OXFORD BROOKES UNIVERSITY

P00358 High Speed

Mobile Communications
3G Pilot Channel Coverage

Jorge Andrade Nr. 09020258

Jorge Pinto Nr. 09097562
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Table of Contents

1. Introduction.............................................................................................................................................2
2. Objectives................................................................................................................................................3
3. Antenna Radiation Patterns and Sectorisation (Task 1)...........................................................................3
3.1 Antenna configuration for omnidirectional sites (1 sector)..............................................................4
3.2 Antenna configuration for 3-sector sites...........................................................................................4
3.3 Antenna configuration for 6-sector sites...........................................................................................5
3.4 Electrical vs. Mechanical Tilt..............................................................................................................6
4. Path Loss propagation models (Tasks 2 and 3 – done by J. Andrade)......................................................7
5. Pilot Channel Coverage..........................................................................................................................10
5.1 Target Area......................................................................................................................................11
5.2 Scrambling Codes (Task 5 – done by J. Pinto)..................................................................................12
5.3 Pilot Power (Tasks 7 and 8 – done by J. Pinto).................................................................................14
5.4 Coverage and power distribution (Task 11).....................................................................................17
6. Final Site Configuration..........................................................................................................................27
7. Conclusions............................................................................................................................................30
8. References.............................................................................................................................................31

1
P00358 – High Speed Mobile Communications
3G pilot channel coverage

1. Introduction

The Universal Mobile Telecommunication System (UMTS) is the European version of the third
generation (3G) cellular networks. 3G systems are truly multi-service radio networks offering
video, telephony, streaming, and packet-based data services like web browsing or e-mail with
different data rates up to 2 Mbps (indoor low mobility environments). It uses Wideband Code
Division Multiple Access (WCDMA) as the air interface scheme. WCDMA is a spread spectrum
multiple access transmission technique in which a narrowband signal is spread in a large
frequency band (5MHz). Spreading codes are used to separate users or channels, hence each user
uses the same frequency at the same time as opposed to other multiple access schemes (e.g.
TDMA, FDMA). Alternatively, at the base station (Node B) a scrambling code is used to
separate one cell/sector from another. WCDMA are therefore interference limited systems
because all other users in the system are seen as a source of noise. This noise rise (as the number
of users increase) is one of many complex challenges that WCDMA network planning and
optimization engineers face.
UMTS radio system planning has to be done carefully because it’s a totally new system. Radio
propagation is not equivalent to other systems (e.g. GSM/TDMA/FDMA) because it uses higher
frequencies (2100MHz) hence more losses and it requires better signal strength (Eb/No) due to
the higher data rates. The network planning process normally follows the following phases [1]:
dimensioning, configuration, coverage and capacity, code and frequency allocation, parameter
planning, optimization and monitoring. The overall goal is to maximize coverage and capacity
while meeting the key performance indicators and QoS (Quality of Service). Some issues that
have to be considered at the planning stage are the location of the different mobile users, base
station/antennas configuration and locations, traffic distribution, link budget calculations,
propagation models, pilot power, etc.
Interference directly limits capacity of CDMA cell sites. If several pilots from different base
stations reach a given location with relatively equal strength, none of them could be dominant
enough for the phone to lock onto the network, a phenomenon known by pilot pollution. Pilot
power adjustment is therefore an important task in WCDMA network design. It is essential to
create a network plan where cells/sectors have clear dominance areas. Natural obstacles and
buildings should be used to create good dominance areas for WCDMA cells. Base stations

2
P00358 – High Speed Mobile Communications
3G pilot channel coverage

should overlap in fringe areas to accommodate hand-off. The degree of overlap, however,
shouldn’t be too large because that will cause too many unnecessary soft handovers per user on
the average. Most network planners agree that overlap should be 20-30 percent [2]. Another
factor in WCDMA systems is cell breathing. Cell breathing is the increase/decrease of a cell’s
range (maximum allowed path loss) due to the decrease/increase of the number of users,
repectively.
Several network planning software packages such as Aircom ASSET 3G are available to help in
the complex and challenging task of cellular network planning and optimization. These software
tools are normally combined with real data (e.g. clutter parameters) obtained through drive tests
in the target areas.

