%.

Tunnel: the planned coverage levels shall be equal to -77 dBm or higher. The coverage area
probability shall be 95%.

Co mmercial in Confidence Page 11 o f 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

Note: The actual Rx levels inside the train, at the cab radio input antenna connector, will be up to 1.5 dB less than
the values read at the antenna at the top of the train (4m above the running rail). This is due to the feeder loss
between the antenna and the cab radio. This value comes from the XXX technical specification, chapter 24.4

To summarise these requirements in a table:

DL Level at Rooftop antenna (for

DL level for
Normal conditions cab radio)
handsets

Stations -82 dBm -92 dBm

Shunting yards -82 dBm (Huawei Assumption) -75 dBm
Tunnels -77dBm -92 dBm
Open tracks -85 dBm -92 dBm

Table 2-1:DL minimum levels (dBm) in normal conditions

2.2.2 Mobile station

Unlike public GSM networks, there are different kinds of terminals in GSM-R communication
networks:

 2W OPH (operational purpose handhelds),

 2W GPH (general purpose handhelds),

 2W OPS (operational purpose for shunting areas)

 8W Cab Radios.

While indeed being a “mobile station” in the GSM sense of the word, the cab radio is a fixed
piece of equipment that is bolted onto the locomotive driver’s panel.

2.2.3 Fading margin

To take into account Rayleigh fading, a margin is factored into the link budget to ensure the
EIRENE target downlink level is reached with a probability of 95% at 100 meters intervals. This
requirement means that this probability must also be reached at the cell edge, where the signal
is weakest, which in turn translates into a margin on the average downlink level.

2.2.4 Cell Overlap

The distance of coverage overlap is determined according to the timing requirements of two
different GSM-R procedures:

 Cell reselection: procedure triggered by the mobile station in idle and group receive modes
to change cells while remaining able to seamlessly access the network or listen to an
ongoing voice group call

Co mmercial in Confidence Page 12 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

 Handover: procedure triggered and managed by the network to instruct the mobi le station
engaged in either a point-to-point voice or data call, or a voice group call as a talker

Cell Reselection Requirements:

To ensure timely reselection of subscribers in the trains moving in opposite directions, the
parameters CRO, Temporary_Offset, and Penalty_Time are usually set to 0. In the case of cell
reselection in a location area, the C2 of a neighbouring cell must be greater than C2 of the
serving cell for 5s. In the case of cell reselection in different location areas, the C2 of a
neighbouring cell must be greater than C2 plus CELL_RESELECT_HYSTERESIS of the
serving cell for 5s.

When the level of the MS in the overlapped coverage areas of two neighbouring cells is the
same, the timing begins. Cell reselection needs a further 5 seconds to be completed. So, as
this reasoning applies to both direction s, the overlapped coverage range must be equal to 10s
multiplied by the train speed.

Intercell Handover requirements

Rxlevel

Source Target

margin

P/N

Time
T1 T2 T3 T4

Figure 2-3 Inter-cell handover time

T1: delay of the reported level and the filter level. The green solid line indicates the receive
level reported by the MS, whereas the green dashed line indicates the filtered level obtained by
the BSC. The filter length affects the filtered level obtained by the BSC because of the filtering
algorithm. The filter length of traffic channels and neighbouring cell measurement of Huawei
BSC is the length of four measurement reports by default. Generally, the average of the lev el

Co mmercial in Confidence Page 13 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

contained in the four measurement reports is taken for filtering. Assume that signal fading is
linear. The dashed line indicates the filtered level. T1 = Filter length/2 = 4 x 480ms = 0.96s.

T2: time to reach the margin. The margin shown in Figure 3-10 is the sum of the margins of all
the handover algorithms. The typical value of T2 is 2 dB in the GSM-R system.

The typical value of T2 is 1.5s to 2s when the railway moves at 100 m/s.

T3: handover decision time. The rate at which handover decisions are taken is 8 SACCH
frames (1 SACCH frame = 480 ms for TCH or 470 ms for SDCCH) in common handover
algorithms and 4 in fast handover algorithms (these are the default values on Huawei BTS).
Thus, the total time is 1.92s (4 x 0.48s) or 3.84s (8 x 0.48s).

T4: handover execution time. The average delay on the Um interface from Handover
Command to Handover Complete is 350 ms in the high-speed railway scenario.

