Rf-Planning

Introduction to RF Planning

A good plan should address the following Issues :

• Provision of required Capacity.

• Optimum usage of available frequency

spectrum.

• Minimum number of sites.

• Provision for easy and smooth expansion of the

Network in future.

• Provision of adequate coverage.


In general a planning process starts with the inputs from the

customer. The customer inputs include customer requirements,

business plans, system characteristics, and any other constraints.

After the planned system is implemented, the assumptions made

during the planning process need to be validated and corrected

wherever necessary through an optimization process.

We can summarize the whole planning process under the 4 broad

headings

• Capacity planning

• Coverage planning

• Parameter planning

• Optimization

CELLULAR ENGINERING OBJECTIVES

1) To provide adequate coverage - Contiguous coverage of the required areas without appreciable holes

- Adequate depth of coverage (i.e. outdoor or indoor , 2 W or 1.2 W mobiles )

to meet the company’s marketing plans.

2) To provide adequate network capacity - Accommodating traffic in the busiest hour with only a low probability of

blocking (congestion).

3) To accommodate network growth - Extension of coverage in new areas

- Expanding the network capacity so that the quality of service is maintained

at all times.

4) To achieve a cost effective design - Lowest possible cost over the life of the network while meeting the quality

targets.

COST JUSTIFICATION OF CELLULAR RNP

The cellular mobile radio system design can be broken

down in the following elements, which have a mutual

relationship. - Reuse of frequency channels

- Co- channel interference reduction

- A desired minimum carrier to interference ratio (C/I)

- Handover mechanism

- Cell Planning

Historical perspective - Wireless telephony network design is relatively new business with a 10-

15 year history . During this period many new tools and techniques

have been developed:

- More accurate radio coverage prediction

- More accurate facility network design

- Enhanced field measurement analysis to improve network

performance.

- New technology applications ( microcells, repeaters, smart antennas

systems. )

- Better tools and methods to evaluate and predict traffic conditions


The challenge of accurate cellular network planning is still a

complex task.

Potential cost of Opportunities Lost Due to Network Planning

problems

Lost Subscribers

- Lost base subscriber fee revenues

- Lost enhanced service fee revenue

- Lost airtime revenues (local and long distance)

- Damaged reputation will impact competitive strength

Cost Considerations That Include in the Design of a quality

network

- Design optimal network : extensive modeling and numerous

revision of design.

- Acquire radio site candidates that meet the design

criterion.

- Manage delays in permitting / zoning of best candidates

- Extensive testing of radio site performance (coverage )

before commissioning.

- Integration of field measurements in design.


Design Activity to compensate for Improperly designed or

less than than optimal radio site in design.

- Modify cell operational parameters (eg. Handover values

and location)

- Modify output power

- Modify equipment (eg. Change antenna )

- Move site location

- Add new sites (micro or macro cells)


An equation for Costing Comparison of Accurate

Network Planning Option one : Poor design / no redesign

- Weak competitive position

- Lost disgruntled subscribers

- Earn a poor service reputation (Weak attraction for

new subscribers ).

Option two : Quality network design

- Additional design cost (engineering and equipment ).

- Teardown and reinstall cost.

Simple equation for characterizing cost /benefits

- Quality network performances = ( Cost of engineering ,

equipment, installation ) – (Lost revenues, cost of

engineering, equipment, installation )

- The benefits of quality design should farweigh lost

revenues particularly in the fact of competition from

new wireless companies.

DESIGN CONSTRAINTS

The objective of radio planning is a technical

realization of the marketing requirements, taking

into account of the following constraints.

- Technical requirements from the license conditions.

- GSM system specific parameters (e.g. GSM recs 5.05

etc.)

- Manufacturer specific features and parameters.

- Radio communications principles and fundamentals.

- Budgetary factors.

LICENSE CONDITIONS

An example of technical requirements following from a

license. Coverage requirements.

- Class 2 or class 4 coverage of 60 % of the population 12 months from commercial

launch.

- Class 2 or class 4 coverage of 95 % of the population 36 months from the

commercial launch.

Quality of coverage

- Service to be available in 90 % of the declared area and for 90 % of the time.

Grade of Service

- Endeavour to achieve 5 % or better

Frequency Allocation

- One of the major limitations in the GSM 900 system is the number of frequencies

available to a GSM network operator. There is a relatively small bandwidth

available that has to be divided over all the licensed operators. Most network

operators are limited to 30-60 frequencies for handeling all traffic.

- GSM 1800 offers 75 MHz bandwidth

MANUFACTURER SPECIFIC PARAMETERS

- BTS Transmit power

- Receiver sensitivity

- Combiner performances

- Cable loss

- Antenna performance

- Availability of frequency hopping and power control

- Handover algorithm

- Capacity – number of TRX provided.

RADIO COMMUNICATION FUNDAMENTALS

- Propagation loss

- Shadowing

- Multipath fading

- Power link budgets

- Interference effects

- The (un)predictability of radio wave propagation

QUALITY OF SERVICE SPECIFICATIONS

The service requirement from the marketing should include

information on which the technical plan can be based ,

including :

Coverage Quality : Defined as a part of optimizing the

business plan (indoor / outdoor coverae, handheld car,

mobile set). Interference should be taken into account for

coverage quality including margin of 12 dB) :

Co channel C/I

Adjacent channel C/I

Call completion and Dropped call Rates : Dictated by the

lisence conditions and quality of the competing

network(includes Blocking rates 2% etc.)

Service availibility

QUALITY OF SERVICE SPECIFICATIONS

Traffic forecast:

- Longterm forecast and trends for the network must be

developed by the marketing.

- Traffic distributions for the existing coverage areas and

typical densities may be obtained from the network.