2. Objectives

The objective in this work is to undertake pilot coverage planning for a 3G network within the
ring-road boundary of Oxford using Aircom 3G ASSET. The coverage planning and analysis
will cover several points such as node Bs location, height, power distribution, path loss models,
antennas, sectorisation, down-tilting, scrambling code assigning, etc. We will explain why we
have chosen the final pilot coverage configuration as we present it and also discuss how the
coverage is altered as one or several of the points mentioned above changes.
This work is divided in several tasks. The core of the work (Task 11) was done by both elements
of the group.
Tasks 2 and 3 were done by Jorge Andrade – 09020258
Tasks 5, 7 and 8 were done by Jorge Pinto – 09097562

3. Antenna Radiation Patterns and Sectorisation (Task 1)

Since WCDMA systems are very sensitive to interference, it is of the utmost interest not to cause
or receive too much of it. Means of controlling interference in the network planning phase
include site configuration such as sectorisation, height, main lobe direction, beamwidth and tilt
of the antennas. Sectorisation is used primarily as a technique to increase system capacity,
although service coverage is generally improved at the same time. Typically 1 sector

3
P00358 – High Speed Mobile Communications
3G pilot channel coverage

(omnidirectional antenna) is used for low capacity macro-cell, 3 sectors are used for medium
capacity and 6 sectors are used for high capacity macro-cell configuration. Antenna radiation
patterns must be carefully selected for each sectorisation scenario in order to control the levels of
inter-cell interference and soft handover overhead. An optimum antenna beamwidth exists for
each sectorisation profile. Simulations have showed [3] that the best beamwidth for a 3-sector
and 6-sector sites are 65° and 33°, respectively. The figures of merit used in the simulation were
the coverage probability, number of users per cell per site, other-to-own-cell-interference ratio
and soft handover overhead.

For this work, we have chosen to use three site profiles with the following sectorisation: 1 sector
(omni) for rural areas, 3 sectors for suburban areas and 6 sectors for urban areas. The antennas
specifications used for the three sectorisation profiles are presented below.

3.1 Antenna configuration for omnidirectional sites (1 sector)

The antenna chosen for this site is a dual band omnidirectional
antenna from Kathrein [3]. It radiates evenly in all directions in
the horizontal plane. The Vertical radiation pattern is shown in
Figure 1. The major technical specifications are:

 Frequency Range: 1920 – 2170 MHz

 Polarization: Vertical
 Gain: 10 dBi
 Beamwidth (half power): 9°
Figure 1 – Radiation pattern for Omni antenna

This site configuration will be used in rural environments

with low capacity requirements.

3.2 Antenna configuration for 3-sector sites

The antenna that we chose for this site is SmartBeam DualPol Antenna with remote tilt and pan
from Andrews [4]. It has the following specifications:

 Model: SBH-1D6516DS
 Frequency Range: 1710 – 2170 MHz
 Horizontal Beamwidth: 65°

4
P00358 – High Speed Mobile Communications
3G pilot channel coverage

 Vertical Beamwidth: 6.5°

 Electrical Beam Tilt: 0 -10°
 Gain: 18.0 dBi
 Front-to-back ratio: 33 dB
 Polarization: cross-polar (±45°)

The horizontal and vertical radiation pattern of this antenna is shown in Figure 2 for a frequency
of 2110MHz and Tilt 0°. This antenna is going to be used in a 3-sector site with an azimuth of
120°. It can be seen in the picture that between two adjacent sectors the signal strength will be
about 12dB lower than the signal at the main direction. This guarantees that the interference from
other sectors is maintained as low as possible. The side lobes in the Vertical pattern shouldn’t
cause any interference problems because they are also about 12dB lower than the main lobe.

Figure 2 – Horizontal and Vertical Radiation Pattern for the 3-sector site antenna

3.3 Antenna configuration for 6-sector sites

The antenna used for this site configuration is a SmartBeam DualPol Antenna with remote tilt
and pan from Andrews [4]. This antenna has the following specifications:

 Model: SBH-2D3318
 Frequency Range: 1710 – 2180 MHz
 Horizontal Beamwidth: 33°
5
P00358 – High Speed Mobile Communications
3G pilot channel coverage

 Vertical Beamwidth: 6.5°

 Electrical Beam Tilt: 0 -10°
 Gain: 20.1 dBi
 Front-to-back ratio: 40 dB
 Polarization: cross-polar (±45°)

The horizontal and vertical radiation patterns are shown in Figure 3. This antenna is going to be
used in a 6-sector configuration with an azimuth angle between adjacent sectors of 60°. The level
of interference between adjacent sectors is again very low. The signal strength halfway between
adjacent sectors is about 12dB weaker than in the main lobe. Interference shouldn’t cause a
problem either in the Vertical plane. The side lobes are also about 12dB lower than the main
lobe.