Total hand-over time = 0.96s + 1.5s + 1.92s + 0.35s = 4.73s

This value is close to the value of the delay for cell reselection (5 seconds).

As a conclusion, the range of the overlapped coverage area is assumed at design stage to be:

5 seconds x railway speed x 2.

The designed speed for the XXX line is 250km/h, according to the requirements.

Overlap distance: 250000*10/3600 = 695 meters.

This value has been rounded to 700 meters.

2.3 BTS selection

To cover the XXX line, Huawei has selected different types of BTS:

- DBS3900Indoor Macro BTS (BTS3012)

Co mmercial in Confidence Page 14 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

- Indoor Macro BTS (BTS3012)

The aim of this section is to provide more details on the use of DBS in special areas of the XXX
line.

2.3.1 DBS3900 Overview

The DBS3900 was developed by Huawei for multiple wireless technologies (GSM, UMTS, LTE).
The concept of distributed base station consists in physically separating the Radio Tx/Rx
functions from other functions such as clock, BSC interface, alarm management etc:

It consists of the baseband control unit (BBU) and several remote radio unit (RRU), remotely
connected to the BBU via optical fiber. The remote radio units RRU3004 e ach support up to 2
transceivers.

Figure 2-4 DBS3900 Overview

The engineering limits of this product are the following:

 The maximum distance between BBU and RRU is 40km.

 The BBU supports 6 optical connections towards the RRUs

 The RRUs can be chained together. The maximum number of RRUs in a chain is 3. So 3
RRUs can be chained together and the chain is connected via a single port to the BBU

 The maximum number of such chains is 6. If these chains are looped back, then the
maximum of loops is 3, since each loop occupies 2 ports.

Multi-RRU co-cell feature

Co mmercial in Confidence Page 15 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

With this feature, more than one RRU can be configured with the same frequencies and act as
a signal repeater. All the RRUs that share these frequencies belo ng to the same cell. There is
no handover between RRUs. This reduces cost and enhances performance.

Multi-RRU co-cell
1 Cell, Several Sites
f1(BCCH), f2 f1, f2 f1, f2 f1, f2 f1, f2 f1, f2

RRU RRU RRU RRU RRU RRU

1 cell CPRI

BBU

(typ.) 2-3 km in tunnels

(typ.) 6-8 km in open areas

Figure 2-5 DBS3900 Multi-RRU co-cell feature

With this feature the DBS3900 have many advantages compared with repeaters:

 Multi-site Co-Cell feature removes the need for handover inside tunnels.

 No need for separate O&M system

 Unlike repeaters, the distributed radiating units are synchronised with one another, which
removes problems of multipath thanks to mobile’s equaliser.

2.3.2 Tunnel scenarios

In line with best engineering practices, the key point in the tunnel coverage design is to avoid if
possible any handover occurring inside the tunnel.

Using the DBS and the multi-RRU cell-sharing feature, this can be achieved since we can
cover the tunnel using one or several RRUs, and define all these RRUs as belonging to one
single cell.

For the tunnels of the CUSTOMER XXX network, such a solution has been adopted.

Co mmercial in Confidence Page 16 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

L1 L2 L3

A B C

Figure 2-6 DBS for coverage of tunnels

2.4 Other design

2.4.1 Splitter using strategy
Unlike GSM three sections in one site, for GSM-R there are two sections in one site , In
traditional site configuration, two RRU connected to two antenna separately. Handover area at
two cell antenna side lobe，Signal attenuation fast, handover area is small, dissatisfy demand
for high-speed handover.

If two RRUs in the same cell scenario and covering two directions, there is no handover.

If we use one RRU, we use splitter to solve, the same cell through the splitter leads to two
antennas, each covering two directions, non handover.

Figure 2-7 handover area design in the site

Co mmercial in Confidence Page 17 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

2.4.2 Antenna selection

High-gain antennas are usually selected to cover high-speed railways so that the coverage of a
single site increases and the rate of handovers can be reduced.

For tunnels, we choose Yagi or log-periodic antennas, as well as leaky feeders.