Spectral effeciencies : for demonstration within the context of

winning maximum points for a mobile license. The spectral

efficiency is determined by decisions taken in :

- Quality of coverage

- Frequency Reuse plan

- Use of cell splitting

- Design for traffic demend

- Feedback into the business plan

Customer support measures

DEFINITION OF COVERAGE QUALITY

Outdoor coverage:

- Default definitions of coverage

- Refers to 2 Watt class 4 mobiles in the street

- Probability of coverage is 95 % averaged across the cell area.

- Coverage probability at the edge of cells is less than this

value.

In car coverage :

- A supplementary level of coverage for highways

- Refers to a Class 4 mobile inside car or other vehicles.

- Coverage probability is nominally 95% averaged

- Coverage is critically dependent on the position of the

handheld mobile within the vehicle.

8 Watt Coverage

- A Supplementary level of coverage for remote areas.

- Refers to class 2 mobile or class 4 with an 8 watt booster and

external antenna

DEFINITION OF COVERAGE QUALITY

Indoor coverage

- Especially good coverage for city centers and stragetic

locations

- Refers to a class 2 mobile indoors

- Building loss is very variables, so indoor coverages can

never be guaranteed

- Where indoor coverage is provided , outdoor coverage will

be nearly 100 %

BLOCKING RATE ( Grade of Service, GOS )

GOS is defined as the probability that a call will be blocked or

delayed due to unavailability of the radio resource. Example

for license requirement

- 5 % Averaged over a defined sub-network (e.g. weighted

average by traffic load over the worse 10 cells )

- No cell to be worse than 10%

- By a particular date , 8 % of the cells permitted to be

between 2 % and 10 % GOS.

- By a particular date , 5 % of the cells permitted to be

between 2 % and 10 % GOS.

- Ultimate target is that no cells should be worse than 2 %

GOS.

CALL SUCCESS RATE

Call failure may be due to :

- Coverage holes

- Interference

- Congestion

- Problem in fixed network

- Handover failures

- Equipment failures

Call success rate is often expressed as the proportion of calls

connected and held for 2 min.

- Target is normally 90 % at launch of service

- Mature networks achieve in excess of 98 %

- Only applies within a declared coverage area.

By a particular date , 95 % of the calls to the network

boundary should be set up within four seconds and held for

two min.

RADIO PLANNING METHODOLOGY

Overall picture

It is important to create an overall picture of the network

before going into the detailed network planning. This is the

fact the main objective of this presentation.

Coverage Capacity and Quality

Providing coverage is usually considered as the most

important activity of a new cellular operator. For a while ,

every network is indeed coverage driven. However the

coverage is not the only thing. It provides the means of

service and should meet certain quality measures.

The starting point is a set of coverage quality

requirements.

To guarantee a good quality in both uplink and downlink

direction, the power levels of BTS and MS should be

balanced at the edge of the cell. Main output results of

the power link budget are:

- Maximum path loss that can be tolerated between MS

and the BTS.

- Maximum output power level of the BTS transmitter.


• A simple Planning Process Description

Business plan.

No of Subs.

Traffic per Subs.

Subs distribution

Grade of service.

Available spectrum.

Frequency Reuse.

Types of coverage

RF Parameters

Field strength studies

Available sites

Site survey

Capacity

Studies

Plan verification

Quality check

Update documents

Coverage

&C/I study

Search areas

Implement

Plan Monitor

Network

Optimize

Network

Customer

Acquires

sites

Capacity Studies

Coverage plan & Interference studies

Frequency plans and interference Studies

Antenna Systems

BSS parameter planning

Data base & documentation of approved sites

Expansion Plans.


Data Acquisition

OMC Statistics

“A” Interface

Drive Test

Implemented

Planning

Data

Data

Evaluation

Implemented

Recommendation

Recommendations :

Change frequency plan

Change antenna orientation/Down tilt

Change BSS Parameters

Dimension BSS Equipment

Add new cells for coverage

Interference reduction

Blocking reduction

Augment E1 links from MSC to PSTN

Cell Planning Aspects

At the end of it all, a good cell plan should have the following

characteristics :

Coverage as required as predicted.

Co Channel and Adjacent Channel interference levels as predicted.

Minimum antenna adjustments during the optimization process.

Minimum changes to the BSS parameters/database during the

optimization phase.

Should be well phased, requiring optimization only for short periods in

the initial commissioning phase and during

Facilitate easy expansion of the network with minimal changes in the

system.

The Basic Cell Planning Process

The basic approach to cell planning is to provide good coverage and capacity.

Initially, both are not known !!

Hence the planning is based on the projections given by the customer. The

customer based on market surveys and the company plans, may specify :

Number of sites he want in the city

OR

Number of subscribers expected in a city.

Base on the inputs from the customer, the initial planning process begins. From

these we can determine either the capacity that is possible for a given number

of sites OR minimum number of sites needed to provide service to a given

number of subscribers. The site density required for a specific capacity should

also pass the coverage criteria. This aspect will be covered later in the course.


What is the area of coverage needed ?

How many sites are required for this area ? ( cell radius of 1

Km. Means an approximate coverage area of 3 sq. Kms. )

Do we need so many sites ? Can some site be bigger ?

Decide number of sites based on capacity and coverage

requirements.

Divide city into clutter types such as .

>Urban

>Suburban

>Quasi Open

>Open

>Water.

Identify search areas covering all clutter types.

Customer selects a few sample sites.


Survey sites with reference to :

>Clutter heights

>Vegetation levels.

>Obstructions.

>Sector orientations

>Building strengths and other civil requirements

Prepare Power Budgets.

Conduct propagation tests

Calculate Coverage probabilities based on the drive test

results.

