You are on page 1of 37

SENSITIVE TO T-MOBILE

LINK BUDGET FOR MACROCELLULAR UMTS FDD


NETWORK

SCOPE

This document provides a link budget for the design of T-Mobile UK’s
UMTS FDD macrocell radio network. A traditional uplink calculates the
maximum path loss for different uplink bearers. The uplink values are
then used to estimate the supportable bearer data rates on the
downlink.

PURPOSE

This document should be used in cell planning in order to do high level


cellular network design based upon cell range, or for specifying signal
level thresholds in detailed cell planning. It can also be used to
determine the effects of different base station configurations.

Please E-Mail the Corporate DOCUMENT REFERENCE: T-Mobile/CDOC/


Library for further information OWNED BY : Peter Stevens
ISSUE: 0.1
about this document DATE : 11 May, 2002

T-MOBILE CORPORATE LIBRARY


THIS DOCUMENT WHEN PRINTED
WILL BE
DEEMED AS UNCONTROLLED
Printed from: EDMS/WEBSITE
Turning Information Into Knowledge
© 05/08/11 T-Mobile FOR CORPORATE LIBRARY USE ONLY
© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

DOCUMENT TABLE OF CONTENTS


SCOPE
1
PURPOSE
1

1. INTRODUCTION
4
1.1 OVERVIEW
4
1.2 MAXIMUM PATH LOSS DEFINITIONS
4

2. UPLINK LINK BUDGET PARAMETERS


6
2.1 UE MAXIMUM OUTPUT POWER
6
2.2 UE ANTENNA GAIN
6
2.3 NODE B Eb/No
6
2.3 NODE B NOISE FIGURE
10
2.4 INFORMATION RATE
10
2.5 UPLINK POWER CONTROL HEADROOM
11
2.6 BASE STATION RX LOSSES
11
2.7 BASE STATION ANTENNA GAIN
12
2.8 UPLINK SHO COMBINING GAIN
12
2.9 SHO SLOW FADING MARGIN REDUCTION
13
2.10 INTERFERENCE MARGIN
14
2.11 BODY LOSS
15

T–Mobile/CDOC/ Page 2 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

2.12 SLANT LOSS


15

3. DOWNLINK PARAMETERS
18
3.1 NODE B MAXIMUM CARRIER POWER
18
3.2 CARRIER LOADING
18
3.3 MAXIMUM FRACTIONAL POWER FOR PACKET USER
18
3.4 BASE STATION TX LOSSES
19
3.5 DOWNLINK SLANT LOSS
20
3.6 HIGHER FREQUENCY ADDITIONAL PATH LOSS
20
3.7 UE NOISE FIGURE
20
3.8 NON-ORTHOGONALITY
21
3.9 OTHER TO OWN CELL CARRIER POWER RATIO
22
3.10 UE DOWNLINK EB/NO
22
3.4 DOWNLINK POWER CONTROL HEADROOM
24
3.12 DOWNLINK SHO COMBINING GAIN
25

LINK BUDGET RESULTS


26
4.1 UPLINK LINK BUDGET
26
4.2 MAXIMUM DOWNLINK PACKET BEARER RATE AT UPLINK LIMIT
27

T–Mobile/CDOC/ Page 3 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

APPENDIX A - EB/NO DEFINITIONS AND RELATED FACTORS POWER-


CONTROL –TX-POWER-INCREASE AND POWER-CONTROL-HEADROOM
28
A.1 INTRODUCTION
28
A.2 RECEIVED EB/NO
28
A.3 TRANSMITTED EB/NO
28
A.4 POWER CONTROL TX POWER INCREASE
29
A.5 POWER CONTROL HEADROOM
29
A.6 SIMULATION RESULTS
30

DOCUMENT HISTORY
31
DOCUMENT HISTORY

ATTACHMENTS.
List any attachments to this document here.

REFERENCES.

1. ‘Typical dimensioning parameters’, AP8 from TMO – Nokia Radio Performance


Workshop, 18/06/01.
2. ‘CEM Alpha performance vs. realistic profiles’, Jules Vidal (Nortel), presentation to
TMO-Nortel Radio Performance Workshop
3. ‘max.mobil. Siemens Workshop 2001-12-20’
4. ‘Node B system description’, Nokia contract document agreed 2 October.
5. ‘WCDMA Masthead Amplifier(MHA) Product Description’, Nokia User Manual ,
DN0114715, issue 1-0 en.
6. ‘RAN system description’ Nortel contract document, 16/8/01.
7. Presentation to TMO-Nokia link budget meeting, UK, 12/2/02.
8. Minutes (with action response LB1) of TMO-Nokia link budget meeting, UK, 12/2/02
9. ‘WCDMA link performance indicators, simulator principle and and examples’ Kari
Sipila (Nokia), NET/IMN/WNP/NSR, v 0.1.0.
10. ‘WCDMA link budget’, Jussi Reunanen (Nokia), 11/12/2000.
11. ‘WCDMA for UMTS’, Holma & Toskala, Wiley, 2000.
12. ‘Results on the peak to average power effect on coverage’, Ericsson, ETSI SMG2
UMTS-L1 490/98, Oct. 1998.
13. UMTS uplink link budget’, Juan Pedro Benitez, Nortel-TMO Radio
Planning/Performance workshop, 29/11/01

T–Mobile/CDOC/ Page 4 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

14. ‘Link budget for a personal communication network’,


Mo2o,/CDOC/P2.2/KNG/REPT/3253, Ken Ng, Issue 1.0, 6/9/95.
15. ‘Comparison of the performance between vertically polarised and cross polar
antennas’, Brian Williams, One2one/CDOC/6032, Issue 2.0, 21 September 2000.
16. ‘Cross polarised antenna results’, Wolfgang Stoermer, T-Mobil.
17. ‘WCDMA radio access network RF dimensioning guidelines’, EEM/TG/N-99:0042,
8/12/99.
18. ‘Power Divider Unit (WPD) description’, DN015603, DRAFT 1, Nokia.
19. ‘Default power settings for common DL channels in WCDMA RAN1- 1.5’, Kari Sipila,
Nokia presentation to TMO radio performance workshop, 18/06/2001
20. ‘Nokia UltraSite WCDMA Antenna System Product Overview’, DN00309809.
21. ETSI TS GSM 03.30.
22. ‘Nokia WCDMA FDD BTS RF Performance’, Mikko Siira, 15/2/01.
23. ‘RTT Revision – Performance results’, ETSI SMG2 351/98.

1. INTRODUCTION

1.1 OVERVIEW

The limited power at the terminal restricts uplink bearer rates to be


lower than those possible on the downlink. On the downlink, PA power
is shared, and by allocating user a large fraction of the total carrier
power it is possible to support high data rates even at the cell edge.
This sharing of power makes the downlink link budget more elastic
than the uplink.

It is therefore proposed to base cell planning for macrocells on uplink


bearer requirements. The uplink part of the link budget calculates the
maximum path loss supportable between base station and terminal, for
a range of different bearer rates. The parameters associated with this
uplink calculation are discussed in detail in section 2.

However, it is nevertheless important to understand the networks’


downlink capability. The path losses provided by the uplink link budget
are used to estimate a supportable bearer rate on the downlink at the
cell edge as defined by the uplink.

