02 MN1790 Coverage Planning

8/9/2019 02 MN1790 Coverage Planning

1/99

MN 1790 2 - 1

©

T E C H C OM

C on s ul t i n g ®

Coverage Planning: ContentsCoverage Planning: Contents

Definition of Terms

• Characteristics of Radio Wave Propagation

• Radio Wave Propagation Models

• Suitable prediction models for Macro-, Micro- and Pico-cells

• Location Probability

• Link Budgets

• Fading

• Fast Fading

• Rice Fading

• Rayleigh Fading

• Slow Fading• Jake's Formula

• Interference Margin

• Noise Figure calculations

• Amplifier Noise


2/99

MN 1790 2 - 2

©

T E C H C OM


Coverage Planning: ContentsCoverage Planning: Contents

• Path Loss Balance

• Cell Coverage Calculation

• Basics about Digital Map Data

• Principles of Planning Tools and their usage

• Measurement Tools supporting Cell Planning

• Cell Types

• Omni versus Sector Cells

• Exercises


3/99

MN 1790 2 - 3

©

T E C H C OM


Definition of TermsDefinition of Terms

To achieve coverage in an area, the received signal strength in UL and DL must be above the so

called receiver sensitivity level:

Coverage: RX_LEV > (actual) receiver sensitivity level

No Coverage: RX_LEV < (actual) receiver sensitivity level

The minimum receiver sensitivity levels in UL and DL are defined in GSM 05.05:

- for normal BTS : -104 dBm

- for GSM 900 micro BTS M1 : -97 dBm



- for DCS 1800 micro BTS M1 : -102 dBm



- for GSM 900 small MS (class 4, 5): -102 dBm

- for other GSM 900 MS: -104 dBm

- for DCS 1800 class 1 or class 2 MS : -100 dBm

- for DCS 1800 class 3 MS : -102 dBm


4/99


5/99

MN 1790 2 - 5

©

T E C H C OM



Maximum output power (before combiner input) for normal BTS / TRX of different power classes:

2.5 – (


6/99

MN 1790 2 - 6

©

T E C H C OM



Maximum output power (per carrier, at antenna connector, after all stages of combining) for micro

BTS / TRX of different power classes:

>0.05 – 0.16 W>0.01 – 0.03 WM3

>0.16 – 0.5 W>0.03 – 0.08 WM2

>0.5 – 1.6 W>0.08 – 0.25 WM1

GSM 1800

micro-BTS

GSM 900

micro-BTS

TRX power class


7/99

MN 1790 2 - 7

©

T E C H C OM



The reference sensitivity performance as defined in GSM 05.05 for the GSM 900 system for

different channel types and different propagation conditions:


8/99

MN 1790 2 - 8

©

T E C H C OM


Characteristics of Radio Wave PropagationCharacteristics of Radio Wave Propagation

Physical Reasons

• Diffraction

• Reflection

• Scattering

• Absorption

• Doppler shift

Technical Problems

• Distance attenuation

(Path Loss)

• Fading

• Inter-symbol Interference

• Ducting

• Frequency shift /

broadening


9/99

MN 1790 2 - 9

©

T E C H C OM


Characteristics of Radio Wave PropagationCharacteristics of Radio Wave Propagation

Exercise:

Which physical phenomena is sketched in the following pictures?


10/99

MN 1790 2 - 10

©

T E C H C OM


Radio wave propagation:

The radio wave propagation is described by solutions of the Maxwell equations.

Exact solutions of the Maxwell equations are not accessible for real space environment with

obstacles which give rise to reflections and diffractions.

However, the full information provided by an exact solution (e.g. exact polarization and phase of

the field strength) is mostly not needed.

What is needed is the the received power level.

What a propagation model should provide is the attenuation of the power level due to the fact that

the signal propagates from the transmitter to the receiver.

Radio Wave Propagation Models


11/99

MN 1790 2 - 11

©

T E C H C OM


Empirical models and deterministic models:

Empirical models are based on measurements. Some empirical models (like the ITU model) are

curves derived from measurements. Others summarize the measurements in formulas (like the

Okumura Hata model) which fit the measured data.

Such models are very simple to handle but also usually rather imprecise. They are limited to

environments similar to the one where the measurements were performed.

Deterministic models are based on simplifying assumption for the general problem. This can be a

mathematical approximation of the original problem (like the finite difference model). Or it can be a

simple model for a special situation of the general problem (like the knife edge model).

Deterministic model can reach a very high precision, but they suffer from a very high complexity.

Semi empirical models are a combination of empirical models with deterministic models for special situations (like knife edge models).



12/99

MN 1790 2 - 12

©

T E C H C OM



Empirical models

Log distance path loss

ITU

Okumura Hata

COST HataDiffraction models

Epstein Peterson

Deygout

Giovanelli

Semi empirical models

Okumura Hata & knife edge

COST Hata & knife edge

COST Walfisch Ikegami

Deterministic models

Ray launching, ray tracing

Finite difference


13/99

MN 1790 2 - 13

©

T E C H C OM


Received power:

P T : Transmitted power

P R : Reveived power

nT R d

c P P ⋅=

)lg()lg()lg(lg d Ad nc L P

P

T

Rα −−=+−==

− 101010Path loss:

d: distance


n

T

Rd c

P

P −⋅=

0

0.2

0.4

0.6

0.8

1.0

2.5 5.0 7.5 10.0


14/99

MN 1790 2 - 14

©

T E C H C OM


0 . 0 0 0 1

0 . 0 0 1

0 . 0 1

0 . 1

1

1 2 5 1 0

n = 4n = 3n = 2

0

0 . 2

0 . 4

0 . 6

0 . 8

1 . 0

2 . 5 5 . 0 7 . 5 1 0 . 0

n = 4n = 3

n = 2

Received power level

as function of distance d

on linear scale.

n R d

P 1∝

Received power level

as function of distance d

on log scale.

n R

d P 1∝



15/99

MN 1790 2 - 15

©

T E C H C OM



2

4

⋅∝

d P

R

π

λ

Example: Free space propagation

? : wavelength in vacuum; , speed of light in vacuum

f: frequency in MHz

d: distance in km

The influence of the surface is neglected completely

f

c=λ

smc 81099792 ⋅= .

