LECTURE 3a

MOBILE RADIO
PROPAGATION AND FADING
References:
Rappaport (Chapter 3 and 4)
Bernhard (Chapter 2)
Garg (Chapter 4)
RADIO WAVE PROPAGATION
Radio waves extends from a frequency of 30 kHz to 300 GHz
In free space, radio waves propagate in straight line (LOS) and are
reflected off objects. Radio waves on the earth are affected by the
terrain of the ground, the atmosphere and the natural and artificial
objects on the terrain.
There are 3 main propagation means on the earth:
Ground wave
Ionespheric or Sky wave
Trophospheric Wave
RADIO WAVE PROPAGATION
Ground Wave
travels in contact with earths surface
reflection, refraction and scattering by objects on the ground
transmitter and receiver need NOT see each other
affects all frequencies
at VHF or higher, provides more reliable propagation means
signal dies off rapidly as distance increases
Tropospheric Wave
bending(refraction) of wave in the lower atmosphere
VHF communication possible over a long distance
bending increases with frequency so higher frequency more chance
of propagation
More of an annoyance for VHF or UHF (cellular)
Ionospheric or Sky Wave
Reflected back to earth by ionospheric layer of the earth atmosphere
By repeated reflection, communication can be established over 1000s
of miles
Mainly at frequencies below 30MHz
More effective at times of high sunspot activity
EFFECT OF TRANSMISSION
Range
Transmission range: communication
possible, low error rate
Detection range: detection of the
signal possible, no communication
possible
Interference range: signal may
not be detected, signal adds
to the background noise
Region
Near-field (Fresnel)
The close-in region of an antenna wherein the angular field
distribution is dependent upon distance from the antenna
Far-field
The region where the angular field distribution is essentially
independent of distance from the source.
If the source has a maximum overall dimension D that is large
compared to the wavelength, the far-field region is commonly
taken to exist at distances greater than 2D
2
/ from the source
For a beam focused at infinity, the far-field region is sometimes
referred to as the Fraunhofer region
distance
sender
transmission
detection
interference
No effect
RADIO PROPAGATION MECHANISMS
Free Space propagation
Refraction
Conductors & Dielectric materials (refraction)
Diffraction
Radio path between transmitter and receiver obstructed by surface
with sharp irregular edges
Waves bend around the obstacle, even when LOS does not exist
Fresnel zones
Reflection
Propagating wave impinges on an object which is large compared to
wavelength
e.g., the surface of the Earth, buildings, walls, etc.
Scattering
Clutter is small relative to wavelength
Objects smaller than the wavelength of the propagating wave
e.g., foliage, street signs, lamp posts
diffraction
shadowing
Radio wave
scattering
Radio wave
reflection
Radio wave
FRIIS TRANSMISSION EQUATION
Power density at any distance, R, in the far field is the total power
transmitted divided by the area of the sphere of radius R
FSPL
Assumes far-field (Fraunhofer region) d >> D and d >> , where
D is the largest linear dimension of antenna
is the carrier wavelength
Suppose we have unobstructed line-of-sight (LOS), the Free Space
Propagation Loss (FSPL) is denoted by:
distance
frequency
) ( log 20 log 20 44 . 32
) (
4
log 20
km MHz

+ + =
(

=
d
f
dB d f
dB
d
FSPL

PEPL Plane Earth Propagation Loss

Path loss for flat reflecting surface
One LOS path and one ground (or reflected) bounce
Ground bounce approximately cancels LOS path above critical distance
PEPL is given by
PEPL
(m) receiver and ansmiter between tr distance
(m) height (MS) receiver
(m) height (BS) r transmitte
) ( ) log( 40 ) log( 20 ) log( 20
log 20
2

+ =
|
.
|

\
|
=
d
h
h
dB d h h
d
h h
PEPL
r
t
r t
r t
Diffraction occurs when waves hit the edge of an obstacle
- Secondary waves propagated into the shadowed region
- Excess path length results in a phase shift
- Fresnel zones relate phase shifts to the positions of obstacles
Model obstructions like hills, building use knife edge diffraction model
Fresnel-Kirchoff diffraction parameter
Single and multiple (Bullington, Millington, Deygout) knife-edge
Diffraction gain (loss) depends on v
DIFFRACTION MODEL
T
R
1st Fresnel zone
Obstruction
|
|
.
|