Section 1.2 discusses the maximum path loss definitions including the
output of the uplink link budget.

Section 2 proposes parameters values to be used in the uplink link


budget.

Section 3 proposes parameter values to be used in the downlink section


of the link budget.

Section 4 provides link budget tables taken from an Excel spreadsheet.

T–Mobile/CDOC/ Page 5 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Appendix A describes in more detail various parameter associated with


Eb/No.

It should be noted that if the above approach is also utilised for mini-
and micro- base stations with limited downlink power, it will imply an
inconsistent level of service in different types of cell. The link budget.

1.2 MAXIMUM PATH LOSS DEFINITIONS

It is important to define clearly the result of the uplink link budget so


that its application is consistent with its calculation.

The intent of this link budget is to include technology and bearer-


dependent factors within the calculation, but exclude more commercial
factors relating to probability of coverage and in-building penetration.
The latter effects may vary form network to network, whereas the
former should stay the same.

The maximum path loss (less body/slant loss) output of the link budget
is therefore defined as the maximum propagation loss in the uplink
band between ‘closest’ base station and terminal on boresight, as
measured by placing 0 dBi vertical antennas at the locations of the
base station antenna and terminal.

With this definition the body loss and slant loss do not form part of the
maximum propagation loss, and therefore it is important that factors
for these are part of the link budget calculation.

For the calculation of a cell range or an outdoor signal level from the
maximum path loss one must reduce the maximum path loss by the
building penetration loss and total shadow fading margin to calculate a
required outdoor mean path loss/signal level. It is assumed that the
propagation loss model does not include a slant loss factor.

Note that macro-diversity or soft handover combining gains are added


within the link budget calculation as technology factors, on the
assumption that in most cell edge situation there is the possibility of
several cells providing service. Any shadow fading margin applied in
calculating a cell range or outdoor signal level threshold should
therefore assume a single serving cell.

T–Mobile/CDOC/ Page 6 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Maximum path loss including body/slant loss

Maximum path loss less body/slant loss

Outdoor path loss

-Body loss -Penetration loss -Antenna gain


-Slant loss -Slow fading/penetration -RX/TX losses
margins (single cell) (jumpers, etc)

Figure 1.2.1 - Illustration of link budget path loss terms

2. UPLINK LINK BUDGET PARAMETERS

2.1 UE MAXIMUM OUTPUT POWER

Four UE categories are defined in the 3GPP Standard in TS 25.101. The


required output power of the various terminal types is given in the
following table:

UE Nominal Tolerance
category Maximum [dB]
Output
Power
[dBm]
1 33 +1/-3

T–Mobile/CDOC/ Page 7 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

2 27 +1/-3
3 24 +1/-3
4 21 ± 2
Table 2.1.1 : Maximum output power of various UE categories

The RFI responses from terminal manufacturers suggests that terminals


with maximum output power of both 21 dBm and 24 dBm are under
development. The majority are planning to provide the 21 dBm class.

It has been noticed that some manufacturers of 2G phones take


advantage of the tolerance range to design phones with a lower output
power than the nominal output level. However, it is not proposed to
plan for a average shortfall in UE output power from the nominal value
until more measurements are available.

A UE maximum output power of 21 dBm is recommended.

2.2 UE ANTENNA GAIN

There is a possibility that some data terminals will have directional


antennas with positive antenna gain. However, this cannot be relied on
for most terminals, and hence no antenna gain is assumed.

A UE antenna gain of 0 dBi is recommended.

2.3 NODE B Eb/No

Eb/No is a receiver performance metric given by the ratio of received


energy per bit to noise density at the point which receiver is just able to
decode the received signal with the required quality (typically FER).
Eb/No depends on the service, the velocity of the UE, and the degree of
multipath.

The node B Eb/No figures given here assume the use of diversity in the
node B, with Eb defined as the energy per bit per receiver branch.

There are no measurement results available at this time, and therefore


Eb/No values are derived from supplier simulation results and
specification values. It is assume that fast power control is being used.
The power control headroom factor is used to compensate for the fact
that UE at cell edge will have a capped maximum power.

A series of TMO workshops was held with TMO’s preferred suppliers in


order to obtain performance data for UMTS network design. The

T–Mobile/CDOC/ Page 8 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

following tables summarise the uplink Eb/No values for pedestrian @


3km/h and vehicular @120km/h mobiles for different bearers [1,2,3].

Bearer Data Nokia Nortel Siemen


Service Rate s
[kbps]
AMR Voice 12.2 4.0 6.3 3.3
Circuit 64 2.0 4.8 2.2
Circuit 128 1.5 4.4 1.8
Circuit 384 - 4.2 0.6
Packet 64 2.0 3.2 1.4
Packet 128 1.5 2.8 1.1
Packet 384 1.0 2.6 0.2
Table 2.3.1: Supplier Uplink (NodeB) Eb/N0 Values [dB] for 3 km/h users

Bearer Data Nokia Nortel Siemen


Service Rate s
[kbps]
AMR Voice 12.2 5.0 6.4 6.1
Circuit 64 - 4.9 3.8
Circuit 128 - 4.4 2.9
Circuit 384 - 3.9 2.7
Packet 64 3.3 2.9 3.6
Packet 128 3.0 2.6 2.9
Packet 384 2.0 2.3 2.3
Table 2.3.2: Supplier Uplink (NodeB) Eb/N0 Values [dB] for 120 km/h users

Assumption
s:
Nokia Voice 20 ms Interl., CS data 40 ms Interl., BLER 1% P-data 10 ms Interl. BLER 10%.
Nortel QoS: PSdata Global RBLER 10^-1, Voice BER 10-3, CS BER 10-6 , 1 dB implementation margin
for HW/SW implementation,
Siemens QoS: Voice Coded BER= 10^-3 for Voice, BER= 10^-6 for LCD, BLER=10%, Voice and CS
Convolutional Coding, PO Turbo Coding, no implementation margin

Due to the small differences in the Eb/No values for 128 kbps and
144 kbps, it is assumed that values for 128 kbps can be substituted
with values for 144 kbps and vice versa.

The differences between the supplier’s figure can be explained by


differences in the simulation assumptions. For instance Nortel, unlike
Nokia, includes the overhead of the 3.4 kbps associated signalling radio
bearer (SRB), and includes a 1 dB implementation margin.

One further source of Eb/No performance figures is the 3GPP


specifications. The table below shows the uplink Eb/No performance
required by 3GPP 25.104 for the 3 propagation cases for which
performance results are provided.

T–Mobile/CDOC/ Page 9 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Bearer Data Case 1 Case 2 Case 3


Service Rate
[kbps]
AMR Voice 12.2 11.9 9 7.2
Circuit 64 9.2 6.4 3.8
(BLER=1%)
Circuit 144 8.4 5.6 2.7
(BLER=1%) (128)
Circuit 384 8.8 6.1 3.1
(BLER=1%)
Packet 64 6.2 4.3 3.4
(BLER=10%)
Packet 144 5.4 3.7 2.8
(BLER=10%) (128)
Packet 384 5.8 4.1 3.2
(BLER=10%)
Table 2.3.3: 3GPP Uplink (NodeB) Eb/N0 Values [dB]

The multipath propagation model cases used by 3GPP for the above are
summarised in the table below. Case 1 has one dominant Rayleigh
faded component. Case 2 has 3 equal power paths with very large
delay spread. Case 3 has the most typical urban/suburban multipath
profile, but unfortunately the results are only applicable for high speed.