The antennas that have been used in the design are as follows:

Frequency (MHz) H-BW (°) Gain Polarization Size (mm)

(dBi) L* W * D

806-960 33 21 X 2578×592×183 mm

806-960 65 10 V Yagi

Table 2-2 Antenna specifications

2.4.3 Feeder and leaky cable selection

 We propose the 7/8 feeder based on the maximum planned distance between the top of
BTS cabinet and antennas

 We propose the leaky cable use in the long tunnels as follows:

Frequency Cable type Impedance Velocity ratio Size Jacket

(MHz) (Ohms) (%) (mm)
L* W *
D

Radiated Mode Halogen Free Low Smoke

30-2000 50 +/- 2 88 1-5/8”
Cable Flame Retardant

Table 2-3 leaky cable specifications

Co mmercial in Confidence Page 18 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

2.5 Propagation model

2.5.1 Propagation model for open areas
Because of the huge propagational character diversity for differen t scenario, Huawei adopts the
following Propagation models:

 Okumura-Hata (Suburban) for urban and mountains scenarios

 Okumura-Hata (Rural) for rural scenarios

The main difference between the Suburban model and the Rural model is the Cm value. In urban
and mountain scenario, high buildings or mountains may block the signal, therefore more margin
should be factored in to account for the more complicated propagation environment than rural
environment.

Table 2-4 XXX GSM -R Project Power Budget for B TS 3012

The Link budget takes into account the assumptions detailed above and yields an estimate of
the cell radius (the radius is measured from the BTS antenna site to the cell edge)
The inter-site distance can be approximately calculated from this cell radius, i.e.:

Suburban: 2 x 3.56 km – 0.7km = 6.42 km

Rural: 2 x 5.66 km – 0.7km = 10.62 km.

Co mmercial in Confidence Page 19 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

In practice, the link budget does not consider various factors that affect the radio wave
propagation, such as terrain and clutter variations. These factors can be theoretically estimated
using RF planning tools in a second stage of the design.

Co mmercial in Confidence Page 20 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

2.5.2 Propagation model for tunnels

Tunnel radio coverage depends on different factors:

 The attenuation of radio wave propagation in a tunnel decreases with increase of signal
frequency. Radio Measurements have demonstrated that the attenuation is inversely
proportional with the frequency.

 Rectangular tunnels normally have less loss than semi-circular tunnels.

 Bends and hills inside the tunnel attenuate the signal. The attenuation in the tunnel due to
bends and hills can be compared with the attenuation caused by bending optical fiber. This
attenuation is inversely proportional with the square of the radius of the bend and
proportional with the used frequency.

 The penetration depth of material of the tunnel walls influences the signal attenuation. High
penetration depth leads to high absorption and high attenuation of the radio signal. The
penetration depth is inversely proportional to the square root of the signal frequency, the
permeability and conductivity of the wall material. Rough surfaces increase the attenuation,
but the impact of rough walls decreases with the increase of the frequency.

 High traffic and fill factor lead to more obstruction for the radio waves, which means higher
losses. The fill factor depends on the size of the tunnel (cross section area), the number of
trains inside the tunnel and the length of the trains. The cross section area is limited by the
maximum speed of the trains on the railway track.

 The siting of the antenna inside the tunnel or close to the tunnel entrance can be crucial for
the system performance. It is therefore important to place the antenna in a position which
provides line-of-sight for as long as possible in the required direction. Antenna can be
mounted on the ceiling of the tunnel or on top of a separate mast very close to the tunnel
entrance.

Co mmercial in Confidence Page 21 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

Figure 2-8 Typical site

Huawei selects the DBS3900 (Distributed Base Station) solution instead of repeater for special
scenario like tunnels.

GSM-R
Link Budget For RRU3004 for S2 configuration
DownLink UpLink
BTS max. transmit power (dBm) 44.80 MS max. transmit power (dBm) 39.00
BTS combiner loss (dB) 3.50 BTS combiner loss (dB) 3.50
7/8 Feeder Length 0.00 7/8 Feeder Length 0.00
1/2 jumper length (m) 5.00 1/2 jumper length (m) 3.00
feeder connector loss (dB) 0.80 feeder connector loss (dB) 0.80
Antenna gain (dBi) 10.00 BTS antenna gain (dBi) 10.00
Min Designed DL Rxlevel(dB) -77.00 Min Designed UL Rxlevel(dB) -113.00
expected shadow fading margin expected shadow fading margin
(dB)
10.50 (dB)
10.50