Verify Power budget sensitivityagainst drive test result ,

modify planning tools parameters.

Prepare final coverage maps.

.

A typical Power Budget

RF Link Budget UL DL

Transmitting End MS BTS

Tx Rf power output 33 dBm 43 dBm

Body Loss -3 dB 0 dB

Combiner Loss 0 dB 0 Db

Feeder Loss(@2 Db/100

M)

0 dB - 1.5 dB

Connector loss 0 dB - 2 Db

Tx antenna gain 0 dB 17.5 dB

EIRP 30 dBm 57 dBm

A typical Power Budget

RF Link Budget UL DL

Receiving End MS BTS

Rx sensitivity -107 dBm -102 dBm

Rx antenna gain 17.5 dBm 0 dB

Diversity gain 3 Db 0 dB

Connector Loss - 2 dB 0 dB

Feeder loss - 1.5 dB 0 dB

Interference degradation margin 3 dB 3 Db

Body loss 0 dB -3 dB

Duplexer loss 0 dB 0 dB

Rx Power -121 dBm -96 dBm

Fade margin 4 dB 4 Db

Reqd Isotropic Rx. Power -117 dBm -92 dBm

Maximum Permis. Path los 147 Db 149 dB

Summary

A good RF Planning ensures that the mobiles receive certain minimum

signal strength for specified percentage of time over a specified area of

coverage.

The MS receive signal strength depends on the path loss depends on the

path loss between the MS and the BTS.

The path loss in a mobile environment includes :

> Free space path loss

>Additional Loss due to Topography of the site ( clutter Factor )

>Confidence level required. (Probability of area coverage )

In general RF Planning means the understanding of :

> Propagation Models

> Coverage aspects

> Link Budgets ( Power Budgets)

> Antenna considerations

> Frequency planning and reuse aspects.

Urban Propagation Environment

This is the most common and yet unpredictable propagation environment for a mobile

system.

Building Penetration:

Building are responsible for the reflection and shadowing of signals. Trees and

foliages also contribute to shadowing as well as scattering of radio signals.

Attenuation of signals by building is measured by taking the difference between the

median signal level in front of the building and inside the bu9ilding. Obviously, the

building attenuation depends on the type of construction and the material used as well

as how big or small it is.

Typically the attenuation values may cause the signal levels to vary by – 40 to +80 Db

The negative value implies that the signal is attenuated and the positive values implies

that the increase in the signal level.

Windows and Doors in general give a good penetration of RF signals. Another

important factor is the angle of arrival of RF signals in to the building. Generally, a

building facing the BTS site has better penetration than the one that is side facing and

without windows.

The furniture used in the building also contributes to attenuation. Typically a

furnished building gives a loss of 2-3 dB more than an empty one.

Propagation Environment Some Typical values for Building Attenuation

Type of building Attenuation

in dBs

Farms, Wooden houses, Sport halls 0-3

Small offices,Parking lots,Independent houses,Small

apartment blocks

4-7

Row Houses, offices in containers, Offices,

Apartment blocks

8-11

Offices with large areas 12-15

Medium factories, workshops without roof tops

windows

16-19

Halls of metal, without windows 20-23

Shopping malls, ware houses, buildings with

metals/glass

24-27

Propagation Models

• Classical Propagation models :- • Log Distance propagation model

• Longley – Rice Model (Irregular terrain model )

• Okumara

• Hata

• Cost 231 – Hata (Similar to Hata, for 1500-2000 MHz band

• Walfisch Ikegami Cost 231

• Walfisch-Xia JTC

• XLOS (Motorola proprietary Model )

• Bullington

• Du path Loss Model

• Diffracting screens model

Propagation Models

• Important Propagation models :-

• Okumara Hata model (urban / suburban areas )( GSM

900 band )

• Cost 231 – Hata model (GSM 1800 band )

• Walfisch Ikegami Model (Dense Urban / Microcell

areas )

• XLOS (Motorola proprietary Model )

Okumara Hata Models

In the early 1960 , a Japanese scientist by name Okumara conducted

extensive propagation tests for mobile systems at different frequencies.

The test were conducted at 200, 453, 922, 1310, 1430 and 1920 Mhz.

The test were also conducted for different BTS and mobile antenna heights,

at each frequency, over varying distances between the BTS and the

mobile.

The Okumara tests were valid for :

• 150-2000 Mhz.

• 1-100 Kms.

• BTS heights of 30-200 m.

• MS antenna height, typically 1.5 m. (1-10 m.)

The results of Okumara tests were graphically represented and were not easy

for computer based analysis.

Hata took Okumaras data and derived a set of empirical equations to

calculate the path loss in various environments. He also suggested

correction factors to be used in Quasi open and suburban areas.

Hata Urban Propagation Model

The general path loss equation is given as :-

Lp = Q1+Q2Log(f) – 13.82 Log(Hbts) - a(Hm)+{44.9-6.55 Log(Hbts)}Log(d)+Q0

Lp = L0 +10r Log (d) path loss in dB

F = frequency in Mhz.

D = distance between BTS and the mobile (1-20 Kms.)

Hbts = Base station height in metres ( 30 to 100 m )

A(hm)={ 1.1log(f) - 0.7 } hm - {1.56log(f) - 0.8} for Urban areas and

= 3.2{log(11.75 hm)2 - 4.97 for dense urban areas.

Hm= mobile antenna height (1-10 m)

Q1 = 69.55 for frequencies from 150 to 1000 MHz.

= 46.3 for frequencies from 1500 to 2000 MHz.

Q2 = 26.16 for frequencies from 150 to 1000 MHz.

= 33.9 for frequencies from 1500 to 2000 MHz.