Case 1, speed 3km/h Case 2, speed 3 km/h Case 3, 120 km/h


Relative Delay Average Power Relative Delay Average Power [dB] Relative Average
[ns] [dB] [ns] Delay [ns] Power [dB]
0 0 0 0 0 0
976 -10 976 0 260 -3
20000 0 521 -6
781 -9
Table 2.3.4: 3GPP propagation model cases

It can be seen that the 3GPP Eb/No values are significantly higher than
the values provided by the vendors, particularly for case 1 and case 2.
One reason for that is that the 3GPP values are based on simulations
without power control. To fairly compare them with the vendor values,
the power control headroom should be subtracted in the 3km/h
multipath case 1 and case 2 figure.

Most of the rest of the difference can be attributed to the suppliers


ensuring a large implementation margin is included in the specification
for performance commitments. With the exception of Nortel, supplier
have not commited to their own recommended values.

The available sources assume a variety of quality measures for circuit


switched data. It is not considered worthwhile to design to a very
stringent quality requirement for circuit switched data at cell edge, as
the marketing interest in the circuit switched services is not very great.
Therefore we take a BLER of 1% as the cell edge design target for link

T–Mobile/CDOC/ Page 10 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

budget purposes. Closer to the centre of the cell this not prevent a
more stringent target being applied to power control.

To obtain TMO recommended Eb/No figures the vendor and 3GPP


figures have been combined by taking a weighted mean of
compensated figures. The compensations applied were as follows.

1. To 3GPP, Nokia [and Siemens] figures, overheads for supporting


the SRB were added as shown in the table below, derived from the
ratio between the SRB and the bearer information rate.

Bearer Data Rate SRB


Service [kbps] overhead
[dB]
AMR Voice 12.2 1.1
Circuit 64 0.2
Circuit 144 (128) 0.1
Circuit 384 0
Packet 64 0.2
Packet 144 (128) 0.1
Packet 384 0
Table 2.3.5: SRB overhead for different bearers

2. To the Nokia and Siemens figures a 1 dB implementation margin


was added.

3. To the 3GPP figures, the power control headroom as given in table


2.2.1 was subtracted.

4. To apply the most realistic 3GPP case 3 figures to the 3km/h


value , a reduction of 1.1 dB was applied to account for improved
power control performance at reduced speed. This figure was
obtained by averaging the difference between the supplier’s Eb/No
figures for the two speeds selected.

The weightings applied for the pedestrian @ 3km/h case were as


follows:

Source Weigh
ting

Nokia 20%
Nortel 20%
Siemens 20%
3GPP Case 3 20%
3GPP Case 1 10%
3GPP Case 2 10%
Table 2.3.6: Weighting applied for calculation of uplink 3km/h Eb/No

After performing the weighted mean calculation the


recommended value for uplink Eb/No are as follows:

Bearer Data 3 km/h 120


Service Rate km/h

T–Mobile/CDOC/ Page 11 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

[kbps]
AMR Voice 12.2 6.9 7.7
Circuit 64 4.1 4.5
Circuit 128/14 3.3 3.6
4
Circuit 384 3.1 3.4
Packet 64 3.2 3.9
Packet 128/14 2.6 3.3
4
Packet 384 2.4 3.0
Table 2.3.7: Recommended Uplink (NodeB) Eb/N0 Values [dB]

2.3 NODE B NOISE FIGURE

The table below shows receiver noise performance for the different
vendors.

Noise Nokia Nortel Sieme


Figure [dB] ns

NodeB (w/o 31 2.22 5


TMA)
NodeB (with 2.53 2.04 3
TMA)
Table 2.1.1

The overall Node B noise figure can generally be reduced by the use of
TMA. With a TMA, the reference point for system noise figure and hence
receiver sensitivity is moved from the top of the node B cabinet to the
input to the TMA.

In the case of the Nortel figures, the guaranteed BTS noise figure is
similar to the guaranteed MHA noise figure, so a slight degradation is
seen. However, the MHA typical noise figure is only 1.6 dB, so some
further improvement in the TMA figure might be expected.

The recommended node B noise figure is dependent on the use


of TMAs as shown in the following table.

Node B Noise
Figure [dB]
NodeB (w/o TMA) 3.5
NodeB (with 2.5
TMA)
1 Calculated from static sensitivity and Eb/No data provided in Node B system
description [4].
2 Contract features and requirements, see annex, feature number 23470.
3 Calculated using Friss’ equaltion with TMA NF of 2 dB, TMA gain of 12 dB [5], and cable
loss of 3 dB.
4 Calculated using Friss’ equation with TMA typical NF of 1.6 dB, TMA gain of 12 dB [6],
and cable loss of 3 dB. TMA guaranteed noise figure of 2db equates to 2.4 dB system
noise figure

T–Mobile/CDOC/ Page 12 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Table 2.3.1 Recommended values for node B noise figure

2.4 INFORMATION RATE

This is the bearer information rate provided by the RLC layer to the
upper layers.

2.5 UPLINK POWER CONTROL HEADROOM

This component of the link budget arises because the Eb/No values
used within the link budget are receiver Eb/No values which assume
power control can counter the effects of Rayleigh fading. In reality at
the cell edge, the transmitted power is capped by the maximum output
power.

Further description of this factor can be found in appendix A.

The power control headroom is dependent upon the type of bearer


(notably required BLER/quality), the speed of the UE and the amount of
multipath. The values recommended below are derived from published
Ericsson results for Pedestrian B (Appendix A, section A6).

Nokia apply a power control headroom (also called fast fast margin) of
3 dB for 3 km/h in their link budget [7]. This is derived from simulation
results for Vehicular A [8], but the BLER is unknown. Nortel do not
inlude such a parameter.

A uplink power control headroom is recommended as follows:

Power Control Headroom


Terminal Speed Circuit Packet
0 km/h 2 dB 0.9 dB
120 km/h 0 0
Figure 2.2.1 Recommended values for UL power control headroom

2.6 BASE STATION RX LOSSES

This parameter needs to cover the total losses between the connector
of the antenna and the point at which the noise figure is referenced.

For a Node B without TMA, the reference point is the connector on top
of the Node B cabinet. The total loss is made up of the main feeder
loss, the connector losses and the jumper loss. The main feeder loss is
dependent upon the length and type of cable used.

T–Mobile/CDOC/ Page 13 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Cross-section Attenuation
[dB/100m]
½“ Highflex 17.0
½“ (LDF450) 11.5
7
/8“ (LDF550) 6.7
1 ¼“ (LDF650) 4.8
1 5/8“ (LDF750) 4.2
2 ¼“ (LDF1250) 3.8
Table 2.4.1: Typical values for cable attenuation

For a Node B with TMA the reference point is the receiver input to the
TMA. The total loss is made up of connector losses and jumper loss.

It is recommended that the following base station RX losses


values are used.