Figure 3 - Horizontal and Vertical Radiation Pattern for the 6-sector site antenna

3.4 Electrical vs. Mechanical Tilt

Antenna tilt is defined as the deviation angle of the main beam of the antenna relative to the
azimuth plane. It’s one of the mechanisms to reduce the other-to-own-cell interference ratio.
Less power is delivered to the neighboring base station and most of the radiated power goes to
the area that is intended to be served by that particular base station. There are 2 types of tilting:
electrical and mechanical. They should be combined properly case by case to get the best

6
P00358 – High Speed Mobile Communications
3G pilot channel coverage

coverage with minimum interference. The effects of these two tilt options are shown in Figure 4.
Mechanical tilt widens the antenna lobe horizontally, which can be used in some cases as long as
the amount of interference between adjacent sectors is within tolerable levels. Electrical tilt, on
the other hand attenuates the radiation sideways while keeping the shape of the horizontal pattern
constant. This enables a more accurate network planning.

Figure 4 – Electrical vs. Mechanical Downtilting

4. Path Loss propagation models (Tasks 2 and 3 – done by J. Andrade)

Propagation models are used in the network planning process to predict the signal field strength
of a given transmitter in the computation area. Empirical models are normally used based on a
wide range of measurements in different locations in different environmental scenarios (urban,
suburban and rural). These models use free parameters and different correction factors that can
be tuned by providing real data measurements of the areas considered. One of the widely used
propagation models is the Okumura-Hata model.
In this work, we are going to use the standard Macro-cell model 3 that is supplied and supported
by Aircom ASSET for all cells in our design. The general path loss formula for this model is
given by:
L(dB) = k1 + k2log(d) + K3(Hm) + k4log(Hm) + k5log(Heff) + k6log(Heff)log(d) + k7(diffn)
+ C_Loss (Equation 1)
Where:

7
P00358 – High Speed Mobile Communications
3G pilot channel coverage

d distance from the base station to the mobile station (Km)

Hm height of the mobile station (m)
Heff effective base station height (m)
diffn diffraction loss
k1 constant offset factor
k2 multiplying factor for the log of the distance d
k3 correction factor for the mobile antenna height
k4 multiplying factor for the log of Hm
k5 multiplying factor for the log of the effective antenna height
k6 multiplying factor for log(Heff)log(d)
k7 multiplying factor for diffraction loss calculation
C_Loss clutter specification parameter

The Okumura-Hata model’s propagation loss for an urban area is given by [5]:
LHATA_URBAN(dB) = 69.55 + 26.16log(f) – 13.82log(Heff) + (44.9 – 6.55log(Heff))log(d) – a(Hm)
(Equation 2)
Where f is the frequency (MHz) and a(Hm) is the mobile antenna gain function. For a medium or
small city, a(Hm) is given by:
a(Hm) = (1.1log(f) – 0.7)Hm –(1.56log(f) – 0.8)

This expression does not usually have much meaning in practice because the mobile antenna
height considered is almost always the same (about 1.5 metres). For this value, the expression is
close to zero as we will see. Considering f=2000MHz (UMTS) and Hm=1.5m, we have
a(Hm)=0.05. Substituting in equation 2 and rearranging we get
LHATA_URBAN(dB) = 155.9 + 44.9log(d) – 13.82log(Heff) – 6.55log(Heff))log(d)
(Equation 3)
Comparing equation 3 with equation 1, we can see that they are equivalent. In this case, the
parameters associated with Hm, diffn and C_Loss are considered all together in the 1 st parameter
of equation 3.
For a suburban area the Okumura-Hata propagation model is given by
LHATA_SUBURBAN(dB) = LHATA_URBAN – 2[log(f/28)]2- 5.4 (Equation 4)

8
P00358 – High Speed Mobile Communications
3G pilot channel coverage

For a frequency f=2000MHz and using equation 3 we get

LHATA_SUBURBAN(dB) = 154.4 + 44.9log(d) – 13.82log(Heff) – 6.55log(Heff))log(d) (Equation 5)

Again, we can see that this equation is equivalent to equation 1. Finally, for a rural environment
the Okumura-Hata propagation model is given by

LHATA_RURAL(dB) = LHATA_URBAN – 4.78[log(f)]2 + 18.33log(f) – 40.94 (Equation 6)

For a frequency of 2000MHz and using equation 3 we have

LHATA_RURAL(dB) = 123.4 + 44.9log(d) – 13.82log(Heff) – 6.55log(Heff))log(d) (Equation 7)

This equation is also equivalent with all the previous ones, included the path loss model for a
UMTS macrocell used by ASSET (Equation 1).