( ) ( )d f L lglg. 20204432 ++=


16/99

MN 1790 2 - 16

©

T E C H C OM



Example: 2 ray model

d 1

d 2a

d 2b

d

hBS

hMS

( ) ( )

( ) ( )

d

hhd d

d

hhd hhd d

d d d

d

hhd hhd d

MS BS

MS BS

MS BS

ba

MS BS

MS BS

2

2

2

12

2

22

2

222

2

22

1

=−

++≈++=

+=

−+≈−+=


17/99

MN 1790 2 - 17

©

T E C H C OM




⋅⋅

≈−

∝−−

d

hhk

d d

e

d

e P MS BS

ikd ikd

R

2

22

21

2

444

21

sinπ

λ

π

λ

( ) ( )

−−++=

d

hhk d f L MS BS sinlg.lglg. 2002620204432

( )d hh L MS BS

lg)lg()lg( 402020120 +−−=

d c

hh f

d

hhk

d

hhk hhk d

c

f k

MS BS MS BS MS BS

MS BS

π

π

2

2

=≈

⇒>>

=

sinfor large

f: frequency in MHz

d: distance in km

hBS

: height base station in m

hMS : height mobile station in m

The ground is assumed to be flat and perfectly reflecting.

The model is valid for hBS > 50m and d in the range of km or for LOS microcell channels

in urban areas.


18/99

MN 1790 2 - 18

©

T E C H C OM


80

100

120

140

1601 10 100

900MHz1800 MHz

path loss in dB

distance in km


hBS

= 50 m

hMS

= 1.5m



19/99

MN 1790 2 - 19

©

T E C H C OM



Log-distance path loss model:

n

R

d

d P

−

∝

0

+=

0

100

d

d n L L

d lg

d 0 : reference distance ca. 1km for macro cells or in the range of 1m -100m for micro cells;

should be always in the far field of the antenna

Ld0 : reference path loss; to be measured at the reference distance.

2-3Obstructed in factories

4-6Obstructed in building

1.6-1.8In building LOS

3-5Shadowed urban area

2.7-3.5Urban area

2Free space

Exponent nEnvironment


20/99

MN 1790 2 - 20

©

T E C H C OM



Okumura Hata model:

Based on empirical data measured by Okumura in 60’s Hata developed a formula with

correction terms for different environments.

The Okumura Hata model assumes a quasi flat surface, i.e. obstacles like buildings are not

explicitly taken into account. Thus the Okumura Hata model is isotropic. The different types of

surfaces (big cities, small cities, suburban and rural) are distinguished by different correction

factors in this model.

Parameter range for this model:

Frequency f= 150… 1500MHz

Height base stationh

BS

= 30… 200m

Height Mobile station hMS = 1… 10m

Distance d= 1… 20km


21/99

MN 1790 2 - 21

©

T E C H C OM


[ ]

[ ] [ ]

[ ]

−⋅

−−−=

−+−−−+=

974751123

805617011

556944821316265569

2

.).lg(.

.)lg(..)lg(.

)(

)lg()lg(..)()lg(.)lg(..

MS

MS

MS

BS MS BS urban

h

f h f

hd

d hchd h f L

small cities

big cities (f>400MHz )


Okumura Hata model:

f: frequency in MHz

d: distance in km

hBS : height base station in m

hMS

: height mobile station in m

( )[ ] 94403318784

4528

2

2

2

.)lg(.lg.

.lg

+−⋅=

+

⋅=

f f c

f c suburban areas

rural areas


22/99

MN 1790 2 - 22

©

T E C H C OM


≈−

≈+

=

⋅+−−=

00010

0020

223542126

.

.

)(

)lg(.)(.

MS

MS urban

hd

d chd L

small cities

big cities


Okumura Hata model:

For f= 900MHz, hBS = 30m, hMS = 1,5m the formula reads:

d: distance in km

5128

949

.

.

=

=

c

c suburban areas

rural areas


23/99

MN 1790 2 - 23

©

T E C H C OM



COST Hata model:

The Okumura Hata model cannot be applied directly to systems like GSM 1800/1900 or DECT.

Therefore it was extended to higher frequencies in the framework of the European research

cooperation COST (European Cooperation in the field of scientific and technical research).



Height base station hBS = 30… 200m


Distance d= 1… 20km

[ ]

[ ] [ ]805617011

5569448213933346

.)lg(..)lg(.)(

)lg()lg(..)()lg(.)lg(..

−−−=

−+−−−+=

f h f hd

d hchd h f L

MS MS

BS MS BS urban


24/99

MN 1790 2 - 24

©

T E C H C OM



COST Hata model:

suburban areas

rural areas

city center

The major difference between the Okumura Hata model is a modified dependence on

frequency and additional correction factor for inner city areas

For f= 1800MHz, hBS = 30m, hMS = 1,5m the correction term for the dependence on hMScan again be neglected. For the other terms of COST Hata model the insertion of the values

serves:

)lg(.. d c Lurban

⋅+−= 223524136

( )[ ] 94403318784

4528

2

3

2

2

.)lg(.lg.

.lg

+−⋅=

+

⋅=

−=

f f c

f c

c


25/99

MN 1790 2 - 25

©

T E C H C OM

C on s ul t i n g ®Both models, the Okumura Hata model and the COST Hata model can lead locally

to substantial deviation from the measured attenuation since these models are

isotropic. Local properties of the surface (big buildings, hills etc.) are not taken intoaccount.

9231

141

3

.

.

=

=

−=

c

c

c

COST Hata model:

suburban areas

rural areas

city center



26/99

MN 1790 2 - 26

©

T E C H C OM


ITU model:

The ITU (or CCIR) model was originally developed for radio broadcasting. It is based onmeasurements in the UHF and VHF range which are summarized in graphs

(ITU-R 370-7, ) for the field strength.