RX losses
(dB)
NodeB (w/o TMA) 2.5
NodeB (with TMA) 0.5
Table 2.4.2 : Recommended values for base station RX losses

2.7 BASE STATION ANTENNA GAIN

The planned antenna for 3G deployment is the Kathrein 742 212


multiband adjustable electrical downtilt cross polar antenna. This
antenna provides a boresight gain of 18 dBi in the frequency range
1920-2170 MHz.

A node B antenna gain of 18 dBi is recommended.

2.8 UPLINK SHO COMBINING GAIN

The total gains arising from soft handover can be considered to be


made up of two components: soft handover (SHO) combining gain and
the SHO fading margin reduction.

The SHO combining gain is defined as the reduction in transmitted


power in a soft handover state compared to the power required to
connect to the ‘closest’ of the cells without soft handover. It is assumed
that soft rather than softer handover is applied at the cell edge. Soft
handover means that selection combining is done at the RNC, which is
less effective than softer handover where maximum ratio combining
can be performed by the node B. The uplink SHO combining gain arises
from the diversity benefit in counteracting fast fading effects.

The gain depends upon the difference in path loss with the two serving
base stations. Nokia simulations [8, 9, 10 (fig 9.22)] show a gain of about

T–Mobile/CDOC/ Page 14 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

1.5 dB for equal path loss, falling to 0.5 dB for a 3dB difference and
then 0 dB at 6dB difference. These simulations results assume fast
power control is operating.

At the cell-edge, soft handover has the further benefit of reducing the
power control headroom because of the increased diversity. In
Ericsson’s simulations with no power control for Pedestrian B at 3 km/h
[11], the reduction in transmit power is 2.3 dB for equal path losses and
0.9dB for a 3 dB difference in path loss between the serving base
stations.

Nokia, despite the simulation results described above, recommend an


uplink SHO combining gain (MDC gain) of 0 dB [1,9].

Soft handover will provide some combining benefit in many cell edge
situations, and therefore a small positive gain seems to be justified. In
selecting a value for this parameter, a relatively cautious value is
chosen since there shall always be many locations (e.g. deep in-
building) with a single dominant server where the application of a large
SHO gain is not appropriate.

A uplink SHO combining gain of 0.5 dB is recommended.

2.9 SHO SLOW FADING MARGIN REDUCTION

As mentioned in the previous section, the soft handover benefit can be


split into two parts, an SHO combining gain and a reduction in the slow
fading margin.

The slow fading here encompasses both the outdoor slow fading, the
variation in penetration loss from building to building and the variation
in signal within the building by local shadowing. It is assumed
governed by Gaussian statistics. The slow fading margin reduction
arises when there are two potential serving cells. The slow fading
margin required can be reduced if only the better of two Gaussian
random variables needs to exceed a given level with a certain
probability, than is the case for a single variable of the same standard
deviation.

This fading margin reduction can also be obtained in GSM, but is not as
great for GSM because a hysteresis is generally applied which lowers
the confidence of being connected to the better server.

The fading margin reduction can be calculated relatively easily for the
cell-edge case when the slow fading is uncorrellated from cell to cell.
For equal signals with a standard deviation of 7 dB, the gain is about 6
dB. With a difference of 3.5 dB, this drops to about 4.5 dB.

T–Mobile/CDOC/ Page 15 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

To obtain an area probability, an integration over the cell area is


required. Nokia estimate the gain in the uncorrellated area integration
case to be 5.1 dB.

Unfortunately, slow fading of receive signals from two cell is often


correllated, and this reduces the gain. To understand this better,
consider an in-building user near a cell border. If he is deep within a
building, then it likely that the signals from both local cells are
substantially attenuated. However if the user is situated on the
opposite side of a building from one cell and therefore suffers high slow
fading attenuation, it is quite probable that the other serving cell is
pointing at that side of the

building so is not so strongly attenuated. The correlation in this case is


less and could even be negative!

For correllated fading, Nokia use a ‘Viterbi’ method to calculate slow


fading margin reduction. The gain is reduced to 3.3 dB [7]. In their link
budget a figure of 2 dB is used [7].

Nortel have also provided a similar calculation of SHO slow fading


margin reduction for the correllated case [13]. The results are shown in
the table below.

Outdoor Outdoor Dense


vehicular to indoor urban/ur
suburban ban
Area reliability 94% 94% 94%
Standard deviation 10.2 8.9 8
(dB)
Shadow margin (dB) 9.9 8.1 6.9
SHO slow fading 2.9 3.5 3.1
marging reduction (dB)
Table 2.9.1 Nortel results for slow fading margin reduction

In selecting values for the handover gains, a cautious value is chosen


since there shall always be many locations with a single dominant
server where the application of a large SHO gain is not appropriate.

A SHO slow fading margin reduction of 2 dB is recommended.

2.10 INTERFERENCE MARGIN

The uplink interference margin defines the allowable uplink noise rise
seen in at the base station.

The instantaneous noise rise can be related to loading (fraction of pole


capacity, η UL ) by the simple equation [9 eq. 9.2].

T–Mobile/CDOC/ Page 16 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

1
Noise _ rise[dB] = 10 log
1 − ηUL

This equaltion is plotted in figure below.

Figure 2.10.1 Classical relation between uplink loading and noise rise

A high noise rise, also known as the cell-breathing effect, is undesirable


as it signifies a unstable service area. A noise rise limit of 3 dB is
recommended as a long term planning assumption, which equates to
50% loading.

In the short term however, a lower noise rise limit maybe assumed in
order reflect the low traffic levels expected in the early network. An
initial planning load of 30%, or 1.5 dB noise rise has been proposed.

It should be noted that utilisation of such a low noise rise may limit the
use of high speed uplink bearers. This is especially true for a 384 kbps
bearer which is estimated to add a 33% load to cell 5, exceeding the
total noise rise limits.

A interference margin of 3 dB is recommended in the long


term. In the short term a value of 1.5 dB may be used.

2.11BODY LOSS

The body loss arises from the use of the UE close to the body or other
objects in close proximity. In practice, it shall depend upon the type of
terminal and the way it is used.

It is assumed that the speech service is carried on a handheld terminal


held against the head. For the speech service, the same body loss as
used by T-Mobile UK in their 2G link budget [14].

The use of a UE for data services typically means that the UE may
placed some distance from the body with a resultant lower body loss.
Nevertheless, the UE may still be attached to the body of the user, or
suffer some additional loss from being placed on furniture. It is
therefore recommended that a residual body loss is included in the link
budget for data services. A 2 dB loss in this case is proposed.

The value used within the vendor link budgets are shown below. The
values of 3 dB (voice) and 0 dB (data) used by most vendors seems to
derive from values used in the ITU UMTS technology evaluation.

5 Assumes a Eb/No of 2.9 dB, i=0.7.

T–Mobile/CDOC/ Page 17 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Bearer Nokia Nortel Siemen


Service s

Voice 3.0 3.0 3.0


Data 0.0 1.0 0.0
Table 2.11.1: Supplier values for body loss.

The recommended body loss values are as given in the table


below.

Bearer Body
Service Loss
[dB]
Voice 5.0
Data 2.0
Table 2.11.2: Recommended values for body loss.

2.12 SLANT LOSS

Current plans for 3G antenna deployment assume the use of cross


polarised antennas.