The free space path loss (FSPL) model is a basic model that is used for academic purposes only.
It doesn’t take into account scattering, multipath phenomena, surrounding objects and obstacles
that cause reflection and diffraction, as opposed to the Okumura-Hata model. The FSPL path loss
model is given by

LFSPL(dB) = 20log(4πd/λ) (Equation 8)

Considering the wavelength λ = c/f = (3×10 8/2000MHz) = 0.15 and rearranging equation 8 so
that the distance d comes in Km, we get

LFSPL(dB) = 20log(4π/0.15) + 20log(d*1000)

= 20log(4π/0.15) + 20log(1000) + 20log(d) = 98.46 + 20log(d) (Equation 9)

It is easy to see that this equation is also equivalent to equation 1. The table below summarizes
all the k1 – k7 parameters that make the general path loss formula for the macrocell used by
ASSET represent the different Hata model scenarios as well as the FSPL model. These results
were obtained by comparing Equation 1 with Equations 2 – 9 for a frequency of 2000 MHz and
assuming a mobile antenna height Hm = 1.5 meters.

9
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Parameters Hata Urban Hata Suburban Hata Rural FSPL

K1 155.9 154.4 123.4 98.46
K2 44.9 44.9 44.9 20
K3* 0 0 0 0
K4* 0 0 0 0
K5 −13.82 −13.82 −13.82 0
K6 −6.55 −6.55 −6.55 0
K7* 0 0 0 0
* These parameters were included in K1.

5. Pilot Channel Coverage

This work is more concentrated in coverage rather than capacity. The goal is to ensure the
availability of the service in the entire service area. However, coverage and capacity are
interlinked through interference in CDMA networks. The base stations must be located and
configured such that minimum inter-site interference levels result. There must be areas with
dominant pilot signal reception in order to avoid pilot pollution and an excess handover overhead
(handover ping-pong).
One of the key system performance indicators is the Ec/Io (received energy per chip to noise
energy ratio) of the pilot signal is used to indicate the quality of the radio channel between UE
(user equipment) and the particular cell. Soft handoff algorithm uses Ec/Io of each pilot to decide
on which cell a user equipment is connected to. Ec/Io should be higher than -10dB for a good
network performance [6]. Since the Io is often the same as total overhead power from all the
neighbouring base stations, reducing their collective power reduces this kind of interference.
Another performance indicator is the received mobile power. The mobile’s minimum received
CPICH (Common Pilot Channel) power should be about -100dBm. However, due to the high
speed of the users along the ring road, a minimum power of -90dBm should be targeted for this
area. The table below summarizes these performance indicators as a function of the coverage
level.

10
P00358 – High Speed Mobile Communications
3G pilot channel coverage

5.1 Target Area

The target area in the pilot channel coverage planning is the area within the ring road of Oxford
as shown in Figure 5. The picture shows the concentration of people around the area. Although
we want to cover all the area with the best signal quality possible, the people distribution will
affect our decision in the sectorisation profile of each site. We can also see the height variation
along the target area from North to South and West to East. The area isn’t a flat area. It has hills
and valleys that can cause some problems in the signal coverage. These natural obstacles, on the
other hand, could be used to create good dominance areas for WCDMA cells.

Figure 5 – Target area for coverage (area within ring road of Oxford) and height profile variation

11
P00358 – High Speed Mobile Communications
3G pilot channel coverage

5.2 Scrambling Codes (Task 5 – done by J. Pinto)

The number of possible scrambling codes is very large. In the DL a complex valued Gold
18
code of length 2 1262,143 is used. In order to reduce the number of scrambling codes in use
and so minimize the time it takes a mobile to detect the signaling channels, the scrambling codes
are divided into 512 sets or segments. Each set (segment) consist of 1 primary and 15 secondary

codes associated with it. 512 separate codes would be sufficient for cell planning. Secondary
scrambling codes could be used if adaptive antenna techniques provide spatial separation of
users. The 512 segments are arranged into 64 code groups each containing 8 primary codes and
has its associated with the 15 secondary codes. With this procedure the cell search will be reduce
[7].