Q0 = 0 dB for Urban

= 3 dB for Dense Urban

Path Loss & Attenuation Slope

The path loss equation can be rewritten as :

Lp = L0 + { 44.9 – 6.55 + 26.16 log (f) – 13.83 log (hBTS)-a(Hm)

Where L0 is = [69.55 + 26.16 log (f) – 13.82 log ( HBTS ) – A (Hm)

Or more conveniently

Lp = L0 + 10 log(d)

is the SLOPE and is = {44.9 – 6.55 log(hBTS)}/10

Variation of base station height can be plotted as shown in the diagram.

We can say that Lp 10 log(d)

typically varies from 3.5 to 4 for urban environment.

When the environment is different, then we have to choose models fitting

the environment and calculate the path loss slope. This will be discussed

subsequently.

Non line of Sight Propagation

Here we assume that the BTS antenna is above roof level for any

building within the cell and that there is no line of sight between

the BTS and the mobile

We define the following parameters with reference to the diagram

shown in the next slide:

W the distance between street mobile and building

Hm mobile antenna height

hB BTS antenna height

Hr height of roof

hB difference between BTS height and roof top.

Hm difference between mobile height and the roof top.


• The total path loss is given by:

• Lp = LFS+LRFT+LMDB

• LFS= Free space loss = 32.44+20 log(f) + 20 log(d)

• Where,

• LFS = Free space loss.

• LRFT = Rooptop diffraction loss.

• LMDB = Multiple diffraction due to surrounding buildings.

• LRFT = -16.9 – 10 log(w) +10log(f) +20log(^Hm)+L(0)

Where

hm=hr-hm

L( ) = Losses due to elevation angle.

L( ) = -10 + 0.357 ( -00) for 0< <35

2.5 +0.075 ( -35) for 35< <55

4.0 +0.114 ( -55) for 55< <90


• The losses due to multiple diffraction and scattering components due to building are given by :

LMBD = k0 + ka +kd.log(d) +kf.log(f) – 9.log(w)

Where

K0 = - 18 log (1+ hB)

Ka = 54 – 0.8 ( hB)

Kd = 18 – 15 ( hB/hr)

Kf = - 4 +0.7 {f/925) – 1 } for suburban areas

Kf = - 4 +1.5 {f/925) – 1 } for urban areas

W= street width

hB= hB –hr

For simplified calculation we can assume ka = 54 and kd = 18

Choice of Propagation Model Environment Type Model

Dense Urban

Street Canyon propagation Walfish Ikegami,LOS

Non LOS Conditions, Micro cells COST231

Macro cells,antenna above rooftop Okumara-Hata

Urban

Urban Areas Walch-ikegami

Mix of Buildings of varying heights, vegetation,

and open areas.

Okumara-Hata

Sub urban

Business and residential,open areas. Okumara – Hata

Rural

Large open areas,fields,difficult terrain with

obstacles.

Okumara-Hata

Calculation of Mobile Sensitivity.

The Noise level at the Receiver side as follows:

NR = KTB

• Where,

• K is the Boltzmann’s constant = 1.38x10-20

(mW/Hz/0Kelvin)

• T is the receiver noise temperature in 0Kelvin

• B is the receiver bandwidth in Hz.

Signal Variations

Fade margin becomes necessary to account for the

unpredictable changes in RF signal levels at the receiver.

The mobile receive signal contains 2 components :

• A fast fading signal (short term fading )

• A slow fading signal (long term fading )

Probability Density Function

The study of radio signals involve actual measurement of signal levels at

various points and applying statistical methods to the available data.

A typical multipath signal is obtained by plotting the RSS for a number of

samples.

We divide the vertical scale in to 1 dB bin and count number of samples is

plotted against RF level . This is how the probability density function for the

receive signal is obtained.

However, instead of such elaborate plotting we can use a statistical expression

for the PDF of the RF signal given by :

P(y) = [1/2 ] e [ - ( - y – m )2 / 2 ( )2

Where y is the random variable (the measured RSS in this case ), m is the mean

value of the samples considered and y is the STANDARD DEVIATION

of the measured signal with reference to the mean .

The PDF obtained from the above is called a NORMAL curve or a Gaussian

Distribution. It is always symmetrical with reference to the mean level.


Plotting the PDF :

Plotting the PDF

-100

-80

-60

-40

-20

0

SAMPLES

RS

S

RSS

A PLOT OF RSS FOR A NUMBER OF SAMPLES


Plotting the PDF :

Plotting the PDF

Bin Numbers

P(x

)

ni/N

NORMAL DISTRIBUTION

P(x) = ni/N

Ni = number of RSS within

1 dB bin for a given level.


A PDF of random variable is given by :

P(y) = [ ½ ] e [ - (y-m)2 / 2( )2 ]

Where, y is the variable, m is the mean value and is the Standard

Deviation of the variable with reference to its mean value.

The normal distribution (also called the Gaussian Distribution ) is

symmetrical about the mean value.

A typical Gaussian PDF :


The normal Distribution depends on the value of Standard

Deviation

We get a different curve for each value of

The total area under the curve is UNITY

Calculation of Standard Deviation

If the mean of n samples is “m”, then the standard deviation is

given by:

= Square root of [{(x1-m)2 + …..+( xn-m)2 }/(n-1)]

Where n is the number of samples and m is the mean.

For our application we can re write the above equation as :

= Square root of [{RSS1-RSSMEAN)2+…..+(RSSN-

RSSMEAN)2/(N-1)}]

Confidence Intervals

The normal of the Gaussian distribution helps us to estimate the accuracy with which we can say that a measured value of the random variable would be within certain specified limits.