The different topographic situations are described by the parameters hBSeff and ∆h.

The ITU model describes the radio wave propagation for the ranges

f= 30... 250 MHz and 450... 1000MHz

d= 10... 1000km

Definition:

hBSeff is the antenna height above the mean elevation of the terrain measured in a range from 3km

to 15 km along the propagation path.

∆h is the mean irregularity of the terrain in the range from 10km to 50 km along the propagation

path, i.e. 90% of the terrain exceed the lower limit and 10% of the terrain exceed the upper limit of

the band defined by ∆h.

The curves for the field strength are given for different hBSeff and ∆h = 50m. The correction for

other values of ∆h is given in an additional graph.

Since local effects of the terrain are not taken into account the deviation between predicted and

actual median field strength may reach 20dB for rural areas. In urban areas this value may be well

exceeded.



27/99

MN 1790 2 - 27

©

T E C H C OM


ITU model:


hBSeff

∆h

3km 10km 15km 50km

90%

10%

0km


28/99

MN 1790 2 - 28

©

T E C H C OM


Correction to the ITU model: clearance angle method

An improvement of the ITU model is obtained by considering the maximum of the angle (clearance

angle) between the horizontal line and the elevations in the range of 0 to 16km along the

propagation path. The correction to the field strength ITU model (with ∆h=50m ) is give as graphs

for the clearance angle. The clearance angle correction applies to both the receiving and the

transmitting side.


16km

γ

MS, BS Position


29/99

MN 1790 2 - 29

©

T E C H C OM



COST Walfisch Ikegami model:

For a better accuracy in urban areas building height and street width have to be taken intoaccount, at least as statistical parameters. Based on the Walfisch Bertoni propagation model for

BS antennas place above the roof tops, the empirical COST Walfisch Ikegami model is a

generalisation including BS antennas placed below the roof tops.



Height base station hBS = 4… 50m


Distance d= 0.02… 5km

Further parameter:

Mean building height: ∆h in m

Mean street width: w in m

Mean building spacing: b in m

Mean angle between propagation path and street: ϕ in °


30/99

MN 1790 2 - 30

©

T E C H C OM


b w

dBS

MS

∆hhBS

hMS



ϕ

BS

MS


31/99

MN 1790 2 - 31

©

T E C H C OM



With LOS between BS and MS (base station antenna below roof top level):


)lg()lg(. d f L LOS

2620642 ++=

With non LOS:

++

=

,

,

0

0

L

L L L

L

msd rts

NLOS

0

0

≤+

>+

msd rts

msd rts

L L

L L

free space propagation:

rts L roof top to street diffraction and scatter loss:

⋅−

⋅+

⋅+−

+−∆++−−=

,..

,..

,.

)lg()lg()lg(.

ϕ

ϕ

ϕ

114004

075052

354010

201010916 MS rts

hh f w L

00

00

0

9055

5535

350


32/99

MN 1790 2 - 32

©

T E C H C OM




msd L multiscreen diffraction loss:

)lg()lg()lg( b f k d k k L L f d amsd msd

91

−⋅+⋅++=

hh BS

∆>

( )

−+−

−+−

=

∆∆−

⋅−

=

⋅∆−⋅−

∆−⋅−=

∆−+−

=

,.

,.

,

,

,.

)(.

),(.

,

,

),lg(

1925

704

1925

704

1518

18

508054

8054

54

0

118

1

f

f

k

h

hhk

d hh

hhk

hh L

f

BS d

BS

BS a

BS

msd

hh BS

∆≤

hh BS

∆>

hh BS

∆>

hh BS ∆≤

hh BS

∆≤

hh BS

∆≤ 50.>d

and

and

50.≤d

Medium sized cities and suburban centres

with moderate tree density

Metropolitan centres


33/99

MN 1790 2 - 33

©

T E C H C OM




Although designed for BS antennas placed below the mean building height the COST Walfisch

Ikegami model show often considerable inaccuracies.

This is especially true in cities with an irregular building pattern like in historical grown cities. Also

the model was designed for cities on a flat ground. Thus for a hilly surface the model is not

applicable.


34/99

MN 1790 2 - 34

©

T E C H C OM


Lee micro cell model:


This model is based on the assumption that the path loss is correlated with the total depth B of

the building blocks along the propagation path. This results in an extra contribution to the LOS

attenuation

)()( Bd L L LOS

α +=

)(d L LOS

)( Bα For both and can be read off graphs based on extensive measurements.

This model is not very precise and large errors occur in the following situation:

• When the prediction point is on the main street but there is no LOS path

• When the prediction point is in a side street on the same side of the main street as the BS.


35/99

MN 1790 2 - 35

©

T E C H C OM



Diffraction knife edge model:

Diffraction models apply for configurations were a large obstacle is in the propagation path and theobstacle is far away from the transmitter and the receiver, i.e.: and 21 d d h ,h

The obstacle is represented as an ideal conducting half plane (knife edge)

hMS

hBS

d 1

h

d 2

Huygens secondary source


36/99

MN 1790 2 - 36

©

T E C H C OM




Huygens principle: all points of a wavefront can be considered as a source for a secondary wavelet⇒sum up the contributions of all wavelets starting in the half plane above the obstacle

Phase differences have to be taken into account (constructive and destructive interferences)

Difference between the direct path and the diffracted path,

the excess path length

Phase difference: with Fresnel Kirchoff diffraction parameter.

Note: this derivation is also valid for

( )

21

21

2

2 d d

d d h +≈∆

2

2

2υ

π

λ

π ϕ =

∆=

( )

21

212

d d

d d h

λ υ

+⋅=

0


37/99

MN 1790 2 - 37

©

T E C H C OM




Diffraction loss:

−+−=

−= ∫

∞

ν

π ν du

ui i

E

E L

D

D

22

12020

2

0

explglg)(

0 E

D E

field strength obtained by free field propagation without diffraction (and ground effects).

diffracted field strength

Shadow border region:

+≈≈

)lg(.)(

υ

ν

20513

0

D L

,

,

0

0

>>


38/99

MN 1790 2 - 38

©

T E C H C OM




Fresnel Zone:

Condition for the nth Fresnel Zone:

d1 d2

r Fnl1 l2

22121λ ⋅=−−+ nd d l l

Fnr d d >>

21,

Fn

Fn

r

hn

nd d

d d r d d l l

2

22

1

21

212

2121

=⇒

⋅=

+≈−−+

ν

λ

The diffraction parameter ν can be rewritten with quantities describing the Fresnel zonegeometry.