Primary Group Number Primary Scrambling Code Number

0 0, 16, 32, 48, 64, 80, 96, 112

1 128, 144, 160, 176, 192, 208, 228, 240
M M

62 7936, 7952, ……8048

63 8064, 8080, ……8176
Scheme of relation between code groups and codes per group

In the ASSET3G project there are 8 code groups from 0 to 7, and 64 codes per group from 0 to
63 cell/sector.

How is the scrambling code ID calculated?

As we can see on the Figure 6, it is showing the relation between the code sets and the code
groups. In order to calculated the scrambling code in our coursework we need to use just the
values from the primary ID in order to get the right values or stay within the 512 primary
scrambling codes [8].

12
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Hence per code sets:

Primary ID = 16i, i = 0 to 511

Secondary ID = 16i + k, k = 1 to 15

per code groups:

Primary ID = 16 x 8 x j + 16 x k

k = 0 to 7

j = 0 to 63

Figure 6. Relation between code sets and code groups [8]

According with the data on the ASSET 3G software, we should be using just the 1 from the
primary ID excluding the other 15 secondary codes. Hence:

SCID = (1 x 8) x j + (1 x k)

j = 0 to 63 and k = 0 to 7

13
P00358 – High Speed Mobile Communications
3G pilot channel coverage

5.3 Pilot Power (Tasks 7 and 8 – done by J. Pinto)

As we can see on Figure 7, it shows the differences between the Pilot Power and Max Power and
the Antenna EIRP values respectively. In our node 11 the cell parameters windows shows that the Pilot
powers is equal to 38 dBm and the EIRP is 56 dBm and the Maximum Transmitter Power is equal to 43
dBm with a EIRP equal to 61 dBm those values will change depending of the necessity of the coverage in
our network. But the difference between them will be always the same because by definitions the
difference on the Antenna EIRP is related to the Gain of our Antenna. In this case our Gain is equal to 18
dBi hence. See figure 8

Figure 7 - Pilot Power, Maximum Transmitted power, EIRP values.

EIRP = Power + Gain

EIRP = Pilot Power + Gain = 8 dBW + 18 dBi = 26 dBw / 56dBm

EIRP = Max Power + Gain = 13 dBw + 18 dBi = 31 dBw / 61 dBm

14
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Figure 8 - Antenna Gain value.

The Common pilot channel (CPICH) signals are used by Mobile station for channel quality
estimation, cell selection, and handover. The strength of the CPICH signal determines the
coverage area of the cell, impacts the network capacity, and thereby the quality of service, and is
therefore a crucial parameter in network planning and optimization. Pilot power is the most
important parameter that allows us to control the strength of the CPICH signal. The more power
is spent for pilot signals, the better coverage is obtained [9]. For all these reasons the Pilot
channel must be broadcasted with more power than the other channels.

As we can see on figure 8 our pilot channel value is 38 dBm without the gain.

8/ 10
38 dBm=8 dBw → 10 =6 . 3 w

15
P00358 – High Speed Mobile Communications
3G pilot channel coverage

But in order to compare those values with the realistic numbers given on figure 9 taken from
Sitefinder we need to use EIRP (Power + Gain) values from Pilot Power and Maximum
Transmitter power respectively.

Name of Operator Vodafone

Operator Site Ref. 49018

Station Type Macrocell

Height of Antenna 10 Metres

Frequency Range 2100 MHz

Transmitter Power 28.881 dBW

Maximum licensed power 32 dBW

Type of Transmission UMTS

Figure 9 - TX Power and Maximum Power taken from Sitefinder

Hence Antenna EIRP value from Pilot power is very similar with the Transmitter power
taken from the sitefinder.

26 / 10
56 dBm=26 dBw →10 =398 w
And the Antenna EIRP value from the Maximum Power is very similar to the Maximum
Licensed power shown on figure 9.

31 /10
61 dBm=31dBw →10 =1258 w

So the values shown on figure 9 for the Transmitter Power and the Maximum licensed power are
equivalent to the EIRP values we have from our antenna specification on ASSET 3G.

16
P00358 – High Speed Mobile Communications
3G pilot channel coverage

5.4 Coverage and power distribution (Task 11)

We started our coverage project using the Hata path loss propagation model for urban areas.
Only one carrier and one terminal type were defined to predict the coverage.