\
| +

+
=
+
=
2 1
2 1
2 1
2 1
2 1
2 1
) (
2 ) ( 2
d d
d d
h
where
d d
d d
d d
d d
h v

R T
d
1
d
2
h

DIFFRACTION MODEL
( )
4 . 2
225 . 0
log 20 ) (
4 . 2 1
) 1 . 0 38 . 0 ( 1184 . 0 4 . 0 log 20 ) (
1 0 )) 95 . 0 exp( 5 . 0 log( 20 ) (
0 1 ) 6 . 0 5 . 0 log( 20 ) (
1 0 ) (
2

|
.
|

\
|
=

=
=
=
=
v
v
dB G
v
v dB G
v v dB G
v v dB G
v dB G
d
d
d
d
d
FADING
Rapid fluctuation of the amplitude of a radio signal over a short
period of time or travel distance (sub-wavelength)
Cause by: multipath waves and Doppler shift
Large scale
mean signal attenuation versus distance
variation about the mean
Small scale
time spreading: flat fading and frequency selective fading
time variance of channel: fast fading and slow fading
MULTIPATH FADING
FADING
Two major components
Long term fading m(t)
Short term fading r(t)
Received signal s(t)
s(t) = m(t) r(t)
PATH LOSS AND FADING
FADING
Short term fading
Also known as fast fading caused by local multipath effect by NLOS
Observed over distance = wave length
30mph will experience several fast fades in a sec
Given by Rayleigh Distribution (Rayleigh fading)
The distribution can be formed using the square root of sum of the
square of two Gaussian functions
r = ( A
c
2
+ A
s
2
)
Ac and As are two amplitude components of the field intensity of the
signal
Long term fading
Long term variation in mean signal level is also known as slow fading
Caused by movement over large distances, shadowing effects and
wave diffraction around buildings, hills etc, moving receivers
experience slow variations of the signal level
The probability density function is given by a log-normal distribution
i.e.normal distribution on a log scale (log-normal shadowing)
A small deviation of the power level is advantageous for a good
transmitting quality. Typical values are 3 to 8 dB
RAYLEIGH and LOG-NORMAL
RICIAN AND RAYLEIGH DISTRIBUTION
PATH LOSS MODEL
Detail path loss model hard to factor in overall system design
Most important characteristic is power falloff with distance
Radio propagation models
Analytical models
Empirical models
Composite
Applications:
Predict large scale coverage for mobile systems
Estimate and predict SNR
PATH LOSS EXPONENT
Path loss is a function of
T-R distance d
Path loss exponent n
Standard deviation s
Estimation path model parameters from measured data by linear
regression
The estimation error probability is also available
Use path loss models for link budget design
Estimate the percentage of coverage area for a signal:
] ) ( y[ Probabilit > b P
r
LOG-DISTANCE MODEL
) log( 10 ) ( ] [
) ( ) (
exponent loss path
and distance ally with logrithmic
decreases power signal received average
model loss path distance - Log
0
0
0
d
d
n d P dB P
d
d
d P
n
d
L L
n
L
+ =

o
With fading (log-normal)
[dB] ) ( [dBm] dBm] )[ (
) ( ] dB )[ (
d P P d P
X d P d P
L t r
L L
=
+ =

PERCENTAGE AREA COVERAGE
| |
| |
) ( 1 ) (
2
1
2
1
2
exp
2
1
) (
) (
) (
) (
) (
2
z Q z Q
z
erf dx
x
z Q
d P
Q d P P
d P
Q d P P
z
r
r r
r
r r
=
(