For obstacles outside the 1st Fresnel zone:

For obstacles outside the 5th Fresnel zone:

dB L D

112 .)( ±=−


39/99

MN 1790 2 - 39

©

T E C H C OM



Diffraction multiple knife edge Epstein Petersen model:

The attenuation of several obstacles is computed obstacle by obstacle with the single knife edgemethod, i.e. first diffraction path: l 1l 2 , second diffraction path: l 2 l 3 .

The model is valid for . j i d h


40/99

MN 1790 2 - 40

©

T E C H C OM



Diffraction multiple knife edge Epstein Petersen model:

.

( )

21

21

11

2

d d

d d h

λ υ

+⋅=

)()(21

ν ν D D Dtotal

L L L +=

The Fresnel integral is replaced by an empirical approximation:

( )[ ]

+−+−+≈

≈

110102096

0

2

..lg.)(

υ υ ν

D L

..

,.

780

780

−≥

−<

ν

ν

This model is rather unprecise. The error grows with the number of obstacles.

( )

32

32

22

2

d d

d d h

λ υ

+⋅=


41/99

MN 1790 2 - 41

©

T E C H C OM



Diffraction multiple knife edge Deygout model:

This model is recursive. First the attenuation of the main obstacle is computed (in this example O1with the path l 1s1). In the second step the possible (main) obstacles along the paths to and from the

main obstacle are computed (here O2 with l 2 l 3 ). This procedure is continued until all obstacles are

taken into account.

d 1

h1

d 2

h2

d 2 d 3

l 1

l 2

l 3

s1

O 1

O 2

H 2


42/99

MN 1790 2 - 42

©

T E C H C OM



Diffraction multiple knife edge Deygout model:

.

( ))(

321

321

11

2

d d d

d d d h

+++

⋅=λ

υ

),()()(2121

OOC L L L D D Dtotal

−+= ν ν

( )

32

32

22

2

d d

d d h

λ υ

+⋅=

p

q

pOOC

2

21

1

22012

⋅

−−=

π α

lg),(

,

)(

arctan

++

= 313212

d d

d d d d α

( ),)(

321

321

1

2

d d d

d d d

h p +

++

⋅= λ ( )

)(123

321

2

2

d d d

d d d

H q +

++

⋅= λ

Correction term:

The correction term is chosen such that the result coincides in a good approximation with

an exact solution. After n steps this models may cover up to 2n-1 obstacles.


43/99

MN 1790 2 - 43

©

T E C H C OM



Diffraction multiple knife edge Giovanelli model:

Also the Giovanelli model is recursive. The recursion procedure is the same as for the Deygoutmodel. Instead of taking a correction term in the attenuation the receiver is considered at an

effective position at an height heff . .

d 1

h1

d 2

heff

d 2

d 3

l 1

l 2

l 3

O 2

O 1

H 1 H

2

effective

receiver positionh2


44/99

MN 1790 2 - 44

©

T E C H C OM



Diffraction multiple knife edge Giovanelli model:

.

( ))(

321

321

11

2

d d d

d d d h

+

++⋅⋅=

λ υ

)()(21

ν ν D D Dtotal

L L L +=

( )

32

32

22

2

d d

d d h

λ υ

+⋅=

The attenuation predicted by this model is between the values obtained from theEpstein Peterson model and the Deygout model without the correction term.

eff h

d d d

d hh

321

1

11

++

−= )(12

2

3

2 H H

d

d hh

eff −+=


45/99

MN 1790 2 - 45

©

T E C H C OM



Semi empirical models:

Semi empirical model combine deterministic models like knife edge models with empirical modelslike Okumura Hata or COST Hata.

The mentioned empirical models are only valid for a quasi flat surface. In combination with knife

edge models they can be extended to hilly surface or a mountain area.

The combination of empirical and deterministic models requires usually additional correction terms.

For the specific combination of models and their correction terms most user develop their own

solution which they calibrate with their measurements. .


46/99

MN 1790 2 - 46

©

T E C H C OM



Deterministic models:

Ray tracing and ray launching:

With the methods of geometrical optics all possible propagation paths from the transmitter to

the receiver are determined and summed up, i.e. there is a free space propagation from the

antenna to the first obstacle or from obstacle to obstacle and at the obstacle the ray is reflected or

diffracted until it reaches the antenna. The algorithm takes only rays with an adjustable maximum

number of reflections and diffractions.

With this method a very high precision for the prediction of the path loss can be obtained.

• For this method a digital map with high accuracy is required.

• For the reflection and diffraction attenuation factors have to be specified which depend on

the building surface (e.g. glass or brick wall).

• The algorithm is very complex and computer power consuming.

However, there are continuous improvements for hardware, software and algorithms.


47/99

MN 1790 2 - 47

©

T E C H C OM



Deterministic models:

Finite difference algorithm:

Since the solution to field equation are inaccessible the partial derivatives for the fields are

replaced by finite differences. This is obtained by introducing a grid and considering the the fields

only at the nodes of the grid. The derivatives become differences along the edges of the grid. The

partial differential equation becomes a linear equation system. However, the linear equation

system involves very large matrices for realistic problems to be treated with a sufficient precision.

With this method a very high precision for the prediction of the path loss can be obtained.

• For this method very precise surface data are required.

• The surface data have to be parameterised in an appropriate way for the grid.

However, as for the ray launching and ray tracing method, there are continuous improvements

for hardware, software and algorithms.