System Loss for Cab Radio 0.00 System Loss for Cab Radio 3.00
allowed UL Propagation loss
allowed DL Propagation loss (dB) 116.42 143.85
(dB)
Maximum Allowable Path Loss
(dB) 116.42

Co mmercial in Confidence Page 22 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

Table 2-5 Link Budget results for Tunnels using Yagi antenna

The propagation model of the tunnel is Lm（max path loss allowed）= 20log (f) +30log
(L)-28dB+F (fill loss)
Cell radius in the tunnels = 170m (with f=900MHz and F=12dB)
There are therefore 2 scenarios based on the length of the tunnel:

 Scenario n°1: if the length of the tunnel is less than 170m: then the coverage solution
consists of DBS3900 plus Yagi antennas radiating into the tunnel.

Figure 2-9 Coverage Scheme for the Tunnels

 Scenario n°2: if the length of tunnel is more than 170m: coverage solution is DBS3900
to cover the tunnels with RRU and Leaky Cable. Yagi antennas in the entrances and exits
to cover the area out of the tunnel.

Figure 2-10 Coverage Model for long Tunnels

In this case, the number of RRUs is designed based on the length of the tunnel.

Note: the tunnels are covered by several RRUs, but with one single cell, to ensure there is no
handover inside. This is the multi-RRU co-cell feature.

Co mmercial in Confidence Page 23 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

The link budget of leaky cable in tunnel is calculated as follows.

Min. Signal Reception Strength = RRU TOC Power - Power Splitter Loss( or combiner loss ) -
7/8 Feeder Loss - 1/2 Jumper Loss - Connector Loss - Leaky Cable Coupling loss - Leaky
Cable Loss Constant * L / 100 - Fast Fading Margin – Cable Loss between Cab Radio and
Antenna on top of the train

TOC (top of cabinet, measured at antenna connector) output power is 30W per TRX of an
RRU.

The length of the jumper is 2 metres.

The minimum signal level request is -92dBm for handset, -77dBm for cab radio.

According to the formula above, we can get

-77dBm = 44.8dBm - 3.5dB – 3 - 0.5dB - 1dB - 2.8 * L (m) / 100m - 72dB - 3dB

So the longest distance of the coverage length for 900M leaky cable is L = 1.39km.

According to the solution mentioned above, we can estimate the RRU quantity for tunnel
coverage, depending on the tunnel length.

Co mmercial in Confidence Page 24 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

3 Result of Design

3.1 BTS Amount

The breakdown per BTS configuration is as follows:

BTS configuration BTS qua ntity

BTS3012 S2 20
DBS3900 S2/2/2/2 1
Total 21

Table 3-1: BTS count

Along the line 20 BTS and 1 DBS are deployed. In order to guarantee the coverage
requirements, high overlapped coverage scheme is adopted. Since DBS coverage area is
higher when compared to a BTS, Huawei has provided double BBU (one active, second
backup) for each DBS. This way, if active BBU fails, RRUs connected to active BBU will switch
to backup BBU and continue operation.

3.2 Frequency Planning

The available frequency resource in the CUSTOMER project is 876-880 MHz for uplink and
921-925 MHz for downlink. This corresponds to the ARFCN channel numbers 955 to 973
included (i.e. 19 channels).

To minimise interference, the frequency plan should meet the following requirements:

 C/I levels:

o Co-channel interference protection ratio: C/I ≥ 12 dB

o Adjacent-frequency interference protection ratio: C/I ≥ 6 dB

Co mmercial in Confidence Page 25 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

o Interference protection ratio in the case of 400 kHz carrier frequency offset: C/I ≥ -38 dB

 Spacing:

o 400kHz minimum spacing space inside site

o 600kHz space for BCCH of adjacent neighbour sites

o 400kHz space for TCH of adjacent neighbour sites

Considering the primary site model O2 along the railway line, 16 out of the 19 available
ARFCNs can be divided into 8 groups of 2 frequencies.

These groups of 2 frequencies (1st frequency: BCCH, 2nd frequency: TCH) will be used
along the railway line with a periodic repetition. The remaining 3 channels (ARFCN n°959,
961, 960) are used as supplements in railway station sites if necessary.

