15/7/05 1

Company Confidential
1
Site Evaluation
Frequency Planning
BTS Installation and Commissioning
RF Coverage Verification
Optimization
Competitive Comparison
Interference Management
RF Activities in Cellular Systems
15/7/05 2
Company Confidential
2

Site
Evaluation
RF Planning
Installation
Commission
Verification
Build-Out & Turn-up
Site Roll out Coverage Verification
Optimization
Trouble-
shooting
Network
Expansion
Fraud
Detection
In Service
Network Optimization Competitive Comparison
Interference
Management
Continous
RF Activities Application
Segments
15/7/05 3
Company Confidential
3
Coverage Prediction
Frequency Planning
Prediction Tool
Band Clearance
Test Transmission
RF Coverage Verification
Optimization
Troubleshooting
Competitive Analysis
Interference Monitoring
Prediction Tool Modelling
Drive Testing
Commissioning
Maintenance
Trouble Shooting
BTS Testing
Despatch Inspection
Quality Assurance
Troubleshooting
Repairs
MS Testing
Tools
Tools for RF Activities
15/7/05 4
Company Confidential
4

Search Area
Selection
Site Physical
Qualification
Coverage
Prediction
Band
Clearance
Test
Transmission
Reject
Acquire
Action Steps
Site Evaluation
15/7/05 5
Company Confidential
5
 Study of Contour / City Map.

 Identifying potential search zones.

 Correlating with nearby existing sites.

 Drive/Walk through physical Land
Survey.

Search Area Selection
15/7/05 6
Company Confidential
6
 Identifying Potential Sites in Search zone.
 Physical verification of the infrastructure.
 Major obstacles around.
 Future potential of major obstacles.
 LOS to other sites.
 Capturing photographs.
 Logging GPS coordinates (Lat,Long,Alt).
Site Physical Qualification
15/7/05 7
Company Confidential
7
 Uses Prediction Software tool to estimate coverage

 Software has electrical map and the city contour information

 Parameters like frequency,power,antenna parameters,height,
etc are fed to the software

 Based on the city model and these parameters the tool
estimates the coverage area .

 Data for all sites are fed to estimate the level of interference.
RF Coverage Prediction
15/7/05 8
Company Confidential
8
WORK
STATION
PLOTTER
Prediction Tool
RF Planning
15/7/05 9
Company Confidential
9
The data available after test transmission is analyzed by
the measurement analysis system.This system consists
of a work station and a plotter.The work station has
software which contains the map of the geographical
area to covered by the network.This map is accurate and
is in terms of earth coordinates. The test transmission
data is fed to this work station.The work station software
then correlates this data with the map and plots out the
coverage on the map.The coverage level could be preset
in zones of various color like good,average,poor and no-
coverage.With this map representation the sites
capability is determined
15/7/05 10
Company Confidential
10
Reliability of Prediction Tool
 Prediction tool uses either area- to - area or point-to-point prediction models.

 Area-to-Area are based on prediction models like HATA,Walfish,etc. These prediction
tools may give a standard deviation from later actual measured coverage in the range
of 12 - 14 dB and above.

 Point-to-Point model based prediction tool are specific for a particular terrain and
hence are more accurate and will have a standard deviation of 7-8 dBs and is
generally accepted.

 Prediction tool which deviates from actual measured coverage by 2-3 db over 90%
predicted area is considered to be excellent. This level of accuracy can only be
maintained by consistently modeling the planning tool.

RF Coverage prediction
15/7/05 11
Company Confidential
11
 High cost of installing base stations
 Are clear bands really free of interference?
 Minimize risk by assessing interference before committing funds
=
Band Clearance
15/7/05 12
Company Confidential
12
The cost and effort involved in developing cell sites is extremely high. Band
clearing can play a critical role in the site qualification process.

The goal of band clearing is to minimize the risk of interference and to
understand the signal environment before committing funds to develop the
site. Many factors drive the decision to choose one site over another: real
estate issues, accessibility, maintenance issues, RF propagation, etc.

Risk of interference should be a deciding factor when choosing between
multiple candidate sites. Lower risk implies that less time will be spent
combating interference after the network has been turned up. Minimizing
interference will improve network performance and quality of service.

Even when a site has already been committed to, or there is no choice, it is
critical to make the measurements necessary characterize the signal
environment in order to understand the risk and types of interference. This
process will greatly simplify the job of dealing with interference after network
turn-up.
15/7/05 13
Company Confidential
13
 Characterize RF signal environment
 In-Band and adjacent bands
 Long term monitoring
 Noise Floor Characteristics

 Determine types and sources of potential interference both in
uplink and downlink bands.