48/99

MN 1790 2 - 48

©

T E C H C OM


Summary of the application areas of the different models:

+00Finite difference

+++Ray launching ray tracing

-+-COST Walfisch Ikegami

-0+COST Hata & knife edge-0+Okumura Hata & knife edge

-++Giovanelli

-++Deygout

-++Epstein Peterson

-0+COST Hata

-0+Okumura Hata

--+ITU

+++Log-distance path loss

inhouseurbanruralPropagation model

Suitable prediction models for

Macro-, Micro-, and Pico- cellsSuitable prediction models for

Macro-, Micro-, and Pico- cells


49/99

MN 1790 2 - 49

©

T E C H C OM


Location ProbabilityLocation Probability

The propagation conditions of electromagnetic waves in real environments are not stable, but

location (and time) dependent fluctuations appear.

The radio network planner has to take this into account by working with probabilities, e.g. with the

following two coverage probabilities:

• Cell edge probability

• Cell area probability

Typical cell edge probabilities for:

Very good coverage: 95%

Good coverage: 90%

Acceptable coverage: 75%

As will be discussed later, these values correspond to the following cell area probabilities:

Very good coverage: 99%

Good coverage: 97%

Acceptable coverage: 91%


50/99

MN 1790 2 - 50

©

T E C H C OM


Link BudgetsLink Budgets

Before dimensioning the radio network, the link budget for different environments (indoor, outdoor,

in-car) must be considered.

From the link budget, the maximum allowable path loss can be derived.

Body LossBuilding (indoor)

penetration loss

Path Loss

(Fading) Margins

Diversity Gain,

Antenna Gain

Cable LossesBTS


51/99

MN 1790 2 - 51

©

T E C H C OM



MS

Maximum output power [dBm]

Feeder loss [dB]

Antenna gain [dBi]

EIRP [dBm]

Receiver sensitivity [dBm]

BTS

Rx-diversity gain [dB]

Antenna gain [dB]

Head amplifier gain [dB]

Jumper, feeder, connector losses [dB]

Duplexer losses [dB]

Receiver sensitivity [dBm]

Environment

Body loss [dB]

Building (indoor) penetration loss [dB]

Path loss [dB]

Fading margin (lognormal and Rayleigh) [dB]

Interference margin [dB]

Frequency hopping gain [dB]

Terms which enter the link budget:


52/99

MN 1790 2 - 52

©

T E C H C OM



Example of an UL link budget (GSM 900 MHz MS power class 4, BS with tower mounted amplifier,

frequency hopping on, receive diversity used):

UL

Link Budget

Outdoor MS

(Class 4)

Indoor MS

(Class 4)

Car mounted MS

(Class 2)

Units Remarks

MS Max. Output power 33 33 39 dBm

Feeder Loss 0 0 -2 dB

Antenna Gain 0 0 +2 dBi

Environment Body Loss

(900 / 1800) MHz

-5 / -3 -5 /-3 0 dB

Building (Indoor) penetration Loss 0 -18 0 dB

Path loss dB

Fading Margin: lognormal:

for 1sigma=10 and cell area probability=99%

-12 -12 -12 dB

Fading Margin: Rayleigh -3 -3 -3 dB

Interference Margin -2 -2 -2 dB

Frequency hopping gain +3 +3 +3 dB

BS Rx - diversity gain +3.5 +3.5 +3.5 dB

Antenna gain +17 +17 +17 dBi

Tower mounted amplifier gain +6 +6 +6 dB

Jumper + Feeder + Connector Losses -4 -4 -4 dB

Duplexer Losses -0.5 -0.5 -0.5 dB

Receiver Sensitivity -107 -107 -107 dB


53/99

MN 1790 2 - 53

©

T E C H C OM


Fading occurs on different scales due to different causes.

Fading appears statistically but different fading types obey different probability distributions.

Propagation models predict only the average value of the receive level.

An extra margin has to be added due the fading effect.

The common question for all fading effects is: how big to chose the margin such that the receive

level drops not below a given limit with a specified probability?

Fading


54/99

MN 1790 2 - 54

©

T E C H C OM


Fast Fading

Fast fading appears due to multi path propagation. The receive level is affected by interferences

due to different path lengths in the multi path propagation.

The field strength at the receiver is the vector sum of the fields corresponding to the different

propagation paths. Usually the fading is described by the probability function for the absolute value

of the field strength.

The generic situations:

Rice fading:

It exists a dominant path (usually the LOS path):


55/99

MN 1790 2 - 55

©

T E C H C OM


Rice Fading

Rice fading:

+−⋅

⋅=

N

R R

N

R R

N

R R

P

V V

P

V V I P

V V f

2

22

11

0 exp)(

RV

1 RV

0 I

+= ∑=

N

i R N i

V P 1

2

: received signal strength

: received signal from the dominant signal

: modified Bessel-Function of the first kind and zero order.

other noise sources : received power of the non dominant signals including other

noise sources like man made noise.

For the Rice distribution can be approximated by a Gauß distribution:12

1 >> N

R

P V

( )

−−⋅

⋅=

N

R R

N

R

P

V V

P V f

22

1 2

1exp)(π


56/99

MN 1790 2 - 56

©

T E C H C OM


Rice Fading

0

0.1

0.2

0.3

0.4

0 2 4 6 8 10

Absolute value of signal amplitude in V

Probability

Eample: Gaußean distributed signal for: V V R

51 =


57/99

MN 1790 2 - 57

©

T E C H C OM


Rayleigh Fading

Rayleigh fading is the other important special case of the Ricean fading. Rayleigh fading

describes the situation were there is no dominant path, i.e. a non LOS situation.