The total area under the Normal curve is treated as unity. Then for any value of the measured value of the variable, its probability can be expressed as a percentage.

In general, if m is mean value of the random variable within normal distribution and is the Standard Deviation, then,

The probability of occurrence of the sample within m and any value of x of the variable is given by :

P=

By setting (x-m)/ = z, we get,

P=

Confidence Intervals

The value of P is known as the Probability integral or the ERROR

FUNCTION

The limits (m n )are called the confidence intervals.

From the formula given above, the probability

P[(m- ) < z < (m+ )] = 68.26 % ; this means we are 68.34 % confident.

P[(m- ) < z < (m+ )] = 95.44 % ; this means we are 95.44 % confident

P[(m- ) < z < (m+ )] = 99.72 % ; this means we are 99.72 % confident.

This is basically the area under the Normal Curve.

The Concept of Normalized Standard Deviation

The probability that a particular sample lies within specified limits is given

by the equation :

P=

We define z = (x-m)/ as the Normalized Standard Deviation.

The probability P could be obtained from Standard Tables (available in

standard books on statistics ).

A sample portion of the statistical table is presented in the next slide..

Calculation of Fade Margin

To calculate the fade margin we need to know :

Propagation constant

>From formulae for the Model chosen

>Or from the drive test plots

Area probability :

>A design objective usually 90 %

Standard Deviation

>Calculated from the drive test results using statistical formulae or

>Assumed for different environments.

To use Jakes curves and tables.

Calculation of Edge Probability and Fade Margin

From the values of and we calculate :

Find edge probability from Jakes curves for a desired coverage probability,

against the value of on the x axis.

Use Jakes table to find out the correlation factor required –

Look for the column that has value closest to the edge probability and read

the correlation factor across the corresponding row.

Multiply by the correction factor to get the Fade Margin.

Add Fade Margin to the RSS calculated from the power budget

Significance Of Area and Edge Probabilities

Required RSS is – 85 dBm.

Suppose the desired coverage probability is 90 % and the edge probability

from the Jakes curves is 0,75

This means that the mobile would receive a signal that is better than – 85

dBm in 90 % of the area of the cell

At the edges of the cell, 75 % of the calls made would have this minimum

signal strength (RSS).

In Building Coverage

Recalculate Fade Margin.

>Involves separate propagation tests in buildings.

>Calculate and for the desired coverage ( say 75 % or 50% )

>Use Jakes Curves and tables to calculate Fade Margin.

>Often adequate data is not available for calculating the fade margin

accurately.

>Instead use typical values.

Typical values for building penetration loss :

Area 75 % coverage 50 % coverage

Central business area < 20 dB < 15 dB

Residential area < 15 dB < 12 dB

Industrial area < 12 dB < 10 dB

In Car 6 to 8 dB

Fuzzy Maths and Fuzzy Logic

The models that we studied so far are purely empirical.

The formulas we used do not all take care of all the possible environments.

Fuzzy logic could be useful for experienced planners in making right guesses.

We divide the environment into 5 categories viz., Free space, Rural, Suburban, urban, and dense urban.

We divide assign specific attenuation constant values to each categories , say

Fuzzy logic helps us to guess the right value for , the attenuation constant for an environment which is neither rural nor suburban nor urban but a mixture, with a strong resemblance to one of the major categories.

The following simple rules can be used :

Mixture of Free space and Rural :

Mixture of Rural and Suburban :

Mixture of Suburban and Urban :

Mixture of Urban and Dense urban :

Cell Planning and C/I Issues

The 2 major sources of interference are:

• Co Channel Interference.

• Adjacent Channel Interference.

The levels of these Interference are dependent on

• The cell radius ®

• The distance cells (D)

The minimum reuse distance (D) is given by :

D = ( 3N ) R

Where N= Reuse pattern

= i2 + i j + j2

Where I & j are integers.


R

D


Assuming the cells are of the same size .

All cells reansmit the same power.

The path loss is not free space and is governed by the

attenuation constant .

By geometry, for every cell there are 6 interfering cells in the

first layer.

The reuse distance Dand cell radius R are related to the c/I as

given below

(D/R) = 6 (C/I)

The C/I is in absolute value.


Co Channel Interference C/I for Omni Cells

D/R = 3N

C/I = 10 Log [ 1/m (D/R ) ], where m is the number of interferers.

M= 1 to 6 for the first layer of interfering cells.

Assuming = 3.5, m = 6 (worst case ), we calculate the theoretical

C/I available for various reuse plans as shown below :

N D/R = 3N C/I = 10 Log [ 1/6 (D/R) ]

3 3 8.917 dB

4 3.46 13.29 dB

7 4.58 21.80 dB

9 5.19 25.62 dB

12 6 29.99 dB


Adjacent Channel Interference :

Adjacent Chl Interference = - 10 Log [1/m (D/R) ]+

Where is the isolation offered by post modulation filters

Minimum value of is 26 dB , as per EIA standards.

If ( C/I ) for co channel interference is 10 dB, then for adjacent

channel interference it is 36 dB.

Frequency Planning Aspects

The primary objective of frequency planning is to ensure that,

given the limited RF spectrum, we achieve the required capacity

(traffic channels), keeping the interference within specified limits.

There are two types of frequency planning :

>Frequency planning based on Reuse patterns (manual)

>Frequency planning based on heuristic algorithm (automatic)

Manual planning is done by dividing the available frequencies in

to a number of frequency groups (as per a selected reuse pattern

) and assigning frequencies to various sectors / cells.

Suppose we have “n” frequencies . For a 3 cell repeat pattern with

3 sectors, we have 9 frequency groups, each group having n/9

frequencies.

The sectors are labeled A1,A2,A3,B1,B2,B3 and so on..