The use of cross polarised antennas is expected to degrade link budget


performance when compared to vertically polarised antennas.

T Mobile UK has performed measurements comparing the use of cross


polar base station antennas with vertically polarised antennas in
drivearound tests [15] for its 1800 MHz GSM network. The findings
show that the difference in performance with vertically polarised base
station antennas is very dependent upon the relative orientation of the
mobile antenna, and particularly so in more open environments. When
the base station and mobile antenna orientations align, a small gain of
1.9 dB was seen compared to the use of vertically polarised base
station antennas. However when the mobile antenna is perpendicular
to orientation of base station antenna, a loss of up to 16.6 dB was seen.

In suburban or urban areas, the dependency on the mobile antenna


orientation was less marked and a loss of 1.1/0.7 was seen in the
suburban/urban cases when mobile and base station antennas were
aligned, and a loss of 2.4/2.6 dB seen when antennas were
perpendicular.

Further evidence of a slant loss arises from traffic reductions seen in


live sites whose vertically polarised antennas have been swapped for
cross polarised one in capacity upgrade.

In the built-up areas, the slant can be explained by the filtering effect of
reflections from vertical surfaces which tend to depress the horizontal

T–Mobile/CDOC/ Page 18 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

component of the signal compared the vertical one. Transmitting on a


slant therefore reduces the amount of power carried on the dominant
vertically-polarised path.

T-Mobil have also made cross-polar vs. vertically polarised antenna


performance measurements for GSM at 900 MHz [16]. In drivearound
tests the results differ for a roof-mounted vertically polarised mobile
antenna and an in-car antenna. A significant degradation is seen in the
case of a roof antenna, but the degradation was not significant in the
measurements done with the in-car antenna. In further measurements
collected by A-bis interface logging, the average signal level appears to
drop by approximately 2.5 dB with cross polar antennas.

In the table below we show the recommendations of various other


suppliers with respect to their assumed slant losses.

Company Slant loss Reference


(dB)
Ericsson 1.5 [18]
Nortel 1.5 [12]
BT Cellnet 1.5
(Chelsea)
Kathrein 2
CSA 2
Nokia -
Table 2.12.1 Slant loss recommendations from suppliers

A slant loss of 2 dBi is recommended.

T–Mobile/CDOC/ Page 19 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

3. DOWNLINK PARAMETERS

3.1 NODE B MAXIMUM CARRIER POWER

This is the maximum output power per carrier.

Nokia offers two types of power amplifier with output powers of 20W
and 40W. In addition they offer a rollout configuration (ROC) which
splits the transmittted signal over 3 antennas, also called ‘omni
transmit sector receive’ (OTSR).

S111 OTSR/ROC
20W PA 43 dBm [-0, +2] 37.7 dBm6
40W PA 46 dBm [-0, +2] 40.7 dBm
Figure 3.1.1 Nokia Node B output power

It is currently assumed that 40W PAs will be used to maximise cell-edge


bearer rates.

A node B maximum carrier power appropriate to the Node B


type deployed should be used. 46 dBm is recommended for the
preferred 40W PAs.

3.2 CARRIER LOADING

This parameter defines the loading of the carrier power by all channels
supported by the carrier over the period that a high bearer rate packet
channel is utilised. The parameter is needed in order to calculate the
level of intracell interference at the receiver.

A high loading level is selected in order to maximise the available


power for the high speed packet bearers. A 10% headroom is assumed
to support a variation in power through a slot. One example of such a
variation is the multiplexing of the SCH1&SCH2 channels with the P-
CCPCH channels. If there is a difference in power levels allocated to
these groups of channels, the power is discontinuous through the slot.
In the pre-optimised network, Nokia recommend [19] that SCH1&SCH2
are each allocated 5% of downlink power, and P-CCPCH is allocated
3.2%. The P-CCPCH takes up 90% of the slot. This gives a average
loading of 3.4%, but rising 6.6% above this value in the first 256 chips
of the slot.

A carrier loading level of 90% is recommended.

6 5.3 dB splitter insertion loss [18]

T–Mobile/CDOC/ Page 20 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

3.3 MAXIMUM FRACTIONAL POWER FOR PACKET USER

This parameter indicates the maximum fraction of the maximum node


B carrier power that can be allocated to one high data rate packet user.

The available power for such a user is equal to the planned carrier
loading less the power required for common channels and power
required for the non-controllable traffic. It is assumed that one user can
use all of the power allocated to controllable NRT packet users. When
there is more than one user, the fractional power per user is reduced
and the bearer rate is reduced by the packet scheduler.

Nokia have made made some proposals for the powers allocated to the
various downlink channels in early deployment [19]. Another source of
settings for these channel powers are the 3GPP test specification
25.133 and 34.108. Nokia and 3GPP values are compared in the table
below.

Common control Fractional Power


channel
Nokia value 3GPP value
CPICH 10% 10%
SCH1 5% 3.2%
SCH2 5% 3.2%
AICH 1.6% 3.2%
PICH 1.6% 3.2%
P-CCPCH 3.2% 6.3%
S-CCPCH (SF=64) 12.6% 6.3%
Total (assuming 29.2% 28.7%
averaged SCH &
P_CCPCH)
Figure 3.3.1 Fraction of power used for common control channels

It is likely that the fraction of power will be reduced as a result of


optimisation.

For the purposes of determining an appropriate fraction of power


available for NRT high speed packet users, we assume that initially 30%
of maximum carrier power is allocated to the control channels, with a
likely improvement to 20% over time. Assuming the RT traffic utilises
15% of available carrier power initially, rising to 25% with the lower
common channel power, this leaves 45% of power for the NRT users.
This is summarised in the table below.

Initial Optimised
Common channels 30% 20%
Non-controllable RT 15% 25%
trafffic
NRT traffic 45% 45%
Total 90% 90%
Figure 3.3.2 Allocation of power between different types of channels

T–Mobile/CDOC/ Page 21 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

A maximum fractional power for packet user of 45% is


recommended.

3.4 BASE STATION TX LOSSES

The downlink TX losses should include all losses between the node B
output connector and the connector on the antenna. It should include
all connector losses, feeder losses, TMA insertion losses and jumper
cable losses.

The table below shows the MHA insertion losses for Nokia and Nortel.

Nokia Nortel
TMA insertion loss 0.3 dB 7 0.3 dB
Table 3.4.1 TMA insertion loss for Nokia and Nortel

The bias tee is housed within the Node B cabinet of both Nokia and
Nortel.

Allowing for additional connector losses on the TMA , the insertion of


the TMA can be expected to add about 0.5 dB to the total downlink TX
losses.

It is recommended that the following base station TX losses


values are used.

RX losses
(dB)
NodeB (w/o TMA) 2.5
NodeB (with TMA) 3
Table 3.4.2 : Recommended values for base station TX losses

3.5 DOWNLINK SLANT LOSS

The same slant loss is assumed for the downlink as the uplink.

A downlink slant loss value of 2 dB is recommended.

3.6 HIGHER FREQUENCY ADDITIONAL PATH LOSS

The FDD downlink band is 190 MHz higher than the uplink.
Consequently it suffers an higher path loss than the uplink.