All contribution to the received signal are comparable in strength and arrive statistically distributed.

with : averaged field strength, and :

−=

2

2

22

R

R

R

R

R

V

V

V

V V f exp)(

RV

−=

0

0

0

0

1

P

P

P P f exp)(

2

0

2

1

RV P = averaged receive power:


58/99

MN 1790 2 - 58

©

T E C H C OM


0.001

0.01

0.1

1

-30 -20 -10 0 10 20

Power / averaged power in dB

Integrated probability for the power to be below a fading marging for

a Rayleigh distributionProbability

Rayleigh Fading


59/99

MN 1790 2 - 59

©

T E C H C OM


Fast Fading

All described types of fast fading have as characteristic length scale the wavelength of the signals.To combat Fast Fading:

⇒ Use frequency hopping

⇒ Use antenna diversity


60/99

MN 1790 2 - 60

©

T E C H C OM


Slow Fading

σ X d Ld L += )()(

Slow fading denote the variation of the local mean signal strength on a longer time scale.

The most important reason for this effect is the shadowing when a mobile moves around (e.g. in a

city).

Measurements have shown that the variation of the the mean receive level is a normal distribution

on a log scale ⇒ log normal fading.

The fading can be parameterized by adding a zero mean Gaussian distributed random variable .σ

X

Let Pm be a minimal receive level, what is the probability that the receive level is higher

than the minimal receive level, i.e. ?))(Pr( => m R

P d P

Pr

The σ has to be determined by measurements.

( )

−−⋅

⋅=

2

2

22

1

σ σ π σ

P P P X exp)(


61/99

MN 1790 2 - 61

©

T E C H C OM


Slow Fading

To compute the probability that the receive level exceeds a certain margin the Gaussian

distribution has to be integrated. This leads to the Q function:

)(1)(

21

2

1

2exp

2

1)(

2

z Q z Q

z erf dx

x z Q

z

−−=

−=

−⋅= ∫

∞

π


62/99

MN 1790 2 - 62

©

T E C H C OM


Slow Fading

0.001353.00.022752.00.158661.00.500000.0

0.000053.90.001872.90.028721.90.184060.9

0.000073.80.002562.80.035931.80.211860.8

0.000113.70.003472.70.044571.70.241960.7

0.000163.60.004662.60.054801.60.274250.6

0.000233.50.006212.50.066811.50.308540.5

0.000343.40.008202.40.080761.40.344580.4

0.000483.30.010722.30.096801.30.382090.3

0.000693.20.013902.20.115071.20.420740.2

0.000973.10.017862.10.135671.10.460170.1

Q(z)zQ(z)zQ(z)zQ(z)z

Tabulation of the Q function


63/99

MN 1790 2 - 63

©

T E C H C OM


Jake’s Formula

Jake’s formula gives a relation for the probability that a certain value Pm at the cell boundary at

radius R is exceeded and the corresponding probability for the whole cell. It is based on

the log distance path loss model:

+−=

0

0 lg10)()(d

d nd L P d P T R

−−

−+−=

22

11

21exp)(1

2

1)(Pr

b

aberf

b

abaerf P mcell

)(Pr mcell

P

( )σ 2

)( R P P a

Rm −= σ 2

)lg(10 enb =


64/99


65/99

MN 1790 2 - 65

©

T E C H C OM


Log-normal FadingLog-normal Fading

From measurements the standard deviation 1 sigma (σ LNF ) in a certain environment.

Typical measurement values (outdoor, indoor) are given in the following table:

9 dB

9 dB

8 dB

σ LNF(i)

10 dB

8 dB

6 dB

Dense urban

Urban

Rural

σ LNF(o)Environment


66/99

MN 1790 2 - 66

©

T E C H C OM



To achieve a certain cell edge probability σ LNF must be multiplied with a factor given in the

following table:

(Cell edge probability means the probability to have coverage at the border of the cell)

0.000

0.126

0.253

0.385

0.524

0.674

0.842

1.036

1.282

1.645

1.751

1.881

2.054

2.326

50

55

60

65

70

75

80

85

90

95

96

97

98

99

Factor for calculation of

lognormal fading margin

Cell edge probability in %


67/99

MN 1790 2 - 67

©

T E C H C OM



Integrating the Gaussian distribution function over the whole cell area delivers cell area

probabilities. Some example results are given in the following table:

77

91

97

99

50

75

90

95

Cell area probability in %Cell edge probability in %


68/99

MN 1790 2 - 68

©

T E C H C OM


Interference MarginInterference Margin

An interference margin can be introduced in the link budget in order to achieve accurate coverage

prediction in case that the system is busy.

This margin in principle depends on the traffic load, the cell area probability and the frequency

reuse. The required margin will be small if interference level decreasing concepts like frequency

hopping, power control and DTX are used.

Typically, a margin of 2 dB is recommended.


69/99

MN 1790 2 - 69

©

T E C H C OM


Noise Figure calculationsNoise Figure calculations

Thermal Noise:

Every object which is at a temperature T > 0°K emits electromagnetic waves

(thermal noise). Therefore, electromagnetic noise can be related to a temperature.

P = s * e* A * T4

Noise Factor:

The Noise Factor can be calculated from the Noise Temperature as follows:

Noise Factor = Noise Temperature / 290°K + 1

Noise Figure:

The noise figure is the value of the Noise Factor given in dB:

Noise Figure = 10 * log (Noise Factor)


70/99

MN 1790 2 - 70

©

T E C H C OM


Conversion table:

4384.02893.01702.0751.0

4223.92752.91591.9670.9

4063.82632.81491.8590.8

3903.72502.71391.7510.7

3743.62382.61291.6430.6

3593.52262.51201.5350.5

3443.42142.41101.4280.4

3303.32022.31011.3210.3

3163.21912.2921.2140.2

3023.11802.1841.170.1

Noise

Temp.

Noise

Figure

Noise

Temp.

Noise

Figure

Noise

Temp.

Noise

Figure

Noise

Temp.

Noise

Figure

Noise figure in dB

Noise Temperature in °K

Noise Figure calculationsNoise Figure calculations


71/99

MN 1790 2 - 71

©

T E C H C OM


Amplifier NoiseAmplifier Noise

Amplifier:

• An amplifier amplifies an input signal, as well as the noise of the input signal.

• It adds its own noise, which is also amplified.