 Minimize / Understand risk of expensive interference problems
Band Clearance
15/7/05 14
Company Confidential
14
GP-IB

Spectrum Analyzer
Antenna
PC Controller
System
GPS Receiver
RS-232
Band Clearance
15/7/05 15
Company Confidential
15
 Power Statistics
 Estimate Probability of Interference
 Noise Floor Characteristics
 Logging of Signals in the Band
 Channel Occupancy
Band Clearance
15/7/05 16
Company Confidential
16
In order to best understand the signal environment and assess the risk of interference it is
necessary to make statistical measurements of power over a long period of time.

The power statistics are used to estimate the probability of interference in each channel.
These probabilities can be used to compare prospective sites based on interference risk.
Since these measurements are channelized they can also be used as an aid in frequency
planning.

The local noise floor can vary from site to site. Various sources of broadband interference will
have an effect on the level of noise at a given location.

Characterization of the noise floor prior to turn-up will help define power settings required to
achieve a desired signal-to-noise ratio. It can also indicate sites where the noise floor may be
excessively high. Noise floor characterization is particularly important for CDMA systems.

Signal logging is done to get a record (frequency and power) of signals present in the band.

Channel occupancy can characterize the usage pattern of signals in the band and possibly
help to define a frequency plan that can work with existing signals.

Again, long-term monitoring is key to getting the best results.
15/7/05 17
Company Confidential
17
Measure test transmitter signal strength as a
function of location.
Generate Coverage Map
Evaluate foliage and shadowing effects.
Help set modeling parameters in RF planning
software.
Calibrate planning software tool.
Test Transmission
15/7/05 18
Company Confidential
18
1. Test Transmitter
2. Drive System
Two Components
BTS
Simulator
Power
Amplifier
Test Transmission
15/7/05 19
Company Confidential
19
The typical configuration for a pre-installation RF coverage measurement
system has two major components -- a transportable signal source and a
drive system.

The signal source is placed at the location of the prospective cell site. The
transmitter is elevated to the proposed antenna height of the cell site, often
using a scissor lift or a crane. It may be desirable to execute the drive test
with various antenna heights. Initial measurements are made with a
continuous wave (CW) signal. Typically CW testing provides adequate data
for pre-installation coverage assessment.

In some cases a modulated signal source may be used. A decision must be
made to trade off time to turn-up for a more complete coverage data set. A
modulated signal source also requires a more sophisticated measurement
at the receiver to capitalize on the modulation.

The drive system contains a receiver to measure signal strength and a
mechanism for determining location (typically GPS vehicular navigation, or
both).
15/7/05 20
Company Confidential
20
System Controller
GIS
GP-IB

GPS Receiver
RS-232
RS-232
GSM Test MS
GSM BTS
Receiver
Drive Test System
Test Transmission
15/7/05 21
Company Confidential
21
 Test Transmitter can be Single Channel CW Source; or a GSM
BCH Transmitter

 For CW source, Receiver should be preferred in Drive Test
System.

 Receiver can do CW measurements accurately, because Mobile
does Channel Power measurement.

 For a GSM BCH transmitter, use a different network code, or
preferably activate cell barring, to avoid traffic discrepancies.
Test Transmission
15/7/05 22
Company Confidential
22
 For CW transmission, each measurement value
should be a running average of 50 samples taken over
a distance of 40 wavelengths.

 This process, will result into 95% confidence in the
predicted coverage with the CW transmitter, with reference
to the actual coverage later.
Meeting Lee's Criteria
15/7/05 23
Company Confidential
23
DTS
Tst.Tx
Test Transmission
Test Transmission
15/7/05 24
Company Confidential
24
Once an approximate site is selected,the test transmission is to
be done.The test transmission as the term states is a process
by which a test transmitter ( BTS Simulator ) is temporarily
installed at the site and any of the allotted GSM frequencies is
transmitted. Now this is transmission is received by a Drive Test
System installed in a mobile van which moves around the plan
and gives a plot of signal strength received in the cell.

The received level should be estimated at -85 dbm,which
means a good outdoor coverage and considering on average
15 db indoor margin,this level of outdoor will give somewhat
acceptable indoor coverage
15/7/05 25
Company Confidential
25

Navigation
Signal Strength
GPS / DR !!!
Mapping

Process
GPS/DR is essential if Prediction tool modeling is to be done
Analysis
Measurement
Test Transmission
15/7/05 26
Company Confidential
26
This slide depicts a typical configuration for RF coverage
measurements.
The GPS receiver and / or the vehicular navigation system known as
dead reckoning (DR) measures location.