Assuming that an operator has 32 frequencies, say, from ARFCN 63

to 94, the frequencies could be grouped as shown in the table

opposite.


Say, for 32 frequencies (ARFCN 63 –94 ), for a 3*3 reuse

pattern, the frequencies are grouped as shown below

A1 A2 A3 B1 B2 B3 C1 C2 C3

63 64 65 66 67 68 69 70 71

72 73 74 75 76 77 78 79 80

81 82 83 84 85 86 87 88 89

90 91 92 93 94

OR

A1 B1 C1 A2 B2 C2 A3 B3 C3

63 64 65 66 67 68 69 70 71

72 73 74 75 76 77 78 79 80

81 82 83 84 85 86 87 88 89

90 91 92 93 94


The Frequency reuse could be done in either of 2 ways

mentioned in the tables in the previous slide :

Frequency Planning Aspects Directional reuse :

In a sectorised site, a group of channels (ARFCN) is

transmitted in the direction of antenna orientation , This is

based on tri cellular platform consisting of 3 identical cells

as shown in the diagram in the last slide.

Every cell is considered as an omni logically. The cells

are excited from the corners, separated by 1200

The axes of the diagram represent the 3 directions of

reuse. These are designated as { f(00)}, {f(1200)} and

{f(2400)}

Because we use directional antennas, the worst co

channel interference will be from only one interfering

station in the same direction

Frequency Planning Aspects We form a generic combination of the tricell pattern

using 7 such pattern, as shown in fig. Down. From this

we can see that each of three axes has three parallel

layers.

This result in a total of six or multiples of six frequency

GROUPS.

While assigning frequencies to individual calls we have

to take the directions of reuse into account.

Antenna Considerations

Uniform coverage in all cells

Alignment with hexagonal pattern

Space availability

Connectivity to BSC/MSC

Urban areas may have the following conditions :

Several sites may be needed.

Frequency reuse is unavoidable

In building penetration is must

Building act as RF shield and contain coverage.

Buildings reflect signals and provide coverage to

areas where LOS would have failed.

Such additional paths improve in building

penetration.

Antenna at a very high point may not meet in

building coverage requirements

Tackling Multipath Fading

In general we have the following methods to combat Multipath

fading:

In time domain Interleaving and coding

In Freq. Domain Frequency hopping

In spatial domain Space diversity

In the polarization domain Polarization diversity

The last two are related to Antenna Systems.

Diversity Antenna Systems

A diversity antenna System essentially has :

Two or more antenna

A combiner circuitry.

Signals A and B should have minimum correlation between them

typically the correlation coefficient <0.7


Antenna Spacings :

Separation D/ 900 Mhz 1800Mhz

Horizontal 10 3.3 m 1.7 m

Vertical 17 5.7 m 2.8 m

>Figures in the table are of minimum required separation

>If space is not a constraint, larger separation is always

recommended.

>Horizontal separation is preferred because it provides low

correlation values.

>However, horizontal separation suffers from angular

dependence (demonstrated in the diagram, next page ).

>Vertical separation does not suffer much from the angular

dependence.

>It also requires minimum supporting fixtures and does not

occupy a lot of space.

>But as the distance increases the correlation between the RF

signal at the antenna points increases rapidly, thereby negating

the very advantage of space diversity.


Space diversity can be achieved using:

3 antenna system

2 antenna system

The 3 antenna system provides very good spatial separation

between the two receive antenna and avoids the use of

duplexers. This reduces the risk of generating intermodulation

products.

The 2 antenna system is preferred where the space for the

antenna structure is limited or where the operators want to use

less number o antenna.


Advantages of dual polarization :

Reduced support structure for the antenna

Reduced weight

Slim towers and hence quicker construction and low cost.

Cost of one dual polarized antenna is generally lower than the

cost of two space diversity antenna.

Choice of Dual Polarized type

H/V type :

As most mobile are held at an angle 450, H/V is more likely to

cause balanced signals at the two branches.

The diversity performance is less dependent on the mobile

location

Slant type

Correlation between the two elements is angular dependent.

Unbalanced signals at the two arms of the receive antenna, since

one of the signal could be at the same angle as the mobile

General Antenna Specifications

Typical parameters of importance :

Polarization

Linear polarization :Evector contained in one plain

Horizontal polarization :H Vector parallel to the horizontal

plane

Vertical Polarization : E Vector parallel to the vertical plane

Circular / Elleptical Polarization

The extremity of the E or H field describes a circle or an ellipse in

the direction of propagation

Radiation pattern

This is a plot of electric field intensity as a function of direction from

the antenna, measured at the fixed distance.

General Antenna Specifications

When the main radiation lobe of the antenna is intentionally

adjusted above or below its plane of propagation, the result is

known as a beam tilt. When tilted downward, we get the Downtilt.

Down tilt can be done in two ways :

Electrical down tilt

Mechanical down tilt


Overall picture

It is important to create an overall picture of the network

before going into the detailed network planning. This is the

fact the main objective of this presentation.

Coverage Capacity and Quality

Providing coverage is usually considered as the most

important activity of a new cellular operator. For a while ,

every network is indeed coverage driven. However the

coverage is not the only thing. It provides the means of

service and should meet certain quality measures.

The starting point is a set of coverage quality

requirements.

To guarantee a good quality in both uplink and downlink

direction, the power levels of BTS and MS should be

balanced at the edge of the cell. Main output results of

the power link budget are:

- Maximum path loss that can be tolerated between MS

and the BTS.

- Maximum output power level of the BTS transmitter.


These values are calculated as a result of design

constraints.

- BTS and MS receiver sensitivity.

- MS output power level

- Antenna Gain

- Diversity reception

- Losses in combiners, cables etc.