GTin

Tnoise

G * Tin + G * Tnoise


72/99

MN 1790 2 - 72

©

T E C H C OM



Cascade of amplifiers:

G1Tin

Tn1

G1* Tin + G1 * Tn1

G2

Tn2

G2 * (G1 * Tin + G1 * Tn1) + G2 * Tn2

= G1*G2* (Tin + Tn1 + Tn2/G1)

= G * (Tin + Tnoise)

With Tnoise = Tn1 + Tn2/G1 and

G = G1 * G2

GTin

Tnoise


Equivalent to cascade of amplifiers


73/99

MN 1790 2 - 73

©

T E C H C OM



Friis formula:

Tnoise = Tn1 + Tn2 / G1 + Tn3 / (G1*G2) + ...

GTin

Tnoise


Equivalent to cascade of amplifiers

Tnoise = Tn1 + Tn2/G1

G = G1 * G2


74/99

MN 1790 2 - 74

©

T E C H C OM



Example:

G1Tin

Tn1

G1* Tin + G1 * Tn1

G2

Tn2

G1*G2* (Tin + Tnoise)

With

Tnoise = Tn1 + Tn2/G1

Assumptions:

G1 = 16 Tn1 = 28°KG2 = 20 Tn2 = 200°K

Result:

Gain = 320

Tnoise = 40.5°K

Assumptions:

G1 = 20 Tn1 = 200°KG2 = 16 Tn2 = 28°K

Result:

Gain = 320

Tnoise = 201.4°K

Consequence:

Position of amplifier in chain

is very important


75/99

MN 1790 2 - 75

©

T E C H C OM



Exercise 1:

Calculate the noise temperature of the following system:

G Tnoise ?

Antenna cable

Loss 10 dB

Amplifier in BTS

Gain 25 dB

Noise temperature 240°K


76/99

MN 1790 2 - 76

©

T E C H C OM



Exercise 2:

Calculate the noise temperature of the following system:

Tnoise ?

Cable to antenna mast

Loss 10 dB

G

Amplifier in BTS

Gain 2 dB


G

Mast Head Amplifier

Gain 28 dB



77/99

MN 1790 2 - 77

©

T E C H C OM


Path Loss BalancePath Loss Balance

Since the coverage range in UL should be the same as the coverage range in DL, the radio link

must be balanced:

Maximum allowable path loss in UL = Maximum allowable path loss in DL

Considering the link budget, usually the UL is the bottleneck, i.e. the maximum allowable path loss

is determined by the UL and not by the DL, although:

• The BS receiver sensitivity is usually better than the MS receiver sensitivity.

• Diversity is usually only used in the receive path.

In case of an unbalanced link with weak UL, the UL sensitivity and therefore also the UL coverage

range can be increased by using tower mounted amplifiers.


78/99

MN 1790 2 - 78

©

T E C H C OM


Cell Coverage CalculationCell Coverage Calculation

From consideration of link budget Maximum allowable path loss

Using radio wave propagation formulas (e.g.Hata) Maximum cell size

Exercise:

Consider a class 4 MS of height = 1.5 m. The BTS height = 30 m. Assume Hata

propagation conditions and a cell area probability of 97%. What is the maximum outdoor,

indoor cell radius and in-car cell radius:

a) In a dense urban environment (σ LNF,o= 10 dB; σ LNF,i= 9 dB )?

b) In a suburban environment (σ LNF,o= 8 dB; σ LNF,i= 9 dB)?

c) In an open area (σ LNF,o= 6 dB; σ LNF,i= 8 dB)?

Assume an in-car penetration loss of 6dB.


79/99

MN 1790 2 - 79

©

T E C H C OM


Basics about Digital Map DataBasics about Digital Map Data

The cell planning tools require as one input digital map data (which are often based on paper

maps, satellite photos,…). These digital map data should contain information about, the land

usage ( so called “Clutter” information), about the height of obstacles and they should also containso called vector data (like rivers, streets,…).

A digital map is an electronic database containing geographical information.

The smallest unit on such a map is called a pixel. The typical edge-length of such a pixel is

ranging from several meters to several hundred meters. A digital map is often subdivided into

several blocks consisting of many pixels. The different layers of information in one block always

use the same resolution, whereas different blocks can have different resolutions.

Each pixel should contain information about:

• Land usage (“Clutter” information)

• Height data

• Vector data (like rivers, streets,…)

•…

Before working with these digital data, some pre-processing of the data may be required. Some

ideas are sketched on the following pages.


80/99

MN 1790 2 - 80

©

T E C H C OM



Definition of terms

• Geoid• Spheroid / Ellipsoid

• Geodetic Datum / Map Datum / Datum

Projections

• Are used to transfer the 3 dimensional earth to a 2 dimensional map

• “Nobody is perfect”

• No projection is at the same time exact in area, exact in angle and exact in distance.


81/99

MN 1790 2 - 81

©

T E C H C OM


Geodetic datum — simplified mathematical representation of the size and shape of the earth

1. Local geodetic datum — best approximates the size and the shape of the particular part of

the earth

2. Geocentric datum — best approximates the size and shape of the earth as a whole

spheroidgeoid

The GPS uses a geocentric datum to express its position because of its global extent.



82/99

MN 1790 2 - 82

©

T E C H C OM


Two coordinates systems are implicitly associated with a geodetic datum:

a. Cartesian coordinate systemb. Geodetic (geographic) coordinate system

A third coordinate system is provided by a map projection.



84/99

MN 1790 2 - 84

©

T E C H C OM


Cylindrical projection — true at the equator and distortion increases toward the poles

1. Regular cylindrical projections

a. Equirectangular projection

b. Mercator projection

c. Lambert‘s cylindrical equal area

d. Gall‘s sterographic cylindrical

e. Miller cylindrical projection



85/99


87/99

MN 1790 2 - 87

©

T E C H C OM


Conic projections — true along some parallel somewhere between the equator and a pole and

distortion increases away from this standard

1. Lambert conformal conic

2. Bipolar oblique conic conformal

3. Albers equal-area conic

4. Lambert equal-area conic

5. Perspective conic

6. Polyconic

7. Rectangular polyconic

…



88/99

MN 1790 2 - 88

©

T E C H C OM


Azimuthal projections — true only at their centre point, but generally distortion is worst at the

edge of the map

1. The Gnomonic projection

2. The azimuthal equidistant projection

3. Lambert azimuthal equal-area

4. etc.