So, along the line there will be the following pattern:

Cell A Cell B Cell C Cell D Cell E Cell F Cell G Cell H

BCCH 955 963 967 971 956 962 966 970

TCH 957 965 969 973 958 964 968 972

Table 3-2: Frequency groups

3.3 Capacity Planning

GSM-R capacity planning is a complex task. As the GSM-R service develops, the requirement
on network capacity changes. In GSM-R capacity planning, you must consider the multiple
factors, including point-to-point call, group call, broadcast call, data service, and subscriber
number.

3.3.1 Point-to-Point Traffic

In the GSM-R system, the two parameters, P-Traffic and T-Traffic, determines the amount of
the point-to-point traffic.

P-Traffic
P-Traffic stands for the traffic generated by the daily maintenance staff in the train station.
Currently, it is estimated that there are 10 staff employees in a train station, and the average
traffic of each one ranges from 0.01 to 0.02 Erl, so it is 0.015 Erl in average. Therefore, the
P-Traffic of each base station is 0.15 Erl.

Co mmercial in Confidence Page 26 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

T-Traffic
T-Traffic is the traffic generated by the subscribers when trains pass and stop at the train station.
Generally, 2 trains will pass by a station at simultaneously. If 30 calls are generated when the
train passed by the station, and the conversation lasts for 60 seconds for each call, the T-Traffic
can be expressed by the following equation:

T-Traffic = the number of trains * each call duration (s)/call times per hour (s) = 2*60/3600*30 =
1Erl.

Generally, only 1 train will stop at a base station at a time. If 10 calls are generated by
subscribers when the train stops at the station, and each conversation lasts for 60 seconds, the
T-Traffic can be expressed by the following equation:

T-Traffic = the number of trains * each call duration (H)/call times per hour (s) = 1*60/360*10 =
0.167Erl.

Therefore, the point-to-point traffic is 1.32 Erl in total. That is, point-to-point traffic = P-Traffic +
T-Traffic = (0.15 + 1 + 0.167) Erl = 1,32 Erl.

3.3.2 Group Call Traffic

To estimate group call traffic, you must consider the group calls generated by the group call
subscribers in the trains passing by and stopping at the train station. In addition, you must also
consider the groups call generated by the staff working in the train station.

If a dispatcher starts a group call, the group call needs 3 channels under 1.5 mode. If the group
call is not started by a dispatcher, it needs 2 channels. If the probability for the dispatcher and
the non-dispatcher to start the group call is 50% for each, you must consider the traffic metric
(3*50% + 2*50% = 2.5).

Note: Under 1.5 mode, BSC allocates a dedicated channel to the speaker and the speaker does not seize the uplink
resource of the group call channel. When the conversation finishes, the speaker releases the dedicated channel,
thus improving the conversation quality.
Generally, 2 trains will pass by a station at simultaneously. If 4 group calls are generated when
the train passed by the station, and the conversation lasts for 120 seconds for each call, the
T-Traffic can be expressed by the following equation:

T-Traffic = the number of trains * each call duration (s)/call times per hour (s)*traffic metric =
2*120/3600*4*2.5 = 0.67Erl.

Generally, only 1 train will stop at a base station at a time. If 2 calls are generated by
subscribers when the train stops at the station, and each conversation lasts for 120 seconds,
the T-Traffic can be expressed by the following equation:

T-Traffic = the number of trains * each call duration (s)/call times per hour (s)*traffic metric =
1*120/3600*2*2.5 = 0.17Erl.

If there are 5 group call subscribers in a train station, and each subscriber starts 1 group call,
and the conversation lasts for 120 seconds, the T-Traffic can be expressed by the following
equation:

T-Traffic = the number of group cal subscribers* each call duration (s)/call times per hour (s)

Co mmercial in Confidence Page 27 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

*traffic metric = 5*120/3600*1*2.5 = 0.41 Erl.

Therefore, the group call traffic is 1.25 Erl in total. That is, group call traffic = (0,67 + 0.17 + 0.41)
Erl = 1,25 Erl.

3.3.3 Broadcast Call Traffic

Similar to the group call traffic, broadcast call traffic also requires traffic metric. If a dispatcher
starts a broadcast call, the call needs 1 channel under 1.5 mode. If the group call is not started
by a dispatcher, it needs 2 channels. If the probability for the dispatcher and the non-dispatcher
to start the group call is 50% for each, you must consider the traffic metric (1*50% + 2*50% =
1.5).