The cell ranges are derived with propagation loss

formulas such as Okumara Hata or Walfisch Ikegami,

which are simply to use . Given a maximum path loss,

differences in the operating environment and the quality

targets will result in different cell ranges.

The traffic capacity requirement have to be combined

with the coverage requirements, by allocating

frequencies. This also may have impact on the cell

range.

COVERAGE PLANNING STRATEGIES

The selection of site configurations, antenna and cables in

the core of the coverage planning strategy. The right

choice will provide cost saving and guarantees smooth

network evolution.

Some typical configurations are :

- 3 sector site for (sub)urban areas

- 2 sector site for road coverage.

- Omni site for rural areas.

These are not the ultimate solutions, decisions should be

based on careful analysis.

Cell Range and Coverage Area :

For any site configurations, the cell ranges can be

determined given the equipment losses and gains. The site

coverage areas can be calculated then and these will lead

to required number of sites for a given coverage region. This

makes it possible to estimate the cost, eg. Per km2, to be

used for strategic decisions

After getting the overall picture, the actual detailed

radio network planning is done with a RNP tool.


- Marketing specifications

- Define design rules and parameters.

- Set performance targets.

- Design nominal cell plan.

- Implement cell plan.

- Produce frequency plan.

- Optimize network.

- Monitor performances.

METHODOLOGY EXPLAINED

Define design rules and parameters

- Identify design rules to meet coverage and capacity targets

efficiently

- Acquire software tools and databases

- Calibrate propagation models from measurements.

Set performance targets

- Clear statement of coverage requirements (rollout and

quality)

- Forecast traffic demand and distribution.

- Test business plan for different roll out scenarios and quality

levels.

Design nominal cell plan.

- Use computer tool to place sites to meet coverage an d

capacity targets.

- Verify feasibility of meeting service requirements

- Ensure a frequency plan can be made for the design.

- Estimate equipment requirement and cost.

- Develop implementation and resource plans (including

personal requirements)

- Radio plan will provide input to fixed network planning.


Implement Cell plan

- Identify physical site locations near to nominal or

theoretical locations, using search areas.

- Modify nominal design as theoretical sites are replaced

with physical sites

- Modify search areas in accordance with envolving

network.

Produce Frequency Plan

- Fixed Cluster configration, can be done manually.

- Flexible, based on interference matrix using an automatic

tool.


Optimize the network

- Campaign of measurements

- Analyze results

- Adjust network parameters such as : antenna directions,

handover parameters, and frequencies.

Expand the network

- In accordance with rollout requirements

- In accordance with forecast traffic levels

- To improve coverage quality.

- To maintain blocking performances.

RF Planning Process

1 Understand the Customers requirements

Coverage requirements

In building coverage experiments

Initial Roll out plans

Pre determined number of sites ?

2 Survey

Traffic Distribution and Pattern

Growth areas

High density business/ residential areas

Propagation tests for in building coverage estimates

and model calibrations

3. Prepare Planning Tool

Get Digitized maps

Load maps in the planning tool.

Use survey data and run the programme.

RF Planning Process

4. Draft Plan

Divide the city into number of regions-

Busy business areas

Areas that need excellent inbuilding coverage

areas

Use appropriate model and link budgets to

calculate the number of sites required per region.

5. Fine Tune plan.

Perform more with drive test, confirm plan

predictions.

Review plan with customer and fine tune the plan.

RF Planning Process

Understanding Customer Requirements :

What are the boundaries for the network ?

Are there any special pockets to be covered due to

Govt. requirements ?

What are the areas in which medium to average in

building coverage is acceptable ?

What are the areas where excellent in building

coverage is needed ?

Areas with high growth potential

Need colonies under development

High revenue areas

Shopping malls , offices complex, industrial estates

etc.

RF Planning Process

Initial Implementation Strategy :

High usage, high revenue users first ?

High end residential and business areas ?

Street coverage first ?

Special areas like 5 star hotel, commercial

building with fine in building coverage ?

High way coverage critical ?

Total coverage on day one ?

Number of sites more than the competition ?

Any Budget Limitations ?

Give an ideal plan to start with.

Let the customer cut corners.

Not an easy job !!

RF Planning Process

City Surveys :

Basically a scouting exercise

Looking for :-

Major traffic routes

Markets

Business Centres

Shopping malls

General customer behaviors

Telephone density

Congested areas with narrow lanes

Narrow water canals/lakes/ponds

General city layout

Prestigious residential areas.

VIP areas

Parks/ playground/open areas.

General Building types.. Multistoried, Row houses,

apartments, colonies etc.

Airport coverage

RF Planning Surveys In building Coverage Surveys :

Classify Buildings-

Hotel/restaurants

Commercial

Industrial

Residential

Shopping malls/markets

Propagation tests in a number of buildings in each variety.

Rf signal on road Vs. inside building gives building penetration loss.

Repeat tests in as many buildings as possible to get an estimate of

building loss for the area.

In building coverage affected mostly in ground floor/basement

Typical values (examples only) :

> Hotel restaurants 15 dB

> Commercial buildings 20 dB

> Shopping malls 15 dB

> Industrial Estates 12-15 dB

> Residential buildings 15-20 dB

> Old/Historical buildings 25-30 dB

RF Propagation Test Kits Battery powered Transmitter. 10 or 20 Watts output : frequency in

GSM 900/1800 Mhz.

Portable mast Adjustable upto 5 m. With 1 m

antenna on top, effective height

above ground is 6 m.

Transmit antenna High gain omni or directional antenna

as required

Receiver TEMS mobile Hand held mobile phone with RS232

connection to a laptop. Or an

accurate portable RF sensitivity meter

/ CW receiver if model calibration is

required.