Starting by the south of Oxford, we placed two 3-sector base stations (Node 10 and Node 11) as
illustrated in Figure 10. We have chosen a 3-sector scheme because it’s not a considerable high
populated area. The power coverage displayed “Before” was done with the sites configured with
the default settings (antenna height: 15m, Pilot Power: 33dBm, Tilt: 0°). It can be seen that there
are coverage problems in the zones indicated by the arrows. The received power in some areas is
too weak (sometimes as low as -120dBm). The height profile between the nodes and the
problematic zones explain why the power is not reaching those areas. We moved Node 11 to a
higher place not too far from its old position and changed the antenna azimuth so that sector B
points to the problematic area. The power was increased to 40dBm and the sectors were down
tilted to minimize interference. At Node 10, the height of the antennas was increased as well as
the pilot power. The antenna azimuth was also changed so that sector B now points to the critical
area. The result after these changes is very positive. The received power level has increased to a
minimum value of around -90dBm which is very satisfactory.

Figure 10 – Pilot power distribution for Node 10 and Node 11

The pilot coverage for Nodes 10 and 11 are displayed in Figure 11. It can be seen a coverage
flaw in the direction of Node’s 11 sector A. In that area there is no dominant pilot signal.
Looking at the height profile we can see that the antenna is not high enough to cover that area
properly. After increasing the height of the antenna and also the pilot power we get a better
17
P00358 – High Speed Mobile Communications
3G pilot channel coverage

coverage where the pilot from Node11 sector A is dominant. Some electrical downtilt was also
used to minimize interference to other sites.

Figure 11 - Pilot coverage distribution for Node 10 and Node 11

Going up north, we introduce 3 more base stations (Nodes 12, 9, 14) as shown in Figure 12.
Each one of these sites uses a 3-sector configuration. The justification for this type of
sectorisation is that the population density is not considerably high to justify higher sectorisation
schemes. On the left side of the picture we can see the pilot coverage using the default settings
for Nodes 9, 12 and 14. The arrow highlights the coverage flaws that result. Pilot signals from
Node9B and Node12A are reaching the area with the same strength so there is no dominant pilot
signal. To understand the problem let’s look at the height profile from Node9C as illustrated in
Figure 13.

18
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Figure 12 – Pilot coverage distribution for Nodes 9, 12 and 14

The characteristics of the terrain don’t permit that the pilot signal from Node 9C reaches that
area without any obstructions. To fix the problem, we increased the height of Node9C antenna,
we increased the pilot power and down tilted the antenna so it can cover only that particular area.
Node12A was also down tilted to minimize interference. For the same reasons, Node 9B and
12B were also downtilted to achieve better dominant coverage areas.

Figure 13 – Height Profile seen from Node9C

Going up north again we reach city centre. Figure 10 shows the pilot coverage in that area. It can
be seen three new base stations (Nodes 1, 3 and 7). Node 1 is configured in a 6-sector scheme
because the population density is quite high in this area (city centre). The other two nodes (Node
3 and 7) are placed in relatively low population density areas so we have decided to use 3-sector
sites for each one of them. The screenshot “Before” in Figure 14 shows the pilot coverage when
Nodes 1, 3 and 7 have their default configuration (Tilt: 0°, Antenna height: 15m, pilot power: 33
dBm).

19
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Node1E was slightly downtilted and the pilot power was increased to cover all the area up to the
ring road. To minimize interference Node3B was also downtilted. Node1D and Node1C were
also downtilted to keep the coverage area within the city centre limits.

The area covered by Node7B and Node9A present some coverage problems. As it can be seen
there are places where none the pilots is dominant. Looking at the height profile from Node7B
depicted in Figure 15, we can see that there’s a hill that is limiting the signal propagation. We
then moved Node7 to the top of the hill and downtilted Node7C quite considerably to cover the
area down the hill.

Node 7B and Node14C were downtilted and the pilot power was increased in order to get a
better signal coverage with low interference.

The azimuth of Node9 was slightly moved so that Node9A points to Node’s 7 old location, to fill
the coverage gap.

20
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Figure 14 - Pilot coverage distribution for Nodes 1, 3 and 7

Figure 15 – Height profile seen from Node7B

21
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Moving further north within Oxford’s ring road, we introduce three more base stations (Node2,
Node6 and Node8) as illustrated in Figure 16. Node2 and Node6 are 3-sector base stations
serving an area with medium population density. Node8, on the other hand, is a 6-sector base
station serving an area with relatively high population density (Headington Centre).