7 Original MHA loss is 0.6 dB [20]. Nokia have recently commited to 0.3 dB for new MHA
product.

T–Mobile/CDOC/ Page 22 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

The COST 231 Walfish-Ikegami propagation model is used to estimate


the additional downlink loss [21]. For the mid-band frequencies of 1950
MHz and 2140 MHz, a difference in the intercept point of 1.6 dB was
obtained.

A higher frequency additional path loss value of 1.6 dB is


recommended.

3.7 UE NOISE FIGURE

The UE noise figure requirement is not explicitly stated in the 3GPP


specification, but is implied within the senstivity requirement. The 3GPP
specification requires a senstivity of –117 dBm for static conditions at
BER of 10-3. This is commonly considered to be the equivalent of
requiring a noise figure of 9 dB. Assuming an static Eb/No without
diversity of 6.5 dB [22] and non-orthogonality of 0 for a static channel,
the –117 dBm requirement can calculated as equivalent to a noise
figure of 9.6 dB. The small difference between 9 and 9.6 can be
considered as signal processing implementation margin. GSM terminal
performance is also a useful point of comparison. The receive
sensitivity of –102 dBm is equivalent to a noise figure of 9.7 dB8.

Selecting a noise figure based upon the worst case performance in the
standard is considered slightly too pessimistic for the typical terminal. 1
dB is subtracted from the 9 dB value to give a typical UE noise figure of
8 dB.

The values recommended by the suppliers are given in the table below.

Nokia Nortel

UE noise 8 9
figure
Table 3.7.1 Supplier recommendations for UE noise figure

A UE noise figure value of 8 dB is recommended.

3.8 NON-ORTHOGONALITY

The non-orthogonality is defined as the fraction of the total own cell


carrier power that is seen as interference at the UE receiver. The non-
orthogonality is used to relate downlink Eb/No to the amount of power
needed to sustain a given connection. The non-orthogonality increases
with the amount of multipath present in the propagation channel.

8 Eb/No (modulated bit) of 8 dB, and modulated bit rate of 270 kbps.

T–Mobile/CDOC/ Page 23 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Note that historically, the degree of non-orthogonality has been


historically called the ‘orthogonality factor’ despite the fact that
increasing ‘orthonality factor’ actually meant reduced orthogonality of
the codes so this terminology is mathematically incorrect. The term
non-orthogonality is therefore preferred in this document.

Non-orthogonality factors were calculated for various multipath


propagation conditions for the ITU evaluation of WCDMA [23], as shown
in the table below.

Propagation Model Non-


orthogonality
Indoor Office A 0.1
Outdoor to 0.06
Indoor/Pedestrian A
Vehicular A 0.4
Indoor Office B 0.25
Outdoor to 0.44
Indoor/Pedestrian B
Vehicular B 0.64
Table 3.8.1 ITU WCDMA evaluation non-orthogonality

A method for calculating downlink Eb/No from given test conditions can
be found in 3GPP 25.942 section 13, and this can also be used to
calculate the effective non-orthogonality. This method was used to
calculate non-orthogonality for the 3GPP propagation models.

Propagation Non-
Model orthogonalit
y
3GPP Case 1 0.11
3GPP Case 2 0.67
3GPP Case 3 0.6
Table 3.8.2 3GPP propagation model calculated non-orthogonality
factors

The values utilised by the vendors are as follows.

Nokia Nortel

Non- 0.5 0.4


orthogonality
Table 3.8.3 Supplier values for non-orthogonality

A value of 0.5 is selected as recommeded by Nokia. This value lies


between the figures for the more realistic propagation models of
Vehicular A/Pedestrian B and 3GPP case 3.

A Non-orthogonality value of 0.5 is recommended.

3.9 OTHER TO OWN CELL CARRIER POWER RATIO

T–Mobile/CDOC/ Page 24 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

This factor is used to calculate the amount of interefering signal from


other cells relative to the carrier power received from the serving cell.
The signals from other cells are completely non-orthogonal to the
wanted own cell signal.

The setting of this parameter takes consideration of the fact that the UE
is at the cell edge. The proposed setting of 1 suggests equal power
from own cell and one adjacent cell in the case of 2 dominant local
cells, or 2 adjacent cells at half the power in the case of 3 dominant
local cells. While is it possible to conceive of a tougher scenarios where
several other cells combine to exceed the carrier power from the own
cell, in most such instances soft handover should allow the signals
from several strong serving cell to be combined to counter the
increased interference level.

An other cell to own cell carrier power ratio value of 1 is


recommended.

3.10 UE DOWNLINK EB/NO

The derivation of downlink Eb/No is based upon the same approach as


the uplink of calculating a weighted mean of compensated values from
suppliers and the 3GPP specification.

The tables below summarise the Eb/No figures provided by suppliers.


Not that in the case of Nortel and Siemens, downlink figures are not
known and therefore the uplink figures are used with a 3 dB

Bearer Data Nokia Nortel Siemen


Service Rate s
[kbps]
AMR Voice 12.2 6.5 9.3 6.3
Circuit 64 - 7.8 5.2
Circuit 128 5 7.4 4.8
Circuit 384 - 7.2 3.6
Packet 64 5.5 6.2 4.4
Packet 128 5.0 5.8 4.1
Packet 384 4.5 5.6 3.2
Table 3.10.1 Supplier downlink (UE) Eb/No values [dB], 3 km/h users

Bearer Data Nokia Nortel Siemen


Service Rate s
[kbps]
AMR Voice 12.2 6.5 9.4 9.1
Circuit 64 - 7.8 6.8
Circuit 128 - 7.4 5.9
Circuit 384 - 6.9 5.7
Packet 64 5.0 5.9 6.6
Packet 128 4.5 5.6 5.9

T–Mobile/CDOC/ Page 25 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Packet 384 4 5.3 5.3


Table 3.10.2 Supplier downlink (UE) Eb/No values [dB], 120 km/h users

Another source of downlink performance information is 3GPP specification


25.101, although the results provided there require conversion to obtain
an Eb/No value. The derivation of the Eb/No figures from the performance
requirements in 3GPP 25.101 has been done as described in 25.942, and
the results of this conversion are summarised below.

Bearer Data Case 1 Case 2 Case 3


Service Rate
[kbps]
AMR Voice 12.2 16.3 13.0 9.0
Circuit 64 14.1 10.8 6.2
(BLER=1%)
Circuit 144 13.8 8.5 5.4
(BLER=1%) (128)
Circuit 384 14.1 7.2 5.7
(BLER=1%)
Packet 64 10.2 7.1 5.5
(BLER=10%)
Packet 144 10.0 5.5 4.9
(BLER=10%) (128)
Packet 384 10.0 4.9 4.9
(BLER=10%)
Table 3.10.3 3GPP derived downlink (NodeB) Eb/No Values [dB]

The compensations applied to the above figures in the above tables are
the same as the downlink.

To calculated a weighted mean the following weightings were applied for


the pedestrian case. The weighting for Nortel and Siemens are reduced
compared to the uplink case because the figures used are derived from
uplink simulation with 3 dB added to account for the lack of diversity gain.