|
.
|

\
|
=
|
|
.
|

\
|
=
|
|
.
|

\
|

= <
|
|
.
|

\
|

= >

Q-FUNCTION
z Q(z) z Q(z)
0.0 0.50000 2.0 0.02275
0.1 0.46017 2.1 0.01786
0.2 0.42074 2.2 0.01390
0.3 0.38209 2.3 0.01072
0.4 0.34458 2.4 0.08820
0.5 0.30854 2.5 0.00621
0.6 0.27425 2.6 0.00466
0.7 0.24196 2.7 0.00347
0.8 0.21186 2.8 0.00256
0.9 0.18406 2.9 0.00187
1.0 0.15866 3.0 0.00135
1.1 0.13567 3.1 0.00097
1.2 0.11507 3.2 0.00069
1.3 0.09680 3.3 0.00048
1.4 0.08076 3.4 0.00034
1.5 0.06681 3.5 0.00023
1.6 0.05480 3.6 0.00016
1.7 0.04457 3.7 0.00011
1.8 0.03593 3.8 0.00007
1.9 0.02872 3.9 0.00005
OUTDOOR PROPAGATION MODELS
Predict the signal strength at some point or local area
Consider also the terrain profile, e.g., mountains, trees,
buildings, obstacles.
Obtain models from systematic interpretation of measurement
data
Classifications:
Computer based models:
Longley-Rice model
Durkins model
Measurement model
Okumura model
Empirical model
Hata model
PCS extension and wideband PCS microcell models
Walfish and Bertoni Model
Computer-based models
Longley-Rice model
Model point-to-point propagation
Frequency band 40MHz-100GHz
Use Geometric optic techniques:
Two-ray ground reflection, knife edge refraction,scattering
Can use the terrain path profile if available
Can not add environment corrections, no multipath considerations
Measurement model
Okumura model
Most widely used model in urban areas
Obtained by extensive measurements
Represented by charts (curves) giving median attenuation relative to
free space attenuation
Valid under:
Frequency band: 150-1920 MHz
T-R distance: 1-10 km,
BS antenna height: 30-1000 m
Quasi-smooth terrain (urban & suburban areas)
OUTDOOR PROPAGATION MODELS
OKUMURA MODEL
Okumura model properties
Based completely on measurement, no analytical explanation and
in graphical form
based on extensive measurement in the Tokyo area at frequencies
from 150-1920 MHz. Valid for those frequencies and distence from
1 to 100 km
Model is valid for an urban environment over quasi-smooth terrain
Simple, but accurate for predicting path loss of cellular & land
mobiles. Practical standard for system planning
Okumuras model is very accurate in cluttered environments, but
responds slowly to rapid changes in terrain (as often seen in rural
areas)
Calculate path loss:
Determine free space loss
Look up table for median attenuation A
Add correction factors due to antennas and environments
OKUMURA MODEL
t environmen to due factor gain :
m 3 10 ), 3 / log( 20
m 3 ), 3 / log( 10
) (
m 30 1000 : /200) log( 20 ) (
distance and frequency
n with attenuatio median : ) , (
loss path space free :
loss path median :
) ( ) ( ) , ( dB] [
: Model Loss Path Okumura
3
50
50
A
r
r r
r
t t t
ma
F
A r t ma F
G
h h
h h
h G
h h h G
d f
d f A
L
L
G h G h G d f A L L

> >

=
> > =
+ =

OKUMURA MODEL
m 3 m 10
3
log 20 ) (
m 3
3
log 10 ) (
m 10 m 1000
200
log 20 ) (
> > |
.
|

\
|
=
|
.
|

\
|
=
> > |
.
|

\
|
=
re
re
re
re
re
re
te
te
te
h
h
h G
h
h
h G
h
h
h G
HATA MODEL
Empirical formulation to match Okumura model
Validity: f
c
= 150-1500MHz, h
t
= 30-200m, h
r
= 1-10m
Suitable for large cell, not for PCS microcells (<1km)
Note: f is in MHz, d is in km and antenna
heights are in meters
f
c
=1500MHz, h
r
=1.5m, h
t
=40m,
medium sized city
| |
| |
| |
| | dB 98 . 40 ) log( 33 . 18 ) log( 78 . 4 ) urban ( ) area rural open (
dB 4 . 5 ) 28 / log( 2 ) urban ( ) urban - sub (
city Large
300MHz dB 96 . 4 ) 175 . 1 log( 2 . 3
300MHz dB 1 . 1 ) 54 . 1 log( 29 . 8
) (
citiy sized medium to Small
dB ) 8 . 0 log 56 . 1 ( ) 7 . 0 log 1 . 1 ( ) (
dB log ) log 55 . 6 9 . 44 (
) ( log 82 . 13 log 16 . 26 55 . 69 urban] [
2
2
2
2
=
=