Figure 16 - Pilot coverage distribution for Node2, Node6 and Node8

The pilot channel coverage displayed on the top is when the sites are configured with the default
options. We can see several coverage problems, specially in the area between Node6B and Node
8F. It can also be seen that the pilot signal from Node2B is reaching some of the area covered by
Node1B. Looking at the height profile between Node8F and Node6B (Figure 17) we can see the
difference in height between the two base stations. To get a better signal coverage with dominant
pilot channel areas, we increased the height and the power on Node6B ad slitghly downtilted it to
minimize the own-to-other cell interference. Node8F was higly donwtilted so it can cover only
the area next to it on top of the hill. Node7A was downtilted and its pilot power was increased.

22
P00358 – High Speed Mobile Communications
3G pilot channel coverage

All the other sectors were slightlydowntilted to avoid interference to neigbouring cells. The
results are shown on the bottom part of Figure 16. We can see the improvement in the coverage
with clear dominat pilot channel areas.

Figure 17 - Height profile seen from Node8F to Node6B

Finally we reach the north part of Oxford as shown in Figure 18. Here we can see two base
stations with omnidirectional antennas (Node5 and Node13). In those areas there are only empty
fields with very low population density, so it doesn’t justify higher sectorisation schemes. Node4
is a 3-sector site located in an area with medium population density.

Figure 18 - Pilot coverage distribution for Node4, Node5 and Node13

On the left of the figure it’s displayed the coverage with the nodes configured with the default
parameters. We can see again some coverage faults. The signal from Node2A is reaching the
area that should be covered by Node4B/Node5. To fix the problem, Node2A and Node4B were

23
P00358 – High Speed Mobile Communications
3G pilot channel coverage

donwtilted and the power was increased. The height of Node’s 5 antenna was increased to cover
a larger section of the ring road to its right.

Node4C’s was donwtilted and its azimuth was changed so now it points to the small population
area on the top left side of the picture. The height of the antenna and the power were also
increased to cover the ring road section on the left. The results are shown on the right side of
Figure 18.

The final pilot channel coverage for Oxford is displayed in Figure 19. We can see that in most of
the areas there is a dominant pilot signal. By analysing the parameter Ec/Io we get a high value
(> -10dB) from only one pilot in most of the area covered. These results show us that the
network that we have designed has minimal interference. The pilot pollution problem was
minimized and in most of the areas completely eliminated.

The pilot channel power distribution is displayed in Figure 20. It can be seen that in some few
zones the power is not entirely satisfactory. The power received at those areas sometimes goes
below −100dBm. We tried to work around this problem but we didn’t achieve an overall better
solution. We were also limited by the ring road boundary. In a real scenario some base stations
could be installed just outside the ring road and that would most probably eliminate or at least
attenuate those areas. Therefore, and considering the circumstances we can conclude that the
power distribution achieved is very satisfactory in general.

24
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Figure 19 – Final Pilot channel coverage for Oxford

25
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Figure 20 – Final pilot channel power distribution for Oxford

26
P00358 – High Speed Mobile Communications
3G pilot channel coverage

6. Final Site Configuration

In our design we have used three different sites (omni 1-sector, 3-sector and 6-sector) giving a
total number of 14 sites (two omni, ten 3-sector and 2 6-sector). Each site has its own location
and it’s configured differently depending on the location and the coverage needs. In this section
we will summarize the final parameters for each site and practical problems that could arise in
the location of the base stations proposed.

Node1 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 0° 15 33 4° 0°
Sector B 60° 15 34 0° 0°
Sector C 120° 15 37 4° 4°
Sector D 180° 15 33 4° 2°
Sector E 240° 20 36 2° 0°
Sector F 300° 15 33 4° 0°
Location: West Gate shopping. The only problem that could arise in the installation of this base
station would be getting permissions for that purpose.

Node2 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 0° 15 35 6° 0°
Sector B 120° 15 36 6° 0°
Sector C 240° 15 35 4° 0°
Location: St. Hugh’s College. The only problem that could arise in the installation of this base
station would be getting permissions for that purpose.

Node3 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 0° 15 33 4° 2°
Sector B 120° 15 33 4° 2°
Sector C 240° 15 33 0° 0°
Location: Botley Retail Park. Possible Problems: permissions.

Node4 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 0° 15 33 0° 0°
Sector B 120° 15 35 6° 0°
Sector C 270° 20 36 6° 0°
Location: Residential area near Wolvercote. Possible Problems: permissions and installation of a
mast to support the antennas.