Source Weigh
ting

Nokia 25%
Nortel 10%
Siemens 10%
3GPP Case 3 25%
3GPP Case 1 15%
3GPP Case 2 15%
Table 3.10.4: Weighting applied for calculation of downlink 3km/h Eb/No

For the 120 km/h case, the following weightings were applied

Source Weigh
ting

Nokia 40%

T–Mobile/CDOC/ Page 26 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

Nortel 10%
Siemens 10%
3GPP Case 3 40%
Table 3.10.5: Weighting applied for calculation of downlink 120 km/h
Eb/No

After performing the weighted mean calculation the


recommended value for downlink Eb/No are as follows:

Bearer Data 3 km/h 120


Service Rate km/h
[kbps]
AMR Voice 12.2 10.3 9.7
Circuit 64 7.6 6.8
Circuit 128/14 6.7 6.0
4
Circuit 384 6.6 5.8
Packet 64 6.5 6.2
Packet 128/14 5.7 5.6
4
Packet 384 5.3 5.2

3.4 DOWNLINK POWER CONTROL HEADROOM

As in the uplink case this parameter reflects the fact that the Eb/No
values used within the link budget are receiver Eb/No values which
assume power control can counter the effects of Rayleigh fading. In
reality at the cell edge the transmitted power is limited by the
maximum carrier power.

At a recent link budget meeting and in [10], Nokia suggest that


including such a margin for power control headroom is not necessary,
because the downlink power is shared between users and hence
individual codes powers may peak to invert fades, but the overall
carrier power should stay more or less constant. This might be true for
low data rate services, but for high data rates utilising 45% of the total
carrier power, this can no longer be safely assumed.

Unfortunately no simulation results for the downlink case exists.


Theoretically, the above-described benefit of any averaging over users
is countered by the lack of receive diversity on the downlink which
would be expected to increase the power control headroom because of
the deeper fading. The same value as the uplink is proposed.

A setting for downlink power control headroom of 0.9 dB for


packet users is recommended.

T–Mobile/CDOC/ Page 27 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

3.12 DOWNLINK SHO COMBINING GAIN

The SHO combining gain is defined as the reduction in transmitted power


when in a soft handover state compared to the power required to connect
to the strongest of the cells without soft handover.

On the downlink the transmit signals from two base stations can be
combined within the UE receiver. As a result, SHO on the downlink
provides not only increased diversity, but an effective increase in the total
transmitted power. Therefore downlink combining gain is greater than for
the uplink.

Simply addition of powers from two cells gives a 3 to 1.5 dB gain in


received power when the more distant cell has 0 to 4 dB increase path
loss than the ‘closer’ cell. A diversity benefit can be expected on top of
these figure.

Downlink link level simulation results have been performed by Nokia for
pedestrian A. These show a gain of 5.3, 3.8 and 2.5 dB 9 for a path loss
difference between serving cells of 0, 3 and dB [9]. These gains are rather
high because of the propagation model used is subject to strong Rayleigh
fading. Other Nokia simulations [10] show more modest gains of 4, 2.5
and 1 dB for path loss differences of 0, 3 and 6 dB. Within the link
budget, Nokia recommend an even more modest value of 1.2 dB [7].

A setting of 1.5 dB has been selected for the downlink SHO combining
gain. This considered relatively cautious given the value chosen for other
to own cell interference ratio.

A downlink SHO combining gain value of 2 dB is recommended.

9 Gain assumed in tthe transmit power per cell. Nokia assume gain relative to sum of
transmity powers and is hence 3 dB lower.

T–Mobile/CDOC/ Page 28 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

3. LINK BUDGET RESULTS

Spreadsheet:

Microsoft Excel
Worksheet

4.1 UPLINK LINK BUDGET

Table 4.1.1 shows the uplink link budget calculation for


pedestrian/stationary users assuming macrocell node B with MHA.

Speed 3 km/h
MHA? yes

Speech Circuit Packet Packet Packet unit


64 64 128 384
UE maximum output power 21.0 21.0 21.0 21.0 21.0 dBm
UE Antenna gain 0.0 0.0 0.0 0.0 0.0 dBi
UE EIRP 21.0 21.0 21.0 21.0 21.0

Node B Eb/No 6.9 4.1 3.2 2.6 2.4 dB


Node B noise figure 2.5 2.5 2.5 2.5 2.5 dB
Information rate 12.2 64.0 64.0 128.0 384.0 bps
Node B receive sensitivity -123.7 -119.3 -120.2 -117.8 -113.3 dBm
Power control headroom 2.0 2.0 0.9 0.9 0.9 dB
Base station RX losses 0.5 0.5 0.5 0.5 0.5 dB
Base station antenna gain 18.0 18.0 18.0 18.0 18.0 dB
Uplink SHO combining gain 0.5 0.5 0.5 0.5 0.5 dB
SHO slow fading margin reduction 2.0 2.0 2.0 2.0 2.0 dB
Interference margin 3.0 3.0 3.0 3.0 3.0 dB
Receive level at BS antenna -138.7 -134.3 -136.3 -133.9 -129.4 dBm

Path loss inc. body/slant loss 159.7 155.3 157.3 154.9 150.4 dB
Body loss 5.0 2.0 2.0 2.0 2.0 dB
UL slant loss 2.0 2.0 2.0 2.0 2.0 dB
Path loss less body loss or 152.7 151.3 153.3 150.9 146.4
slant loss
Table 4.1.1 Uplink link budget for UMTS FDD macrocell for speed of 3 km/h and
assuming deployment of MHAs.

T–Mobile/CDOC/ Page 29 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

4.2 MAXIMUM DOWNLINK PACKET BEARER RATE AT UPLINK LIMIT

Table 4.2.1 shows the calculation of the maximum downlink packet


bearer rate at the cell edge defined by the uplink link budget
calculation.

Speed 3 km/h
MHA? yes
Node B Config 40 PA, STSR

UL UL UL UL UL Units
Speech Circuit Packet Packet Packet
64 64 128 384
Node B max. carrier power 46.0 46.0 46.0 46.0 46.0 dBm
Carrier loading 0.9 0.9 0.9 0.9 0.9 %
Max. fract. power for packet 0.5 0.5 0.5 0.5 0.5 %
user
TX code power 42.5 42.5 42.5 42.5 42.5 dBm
Base station TX losses 3.0 3.0 3.0 3.0 3.0 dB
Base station antenna gain 18.0 18.0 18.0 18.0 18.0 dB
BS carrier EIRP 61.0 61.0 61.0 61.0 61.0 dBm
BS code EIRP 57.5 57.5 57.5 57.5 57.5 dBm

UL Path loss less body/slant loss 152.7 151.3 153.3 150.9 146.4 dB
DL slant loss 2.0 2.0 2.0 2.0 2.0 dB
Higher frequency additional 1.6 1.6 1.6 1.6 1.6 dB
path loss
Body loss 5.0 2.0 2.0 2.0 2.0 dB
UL SHO fading margin 2.0 2.0 2.0 2.0 2.0 dB
reduction
DL path loss incl. body/slant 159.3 154.9 156.9 154.5 150.0 dB
loss

UE noise figure 8.0 8.0 8.0 8.0 8.0 dB


Noise density -166.0 -166.0 -166.0 -166.0 -166.0 dBm/Hz
Non-orthogonality 0.5 0.5 0.5 0.5 0.5
Intracell interference density -167.6 -163.2 -165.2 -162.8 -158.3 dBm/Hz
Other to own cell carrier power 1.0 1.0 1.0 1.0 1.0
ratio
Intercell interference density -164.6 -160.2 -162.2 -159.8 -155.3 dBm/Hz
Total noise&interference -161.2 -157.8 -159.4 -157.4 -153.3 dBm/Hz
density

DL Eb/No 5.3 5.3 5.3 5.3 5.3 dB


DL power control headroom 0.9 0.9 0.9 0.9 0.9 dB
DL SHO combining gain 2.0 2.0 2.0 2.0 2.0 dB
Received code energy -99.8 -95.4 -97.4 -95.0 -90.4 dBm
Max DL cell-edge packet 327.2 413.5 380.1 419.2 460.8 kbps
bearer rate
Table 4.2.1 Maximum downlink packet bearer rates at uplink link budget cell edge for
speed of 3 km/h and assuming deployment of MHAs and 40W STSR configuration.