Generally, 2 trains will pass by a station at simultaneously. If 3 broadcast calls are generated
when the train passed by the station, and the conversation lasts for 30 seconds for each call,
the T-Traffic can be expressed by the following equation:

T-Traffic = the number of trains * each call duration (s)/call times per hour (s)*traffic metric =
2*30/3600*3*1.5 = 0.07 Erl.

Generally, only 1 train will stop at a base station at a time. If 1 broadcast call is generated by
subscribers when the train stops at the station, and the conversation lasts for 30 seconds, the
T-Traffic can be expressed by the following equation:

T-Traffic = the number of trains * each call duration (s)/call times per hour (s)*traffic metric =
1*30/3600*1*1.5 = 0.011 Erl.

If there are 5 group call subscribers in a train station, and each subscriber starts 1 broadcast
call, and each conversation lasts for 30 seconds, the T-Traffic can be expressed by the
following equation:

T-Traffic = the number of group cal subscribers* each call duration (s)/call times per hour (s)
*traffic metric = 5*30/3600*1*1.5 = 0.06 Erl.

Therefore, the broadcast call traffic is 0.132 Erl in total. That is, group call traffic = (0.07 + 0.011
+ 0.06) Erl = 0.132 Erl.

In the GSM-R system, non-data services include point-to-point call, group call, and broadcast
call, so the total traffic of the non-data services are 2,70 Erl. (1,32 + 1.25 + 0.132 = 2,702). If the
call loss rate is 0.5%, 8 channels are needed for the non-data services according to Erlang
table.

Co mmercial in Confidence Page 28 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

Table 3-3 Erlang B table

3.3.4 GPRS Data Service

According to the requirement of customer, 2 channels will be reserved for GPRS service.

3.3.5 ETCS Data Service

The ETCS data service is a high priority call for the GSM-R network and requires a dedicated
channel allocation.

We assumed that there will be 2 trains in one cell at a specific time with 2 cab radio located in
each train. So in that case 4 dedicated channels will be enough for the ETCS data service.

3.3.6 Total Traffic Volume

Based on the assumptions and calculations above, 14 (8+2+4=14) channels are needed for the
point-to-point call, group call, broadcast call, GPRS and ETCS Data.

In addition, another 2 signaling channels must be considered, so a site needs 16 channels in

the GSM-R system in normal conditions. In this case, it will be enough to configure 2 TRXs for
each cell.

3.4 Coverage Prediction

Coverage prediction is performed with Huawei’s RF planning tool GENEX U-Net.

To establish the coverage prediction, the RF Planning assumptions and site distance reference from
the link budget is used as initial setting inputs. Together with the digital map which provides rich and
precise geographical information, the RF planning tool processes the inputs to calculate the
propagation loss. The coverage prediction is then represented in graphical format.

The coverage results provided in the subsequent sections have considered the factors above.

Co mmercial in Confidence Page 29 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

 Cab Radio

Figure 3-1 Coverage prediction of cab radio

Co mmercial in Confidence Page 30 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

 Handset

Figure 3-2 Coverage prediction of handset

Co mmercial in Confidence Page 31 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

4 Acronyms and
Abbreviations

Abbreviation Full Name

BBU Baseband Unit
BSC Base Station Controller
BSS Base Station Sub-system
BTS Base Transceiver Station
DBS Distributed Base Station
DTRU Double-Transceiver Unit
EIRENE European Integrated Railway Radio Enhanced Network
GSM-R GSM for Railways
RRU Remote Radio Unit
TRX Transceiver
QoS Quality of Service
SGSN Serving ORRISSARS Support Node
GSM Universal Mobile Telecommunications System

Co mmercial in Confidence Page 32 of 33

 CUSTOMER XXX GS M-R Project Radi o Network High Level Design Report

Customer Acceptance Sign-off Sheet

It is agreed that, from XXXXX to XXXXX , the GSM-R Radio

Network Design HLD Report of XXX GSM-R Project is accepted by
both CUSTOMER experts ,XXXX experts and Huawei experts.

CUSTOMER Representative CUSTOMER Representative Huawei Representative

Name: Name: Name:

Signature: Signature: Signature:

Date: Date: Date:

Co mmercial in Confidence Page 33 of 33