=
=
+
+ =
c c p p
c p p
r
r
r
c r c r
t
r t c p
f f L L
f L L
f h
f h
h a
f h f h a
d h
h a h f L
COST231 - HATA MODEL
Hata Model - PCS Extension
Setup by EURO-COST: COST-231 committee
Valid for
1.5-2GHz PCS systems
base station height, ht = 30 - 200 m
Mobile height, hr = 1 - 10 m
Distance, d = 1 - 20 km
Environment: Urban areas
| | | |
centres an metropolit for dB 3
density tree moderate with centres
suburban and city sized medium for dB 0
8 . 0 ) log( 56 . 1 7 . 0 ) log( 1 . 1 ) a(
where
log ) log 55 . 6 9 . 44 (
) ( log 82 . 13 log 9 . 33 3 . 46
r
=
=
=
+ +
+ =
M
r
M t
r t c p
C
f h f h
C d h
h a h f L
OTHER MODELS
Walfisch-Ikegami Model
Valid between 800 and 2,000 MHz and over distances of 20 m to 5 km
Useful for dense urban canyon-style environments where antenna
height is lower than the average building height Signals are guided
along the street, like an urban canyon
The Walfisch-Ikegami Model includes a diffraction constant and the
street width
Walfish & Bertoni model:
Consider the impact of rooftops and building height
Considered in IMT-2000 evaluation
Wideband PCS microcell model
Measurement in the microcells
Results:
Two-ray ground reflection model is good for LOS microcells
Log-distance path loss model is good for OBS microcells
Ibrahim and Parsons model - equations developed to best fit data
observed at London. (freq. 168-900 MHz)
Lees model - Use at 900MHZ with 3 parameters (median trasmission
loss, slope of the path loss curve and adjustment factor)
INDOOR PROPAGATION MODELS
Predict the pass loss inside buildings
Investigated only recently with the development of PCS
Indoor environments: small distance, great variability due to
layout, materials, type, etc.
Practical variations: door open/close, antenna positions, not far-
field
Model classification: LOS, OBS
Partition loss in the same floor
Hard partition (non-movable)
Soft partition (movable)
Path loss differs with partitions
Partition loss between floors
Floor attenuation factor
Log-distance path loss model applicablesame as log-normal
shadowing model
Thermal noise occurs in both media and communications equipment
due to random electron motion and is characterised by a uniform
distribution of energy over the frequency spectrum with a Gaussian
distribution of levels
Amount of thermal noise in B Hz of bandwidth:
P
n
= kT
e
B (W)
k is Boltzmanns constant 1.3803x 10
-23
J/K
T
e
is effective noise temperature of the device (K)
B is Bandwidth (Hz)
where T
e
=T
o
(N
f
-1) (N
f
- Noise Figure of the receiver and T
o
= 290K)
To include the effect of noise seen by the antenna and the feed line
between the antenna and the receiver, then:
P
n
= k(T
ant
+ T
e
+ T
fl
)B = kTB
Thermal noise sets the lower limit of sensitivity of a receiving system
Example, MS receiver noise figure, N
f
= 5 dB = 3.162
MS receiver antenna temperature = 290K
MS receiver bandwidth = 1.25 MHz
(Neglect any feed loss between antenna and receiver)
P
n
= kTB = 1.3803x10
-23
x[290(3.1621)+290]x1.25x10
6
= -138 dBW
NOISE LIMITED CHANNEL
Receiver sensistivity
The minimum available signal power at the input terminal of the receiver
such that there is an adequate SNR at the output of the receiver
e.g. -83 dBm (DECT), -102 dBm (GSM)
Noise floor
P
n
= kTB
e.g. 111.5 dBm (DECT), -120 dBm (GSM)
Signal-Noise ration
SNR = P
r
/P
n
Example
NOISE LIMITED CHANNEL
LINK BUDGET
A link budget tells us the maximum allowable path loss on each
link, and which link is the limiting factor. This maximum allowable
path loss will set our maximum cell size
Instead of solving for propagation loss in the prediction equations,
we can take the maximum allowable loss from the link budget and
calculate the cell radius, d, from the propagation model
The link budget is simply a balance sheet of all the gains and
losses on a transmission path
The link budget usually includes a number of product gains/losses
and margins
Product Parameters in the Link Budget
Transmit Power (EIRP): 30-45 dBm for base stations and 0-30
dBm for mobiles
Antenna Gain ( 18 dBi for base stations) A measure of the antennas
ability to increase the signal (usually by narrowing the beam)
Diversity Gain ( 3-5 dB) By utilizing different time, space, or frequency,
the system can tolerate a weaker signal. This translates to a gain
Receive Sensitivity ( -102 to -110 dBm) The lowest signal a receiver
can receive and still be able to demodulate with acceptable quality
LINK BUDGET
Duplexer Loss ( 1 dB) The loss from using the piece of equipment which
duplexes the uplink and downlink
Filter Loss ( 2-3 dB) The loss from using a filter to make the transmit or
receive signals cleaner
Combiner Loss ( 3 dB) The loss from using the piece of equipment that
can combine multiple frequencies onto one antenna system
Feeder Loss ( 3 dB) The loss from the cables connecting the base station
with the antenna system
Body Loss ( 3 dB) Accounts for the signal blockage created by a mobile
users head (sometimes called head loss)
Fade Margin ( 4-10 dB) Accounts for multipath fading dips for slow
moving mobiles
Typical Margins in the Link Budget
Interference Margin ( 1 dB) Accounts for high interference during busy
hour. Depends on frequency reuse plan, traffic load, etc.
Vehicle Penetration ( 6 dB) Accounts for the attenuation of signal by the
frame of a car
Building Penetration ( 5-20 dB) Accounts for the penetration of building
material for indoor coverage. Depends on the type of building and the desired
quality at the center of the interior
LINK BUDGET
COVERAGE PREDICTION