In a CW drive system the signal strength is simply peak power
measured at the transmit frequency. For more sophisticated
modulation types the receiver typically makes an estimate of the bit-
error-rate (BER). This is done by comparing a reference signal with the
received signal. The reference signal is constructed to match the
known transmit signal.

The resulting output data is displayed on a map. Often the display is
done in the RF planning prediction tool. This allows for comparison of
the measured coverage to that predicted by the model.
15/7/05 27
Company Confidential
27
Three types of GPS Receiver's for Navigation
 GPS
-- accuracy in the range of 40 - 60 m

 GPS with Dead Reckoning
-- with Compass and Wheel/Odometer
-- improves performance during signal loss( street canyon)

 Differential GPS
-- improves the absolute accuracy ( in few meters)
-- Local correction signal is transmitted from a separate Tx.
-- FM receiver with the GPS picks up and applies the correction
Something on GPS
15/7/05 28
Company Confidential
28
GPS Fix !!
 GPS Fix occurs when it gets Satellite Signals.

 A reliable GPS should be at least a 8 channel receiver.

 2D fix : at least 2 satellite available ( lat,long).

 3D fix : at least 3 satellite available (lat,long,alt).
GPS Interface
 GPS communicates on the RS232 interface with the PC.
 Standard interface protocols are TAIP,TSIP or NMEA.
Something on GPS
15/7/05 29
Company Confidential
29






F=1
F=2
F=3
F=4,8
F=5,9
F=6,10
F=7






F=1
F=2
F=3
F=4,8
F=5,9
F=6,10
F=7






F=1
F=2
F=3
F=4,8
F=5,9
F=6,10
F=7
F= 1,2,3,4,5,6,7,8,9,10
 GSM uses concept of cells
 One cell covers small part of network
 Network has many cells
 Frequency used in one cell can be used
in another cells
 This is known as Frequency Re-use
Clusters
Co-Channel ( Re-use ) Cells
Frequency Re-use
Frequency Planning
15/7/05 30
Company Confidential
30
GSM uses the concept of cells. One cell covers a small
part of the network. A GSM network will have several
cells. Since a cell has limited area, the frequency used in
this cell can be re-used in some other cell. This is known
as frequency reuse. By using this concept, all cells will
have appropraite frequencies and hence can be
increased, by increasing cell and re-using the
frequencies. The cells which use the same frequency
numbers are known as re-use or co-channel cells.
Adjacent cells should not use the same frequency, as
they would interfere with each other and disturb the
speech.
15/7/05 31
Company Confidential
31
Objective
 Optimum uses of Resources

 Reduce Interference
Frequency Planning
15/7/05 32
Company Confidential
32






A






A
Q = D
R
C / Ic = 9 db
Q = Sqrt ( 3 x N )
Co - Channel Re-use factor
15/7/05 33
Company Confidential
33
Frequency reuse implies that in a given coverage area there are
several cells that use the same set of frequencies.These cells
are called co-channel cells,and the interference between
signals from these cells is called co-channel interference.An
increase in transmit power and decrease in cell size leads to
this problem . Considering each cell size to be same co-channel
interference becomes the function of the radius of the cell ( R )
,and the distance to the center of the nearest co-channel cell
(D). This ratio of D/R is termed as co - channel reuse ratio ( Q )
. By increasing Q the spatial separation between two co-
channels is increased thereby reducing interference.
A small value of Q provides larger capacity by more
reuse,where as a large value of Q provides improved
transmission quality ,due to a smaller level of co-channel
interference.
15/7/05 34
Company Confidential
34
 Adjacent ARFCN's should not be used in the same cell

 It will have no problems in Downlink*, but will have high risk of
uplink interference (due to mandatory uplink power control ).
* If Downlink dynamic power control is not used
- 70 dbm ( C/Ia = 20 )
- 90 dbm ( C/Ia = -20 )
5 dbm
33 dbm
 Since all the ARFCN's in a cell are frame
synched, Timeslot numbers will align on
all the ARFCn's
Adjacent-Channel Re-use Criteria
15/7/05 35
Company Confidential
35
 Adjacent ARFCN's can be used in adjacent cells, but as far as
possible should be avoided.

 As such separation of 200 Khz is sufficient, but taking into
consideration the propagation effects, as factor of protection
600 Khz should be used*.