27
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Node5 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A - 13 35 - -
Location: Open Field near Marston. Possible Problems: permissions, mast installation and
connection to core network.

Node6 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 0° 12 33 0° 0°
Sector B 120° 20 35 2° 2°
Sector C 240° 12 33 4° 2°
Location: Residential area in Marston. Possible Problems: permissions, mast installation.

Node7 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 0° 15 36 6° 2°
Sector B 120° 15 38 6° 2°
Sector C 240° 15 33 8° 2°
Location: Oxford Brookes University, Gipsy Lane. Possible Problems: permissions.

Node8 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 0° 20 35 2° 0°
Sector B 60° 15 33 0° 0°
Sector C 120° 15 33 0° 0°
Sector D 180° 15 34 6° 3°
Sector E 240° 15 33 4° 0°
Sector F 300° 15 35 6° 2°
Location: Headington Shops. Possible Problems: permissions.

Node9 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 340° 18 33 4° 3°
Sector B 120° 15 33 4° 0°
Sector C 230° 20 38 4° 2°
Location: Headington Shops. Possible Problems: permissions.

Node10 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 80° 20 35 2° 0°
Sector B 200° 20 40 0° 0°

28
P00358 – High Speed Mobile Communications
3G pilot channel coverage

Sector C 320° 15 35 6° 0°
Location: NHS building, Cowley Centre. Possible Problems: permissions.

Node11 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 60° 24 38 4° 4°
Sector B 180° 15 40 6° 0°
Sector C 300° 15 41 8° 0°
Location: Field near residential area in Rose Hill. Possible Problems: permissions, mast
installation.

Node12 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 45° 15 33 4° 2°
Sector B 165° 15 35 2° 2°
Sector C 285° 15 33 0° 0°
Location: Open Field near Hinksey Stream. Possible Problems: permissions, mast installation,
connection to the core network.

Node13 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A - 20 35 - -
Location: Open Field near Western Bypass. Possible Problems: permissions, mast installation,
connection to core network.

Node14 Azimuth Height (m) Pilot Power (dBm) Elect. Tilt Mech. Tilt
Sector A 45° 15 33 0° 0°
Sector B 165° 15 33 0° 0°
Sector C 285° 15 35 6° 2°
Location: Residential area in south Headington. Possible Problems: permissions, mast
installation.

7. Conclusions

The work carried out enabled us to get a feeling of the complex task that is network planning and
optimisation for a 3G cellular infrastructure. In this particular project however, we were not
working under any particular constraints which wouldn’t be the case if we were working for a
real network operator. Nevertheless, that didn’t make the job easier. The choice of the best site

29
P00358 – High Speed Mobile Communications
3G pilot channel coverage

locations, configuration, antennas and path loss models proved to be a big challenge, especially
in a cellular network where interference is the limiting factor.
The results of our design for Oxford were very satisfactory. We managed to configure the base
stations in such a way reducing inter-site interference and establish cell dominance in each
coverage area with respect to pilot signal power. The coverage gaps in terms of received pilot
power level were very few. Unfortunately we didn’t manage to solve the situation due to our
limitation in the fact that the location of the bases stations cannot be outside the ring road
boundary. Overall, we are satisfied with the final results of what we considered to be a
challenging work.

8. References

[1] Dinam, E., et al.,UMTS Radio Interface System Planning and Optimization, Technical paper,
2002

[2] http://www.umtsworld.com/technology/coverage.htm

[3] Kathrein Group: http://www.kathrein.com/

[4] Commscope: http://www.commscope.com/andrew/eng/index.html

[5] http://en.wikipedia.org/wiki/Hata_Model_for_Urban_Areas

[6] http://www.cdmaonline.com/

[7] Dr. Childs, G. High Speed Mobile Communications lecture notes. Oxford: Oxford Brookes
University, 2010.

[8] Andrew Richardson, WCDMA design handbook. Cambridge University Press, 2005

[9] Pilot power optimization and coverage control in WCDMA mobile networks. Retrieved from
http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6VC4-4KRY3YS-
2&_user=558526&_coverDate=12%2F31%2F2007&_rdoc=1&_fmt=high&_orig=search&_sort

30
P00358 – High Speed Mobile Communications
3G pilot channel coverage

=d&_docanchor=&view=c&_searchStrId=1324297468&_rerunOrigin=google&_acct=C000028
481&_version=1&_urlVersion=0&_userid=558526&md5=f62fdb3051559770408ca1b3838a9d6
5

31