89/99

MN 1790 2 - 89

©

T E C H C OM


Compromise projection

1. Gall‘s projection

2. Miller projection

3. Robinson projection

4. Van der Grinten Projection



90/99

MN 1790 2 - 90

©

T E C H C OM


For transformation of parameters (Latitude and Longitude) from the 3 dimensional representation into

a 2 dimensional rectangular system often a combination of WGS-84 ellipsoid & UTM rectangular

coordinate system is used (like e.g. for GPS).UTM (Universal Transverse Mercator) system defines 2 dimensional positions using zone numbers

and zone characters for longitudinal and horizontal scaling:

UTM zone number (1-60):

longitudinal strips: range: 80° south latitude - 84° north latitude, width: 6 degree

UTM zone characters (using 20 characters, also called designators):

horizontal strips: range: 180° east - 180° west longitude, width: 8 degree



91/99

MN 1790 2 - 91

©

T E C H C OM


Hints concerning the usage of maps:

• Avoid in any case the referencing of geodetic co-ordinates to a wrong geodetic datum.Referencing to a wrong datum can result in position errors of several hundred meters! (In

the meantime people agreed to use in the future the World Geodetic System 1984

[WGS-84] for all maps.)

• Remember that e.g. different nations may use different geodetic datum.

• If a datum conversion is necessary a careful transformation of seven parameters is necessary:

3 for translation, 3 for rotation, 1 for scaling

• For daily work, try to use the same geodetic datum: in your planning tool(s), for your

GPS systems, and for your paper maps.

• Prefer the following map scales:

1:50000 (for rural areas and 900 MHz cell planning)

1:20000 (for rural areas and 1800/1900 MHz cell planning)

1:10000- 1:5000 (for urban areas and for micro cell planning)

In the maps, height information should be included as contour lines.



92/99

MN 1790 2 - 92

©

T E C H C OM


Principles of Planning Tools and their usagePrinciples of Planning Tools and their usage

Main Task of radio network planning tools:

• Coverage planning

• Capacity planning

• Frequency planning

• Link Budget calculations

• Propagation predictions

• Propagation model fine tuning

• Co- and adjacent channel interference analysis

• Macro, micro cell planning

• Handling of multi-layer structures

• Repeater system handling

• Microwave planning

• …


93/99

MN 1790 2 - 93

©

T E C H C OM



Remarks to radio network planning tools and required digital map data:

Tools using empirical propagation models require map data with less resolution compared to tools

working with deterministic propagation models.

In case empirical propagation models are used:

• Typical pixel size: 50m x 50m to 200m x 200m

• Using statistics, the signal variation around the mean value is taken into account

• In case that the BS antenna is higher then the surrounding, the clutter correction term of the

target pixel contain most propagation effects. For the clutter boundaries often several pixels

before the target pixel are taken into account.

In case deterministic propagation models are used:

• Digital data with high resolution are required (often very expensive)

• Typical pixel size: 2m x 2m to 10m x 10m

• Mostly used for big cities only


94/99

MN 1790 2 - 94

©

T E C H C OM



Remarks to tools and required computational time:

Depending not only on the hardware used but also on the algorithms behind the software,

the computational time required by different tools varies significantly.


95/99

MN 1790 2 - 95

©

T E C H C OM



Planning tools do not run fully automatically but always require some input and anintelligent and creative usage.

Remember:

Garbage in Garbage out


96/99

MN 1790 2 - 96

©

T E C H C OM


Measurement Tools supporting Cell PlanningMeasurement Tools supporting Cell Planning

Fine tuning (calibration) of propagation models:

Why? When? How?

• Since propagation models does not necessarily describe exactly the real situation, a fine tuning

of the models is necessary (e.g. clutter data may vary from country to country).

• This tool tuning should be done in the start phase of the network planning (i.e. before a detailed

plan is performed).

• A test transmitter is located at typical site locations, a test receiver measures the RX_LEV

along predefined measurement routes. These measured values are taken as input for the tool fine

tuning.


97/99


98/99

MN 1790 2 - 98

©

T E C H C OM


Omni versus Sector CellsOmni versus Sector Cells

Omni sites:

Advantages of omni sites:

• Trunking gain (especially interesting for those networks having only a few frequencies)

• Omni antennas are usually less bulky than sector antennas

• Suitable in those areas, where the surrounding terrain limits the coverage (before the

maximum omni cell radius is reached)

Disadvantages of omni sites:

• In case of horizontal antenna diversity: Diversity gain depends on direction

• Greater reuse distance required

• Less flexibility in network optimization (concerning antenna tilt, power control

parameters, handover parameters)

• TX/RX antenna separation difficult (usually TX/RX antennas are mounted on different

vertical levels to achieve sufficient separation)

• Limited mounting positions: no wall mounting possible


99/99

©

T E C H C OM


ExercisesExercises

1) Consider:

an extended cell with 100 km cell radius covering a sea area (clutter term: 30 dB),a 900 MHz mobile station of power class 4,

a BS with the GSM minimum receiver sensitivity,

an (BS) antenna gain of 15 dBi.

What should be the height of the BS antenna?

2) Consider:

a mobile station with 2 Watts output power maximum,

a BS receiver sensitivity of –104 dBm,

an (BS) antenna gain of 15 dBi.

For a satellite carrying the BS, what would be the maximum radius for the satellite orbit.

3) How many sites can be saved in principle if TMAs with 6 dB gain are used in the

network? Use typical values and Hata’s propagation formula for calculation.

Date post:	01-Jun-2018
Category:	Documents
Upload:	taringtrenggana
View:	219 times
Download:	0 times

02 MN1790 Coverage Planning

Documents