 In the worst, Adjacent ARFCN's can also be used in adjacent
cells by setting appropriate handover parameters ( discussed
later in optimization)

* Practically not possible in most of the networks due to tight reuse
Adjacent-Channel Re-use Criteria
15/7/05 36
Company Confidential
36


Omnidirectional Cell
BTS






Sectorial Cell
BTS
 Low gain Antennas
 Lesser penetration/directivity
 Receives Int from all directions
 Lower implementation cost
 High gain Antennas
 Higher penetration/directivity
 Receives Int from lesser directions
 Higher implementation cost
Cell Configuration
15/7/05 37
Company Confidential
37


3,6,9
A










A
B





C

3,6,9
B

3,6,9
C
 Receives Interference from all directions
Interference in Omni-Cells
15/7/05 38
Company Confidential
38




A1
A2
A3
3
6
9



B1
B2
B3
3
9
6



C1
C2
C3
3
6
9
 Receives
Interference
from lesser
directions.
Sectored Cells
15/7/05 39
Company Confidential
39
 Re-use Patterns ensures the optimum separation between Co-Channels.

 Re-use pattern is a formation of a cluster with a pattern of frequency distribution in
each cell of the cluster.

 Same cluster pattern is then re-used.
Preferred Re-use Patterns
Omni - Cells : 3 cell, 7 cell, 12 cell, 14 cell, 19 cells etc
Sector - Cells : 3/9 , 4/12, 7/21
Re-use Patterns
15/7/05 40
Company Confidential
40


















A1
A2
A3
B1
B3
C1
C2
C3


















A1
A2
A3
B1
B2
B3
C1
C2
C3


















A1
A2
A3
B1
B2
B3
C1
C2
C3


















A1
A2
A3
B1
B2
B3
C1
C2
C3


A1






B1
B2
B3






A1
A2
A3
B2






C1
C2
C3



C2
C3



C2
C3



C2
C3


A1
3/9 Re-use Pattern
15/7/05 41
Company Confidential
41


















A1
A2
A3
B1
B3
C1
C2
C3


















A1
A2
A3
B1
B2
B3
C1
C2
C3


















A1
A2
A3
B1
B2
B3
C1
C2
C3


















A1
A2
A3
B1
B2
B3
C1
C2
C3


A1





B1
B2
B3






A1
A2
A3
B2






C1
C2
C3




C2
C3




C2
C3




C2
C3


A1
Using ARFCN's 1,2,3,4,5,6,7,8,9 , do the channel allocation for
the below cells using 3/9 pattern
Exercise !!!
15/7/05 42
Company Confidential
42
Adjacent Channel Interference is very difficult to avoid
within the cluster itself.


















1
4
3
2
8
5
7
9
6
Frequency Allocation in 3/9
patterns
15/7/05 43
Company Confidential
43










D1
D3
B1
B3












C1
C2
C3
D1

















A1
A2
A3
B1
B2
B3
C1
C2
C3


















B1
B2
B3
A1
A2
A3
C1
C2
C3


C1






D1
D2
D3



D2
D3



B2
B3



B2
B3








D2 C1
C3
B2






D2
D3
A1
A2
A3






B1
B2
B3
C2






D1
D2
D3


A1
4/12 Reuse Patterns
15/7/05 44
Company Confidential
44
Using ARFCN's 1,2,3,4,5,6,7,8,9,10,11,12 do the channel
allocation for the below cells using 4/12 pattern.


















D1
D2
D3
C1
C3
B1
B2
B3


















C1
C2
C3
D1
D2
D3
A1
A2
A3


















A1
A2
A3
B1
B2
B3
C1
C2
C3


















B1
B2
B3
A1
A2
A3
C1
C2
C3


C1






D1
D2
D3






B1
B2
B3
C2






D1
D2
D3




D2
D3




B2
B3




B2
B3


A1
Exercise
15/7/05 45
Company Confidential
45












1
3 5






2
4 6






7
9
11
12
10
8
4/12 pattern avoids adjacent channels in adjacent cells
4/12 Pattern Channel Allocation
15/7/05 46
Company Confidential
46
 Larger reuse patterns give reduction in interference

 Re-use patterns becomes more effective with sectorial cell
configurations.

 To implement large patterns ( like 4/12, 7/21) , more channels
are required.

 So with less resources, the best way to plan is :

1. Use optimum no of channels per cell.
2. Thus, increase the pattern size.
Reuse Patterns Conclusion