T–Mobile/CDOC/ Page 30 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

T–Mobile/CDOC/ Page 31 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

APPENDIX A - EB/NO DEFINITIONS AND RELATED FACTORS POWER-


CONTROL –TX-POWER-INCREASE AND POWER-CONTROL-HEADROOM

A.1 INTRODUCTION

This paper describes different parameters related to Eb/No and proposes


values to use in link budgets and simulation results.

A.2 RECEIVED EB/NO

Received Eb/No is related to the average received power as follows:

Pr = Eb/NoRX.(Io+No).R

where Pr is received power, Io is the effective interference spectral


density, No the noise spectral density and R is information rate.

Note that the information rate is the information rate after channel
decoding. It does not include bits required for DPCCH, MAC, RLC, and
associated SRB. However the received power Pr should include the
overhead for such bits, so that such overheads are also included in the
Eb/No value.

The received Eb/No is generally provided by link simulations, and values


are dependent upon the multipath propagation channel used and the link
quality requirement (FER, BLER, RBER, etc.).

A.3 TRANSMITTED EB/NO

Transmitted Eb/No is related to the average transmitted power as follows.

Eb/No TX (Io + No) R


Pt =
L

L is the transmission loss between transmitter and receiver. The


transmitted power is related to receive power as follows.

Pr = L. Pt

So why is there a difference between received and transmitted Eb/No?


Unfortunately the above relation does not apply to average powers.

T–Mobile/CDOC/ Page 33 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

P r ≠ L.P t

This is because the transmission loss and transmit power are not
independent, as a result of power control. Under ideal power control, we
can say that Pr is constant and therefore

Pt = Pr / L

The average transmit power is then given by the integral.


P t = Pr ∫ (1 / L ). p ( L).dL
0
where P(L) is the probability of a certain transmission loss. For full
Rayleigh fading and perfect (infinite dynamic range) power control, this
integral is found to be infinite (integral of 1/x function from zero to infinity
is infinite). Infinite average power is required to provide the infinite power
in the deepest fades!

In practice, of course, the utilised transmit power is not infinite, because


of 2 reasons: the multipath diversity reduces the propability of deep
Rayleigh fades; and the dynamic range of the transmitter is limited.
Nevertheless, it is found that fast power control does increase the
transmitted power over that predicted by the received Eb/No.

For fast moving mobiles, the power control is not able to track the fades
and the difference between transmitted and received Eb/No disappears.
Compared to the slow moving case received Eb/No tends to rise, and
transmitted Eb/No reduces.

A.4 POWER CONTROL TX POWER INCREASE

Power control TX power increase (PCTXPI) is defined as the difference


between received and transmitted Eb/No.

PCTXPI = Eb/NoTX - Eb/NoRX

Nokia sometimes refer to this as the Power Rise due to uncorrelated


fading.

A.5 POWER CONTROL HEADROOM

T–Mobile/CDOC/ Page 34 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

When average transmitted power approaches the peak power, it is found


that receiver performance is degraded since it is no longer possible to
invert deep Rayleigh fades because of the limited peak transmitter power.
Therefore one needs to increase the average power over that used when
perfect power control is assumed.

A important limiting case is when power control is deactivated entirely. In


this case transmitted power is independent of path loss and the received
and transmitted Eb/No are the same.

We define Power Control Headroom (PCH) as the difference between the


Eb/No’s for power control enabled and disabled. Two cases of power
control headroom apply, for received and transmitted Eb/No.

PCHRX = Eb/NoTX/RX, no PC - Eb/NoRX, pc

PCHTX = Eb/NoTX/RX, no PC - Eb/NoTX, pc

Nokia sometimes refer to this parameter as the Fast Fading Margin.

A.6 SIMULATION RESULTS

Ericsson have used simulations to quantify these effects for a Pedestrian


A and B uplink channel with receive diversity [12]. Pedestrian A is much
worse than Pedestrian B because of the reduced multipath. The tables
below summarises their results, for frame erasure rates of 1% and 10%
respectively.

FER= 1%
Max – Average Eb/No(rx Eb/No(tx Eb/No(tx Max – Average +
TX Power ) ) )- Eb/No(tx) –
Eb/No(rx Eb/No(rx)@full PC
)
42.3 (Full PC) 4.5 5.2 0.7 43
8.3 4.5 5.2 0.7 9.0
5.3 4.5 5 0.5 5.8
2.7 4.5 5 0.5 3.2
0 (no PC) 6.5 6.5 0 2

FER=10%
Max – Average Eb/No(rx Eb/No(tx Eb/No(tx Max – Average +
TX Power ) ) )- Eb/No(tx) –
Eb/No(rx Eb/No(rx)@full PC
)
43.1 (Full PC) 3.4 4.1 0.7 43.8
9.1 3.4 4.1 0.7 9.0
6.1 3.3 4.0 0.7 6.8
3.3 3.3 3.8 0.5 3.7

T–Mobile/CDOC/ Page 35 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

0 (no PC) 4.3 4.3 0 0.9

The right hand column calculates the difference the actual peak power
used in the simulation and the average transmitted power which would be
derived from the use of received Eb/No applicable for full power control. It
is minimised when no power control is used. In this case it is equal to the
power control headroom (PCHRX).

T–Mobile/CDOC/ Page 36 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

DOCUMENT HISTORY

Issue Date Details


0.1 Draft issue for comment.

Title - LINK BUDGET OFR UMTS FDD NETWORK


DOCUMENT APPROVAL

OWNER/AUTHOR
(SIGNED)

NAME (BLOCK CAPITALS): PETER STEVENS

CHECKED BY (SIGNED) DOCUMENT


REPRESENTATIVE

NAME (BLOCK CAPITALS):

APPROVED BY (SIGNED) MANAGER

NAME (BLOCK CAPITALS):

OWNERSHIP & CONFIDENTIALITY

T- Mobile is the trading name of T-Mobile (UK). No part of this document may be disclosed orally or in
writing, including by reproduction, to any third party without the prior written consent of T- Mobile (UK).
This document, its associated appendices and any attachments remain the property of T - Mobile (UK)
and shall be returned upon request.

T–Mobile/CDOC/ Page 37 of Issue: 0.1


© 05/08/11 T-Mobile SENSITIVE TO T-MOBILE

T–Mobile/CDOC/ Page 38 of Issue: 0.1

You might also like