Positioning system GPS system, with PCMCIA card

Computer Laptop PC with TEMS software and

GPS software

Cables accessories Calibrated cable lengths (10 m) of low

loss feeder with known attenuation

values; 12 Volts battery with

appropriate cable to connect to

transmitter.

Power meter, VSWR meter.

RF Planning Tool Planning Tool preparation and Model Calibration :

There are many planning tool available toaday :

PLANET (MSI)

Cell Cad (LCC)

Odessy (Aethos)

Asset (Aircom)

NetPlan (Motorola)

A planning tool Should be :

Easy to use

Compatible with tools like TEMS

Minimum hardware requirements.

Economical.

Maps collected from authorized sources.

1:50000 or 1:25000 scale

50 m resolution for macro

Less than 30 m resolution for Micro cell planning using “Ray tracing Tool”

Maps are digitized under 3 categories :

LandUse

Digital Terrain Map

Vectors (Roads, Railways, etc.)


Most Planning tools use corrections for the land use or clutter.

Propagation Model tuned by assigning the values to

Clutter factor (Gain or Loss due to clutter )

Clutter Heights (for diffraction modeling)

Different types of clutter are defined in these models/ tools

1. Dense Urban

2. Urban

3. Suburban

4. Suburban with Dense Vegetation

5. Rural

6. Industrial area

7. Utilities (marshalling yards, docks, container depots etc. )

8. Open area

9. Quasi Open Area

10. Forest

11. Water

Too many clutter type definitation complicate planning process 10 to 15 is

typical.


DTM

Provided by the map vendor

Provides contour information as a digital map.

Vectors

Highways

Main Roads

Railways

Canals / water ways.

Coast line

Rivers.

Each categories is digitized as different layer

Displayed separately if required

Map information is set up in the planning tool.

Model calibration carried out.

Model Calibration All tools have provision for manipulating clutter values.

Different tools have different directory structures and means of handling

geographical data.

The procedure mainly talks about ensuring correct data header files to

include.

BTS location

EIRP of BTS

Antenna Type

BTS antenna height

Description of surrounding area.

Procedure uses a general core model equation :

The equation has constant k1 to k6 and a constant of clutter, kclutter

Initial values for the constants are set as per the model chosen (say

Okumara Hata )

PLANET programme is run repeatedly to make RMS error values for

all data files ZERO or a minimum.

For each run of the programme, the values of k1 to k6 are manipulated.

This completes model calibration.

Link Budget and other Steps

Key Points To be Considered :

Coverage Probability

Expected inbuilding coverage

Edge probability

Fade margin required

Maximum permissible path loss ( from the link Budget )

What is the radius of the cell ?

Number of sites required (from coverage point of view )

Is the number of sites calculated as above adequate for capacity ?

Decide on more sites for capacity.

Capacity Calculations

Capacity calculations :

Check if number of sites is enough to give capacity.

Depends on

Spectrum available

This decides the site configuration.

Availability of features like frequency hopping etc.

If Capacity is not met, add more sites.

If number of site is not OK with the customer, then :-

Recalculate site density, for 50 % in building coverage in place of 75 %

Fine Tune The Plan

Use Planning tool to return Coverage predictions

Iterate the process in consultation with the customer.

Finalize Plan and document it.

Search Areas

Planner issues search areas for each site location with information on :

Location

Lat/Long

Antenna heights

Specific target areas if any

Size of search areas

Size acquisition team scouts for buildings.

3-5 alternatives preferred.

Site Selection Central Business area

Small Search areas (100 m)

Avoid near field obstruction.

Antenna at or slightly above the average clutter height.

Orientation is critical.

Try solid structure (lift room ) for antenna mounting.

This helps reduce back lobe radiation problems

Avoid towers on building tops. This reduces interference to neighboring cells.

Residential suburban areas : Larger search areas (200 m)

Location not very critical.

Antenna 3-5 meters above average clutter height.

Antenna orientation less critical.

Site Selection

Industrial area :

A suitable central location.

Avoid proximity to electrical installations like towers, transformers

etc.

Towers are common

Quasi / open Highways

Larger search areas (500 m)

Limited by terrain and not the clutter. Hilly areas need care.

Highways need closer search areas along road.

Tall sites give better coverage.

Extending Cell Range

Extended cell range reduces number of sites.

Cell range improvement achieved through :

BTS transmit power enhancement

BTS sensitivity enhancement

Combination of both


Increasing BTS transmit EIRP:

To maximize BTS O/P power, single carrier cells can be used.

This will avoid the combination losses of multiple carrier cells.

The output power at the top of the cabinet could be set to 40 Watt,

giving an increase in signal strength of 3 Db.

For cells with more than aone carrier, air combination can be

implemented so that the combination loss is minimized.

Another way to maximize Tx and Rx signals is to implement

lowloss feeder cable.

A typical 7/8 “ Andrewscoaxial cable has an attenuation of 3.92

dB/100 m. If a 5/8” Andrews cable with an attenuation of 2.16

dB/100 m is used, then an increase of 1.6 dB can be obtained per

100m.


Improving BTS receiver sensitivity :

Better devices in the BTS receiver.

Using Mast Head amplifiers with very low noise figures.

Better RF cables.


Improvement in the transmit side gives 2 dB advantage.

MHA’s extend the BTS receiver sensitivity to –110 dBm instead of the

usual –107 dBm.

Overall improvements result in 4-5 dB advantage in path loss, leading

to extended coverage.

This improves quality of coverage.

Experiments with MHA’s have shown improvements

In areas with 50 % probability to approximately 70 %

probability.


probability.


probability.


probability.

Documents

Rf-Planning