Terrestrial Propagation:

Large Scale Statistical Models

1
Statistical Models
• Generally, a signal undergoes multiple reflections and
diffractions during propagation

• Therefore, many different attenuated and phase-rotated

versions of the signal are received at the receiver

• With so many different rays, it is difficult to characterize the

power of the received signal using physical propagation
models

• Alternatively, a statistical approach is used in which the signal

characteristics are measured at different points and then a
statistical model is fit to match the measurements
2
Statistical Models
• In a statistical model, signal characteristics are measured
at different points and then a statistical model is fit to
match the measurements

• Generally two signal components are measured:

• Median Path Loss: Average signal attenuation based on the
distance between a sender and receiver

• Local Variations: Deviation around the median path loss

3
Median Path Loss

• Median path loss can also be represented in decibels as:

d
Lp 0
(dB) 10n log
d0
Where β0 is the attenuation at reference distance d0
n is called the path loss exponent which varies between 2 and 5
depending on the environment

4
Log-Normal Shadow Fading
• Large variations around the median are generally
observed in urban environments

• The variations are modelled as a Gaussian distribution:

d
Lp 0
10n log X
d0
where Xσ is zero-mean Gaussian distributed random
variable with standard deviation σ
All quantities in the above expression are in decibels
• This path loss model is called the lognormal shadow
fading model

5
Median Path Loss
d
Lp 0
(dB) 10n log
d0

Environment Path loss exponent, n

Free Space 2
Flat Rural 3
Rolling Rural 3.5
Suburban, Low Rise 4
Dense urban, Skyscrapers 4.5

6
Log-Normal Shadow Fading
• The variations around the median are modeled as a
Gaussian distribution
x dB
1 2 2
fX (x dB ) e dB

2 dB
All quantities in the above expression are in decibels.

-- It is a function which tells the probability of a number

in some context falling between any two real numbers.

7
Local Variations

x dB dB
1 2 2
fLp x dB e dB

2 dB

• Loss relative to median path is greater than 10dB for only

10% of the times with standard deviation of 8dB.
8
Exercise- Availability
• A measurement campaign in a large city indicates that the
propagation can be reasonably well modeled with a loss exponent of
n= 2.9. The shadowing deviation about this loss is 6 dB. What is the
range of coverage if 99% availability is required for a public- safety
radio application? Assume that the receiver sensitivity is -100 dBm
and the required power at 10 meters is 2mW.

9
Example: Availability
• A service is said to be X% available if the received signal
strength is above the receiver sensitivity more than X%
of the times

• What is the range of coverage under the following

conditions:
• 99% availability
• Path loss exponent, n = 2.9
• Shadowing deviation, σ = 6dB
• Receiver sensitivity = -100dBm
• Measured power at 10m is 2 mWatts (3.01dBm)

10
Example: Availability

14dB

99% availability

11
Copyright Pearson Education Inc. ©
Example: Availability
r
PR PT Lp PT dBm 0
dBm 10n log
r0
r
10n log 3.01 dBm PR
r0
3.01 dBm 100 dBm 14 dBm
r
10n log 89.01 dB
10
r
10 log 30.69 dB
10
10 log r 40.69 dB
r 11.72 km
12
Problem
• In previous example, the estimates were based on median path loss.
• In the same city, the deviation about the median path loss was
estimated to be 8 dB. Assuming a log normal model, how much
additional power must be transmitted to cover the same service area
with 90% availability at the edge of coverage when local shadowing is
taken into account.

13
Problem
• A brief measurement campaign indicates that the median
propagation loss at 420 MHz in a midsize North American city can be
modeled with n=2.8 and a fixed loss of 25 dB. Assuming a cell phone
receiver sensitivity of -95 dBm, What transmitter power is required to
service a circular area of radius 10 Km.
• Suppose that measurements were optimistic and n=3.1 is more
appropriate. What is the corresponding increase in transmit power
that would be required??

• Conclude your findings by looking at the difference???

14
Outdoor Propagation Models

• Radio Transmission often takes place over irregular terrain.

• The terrain profile of a particular area needs to taken into account for
estimating the path loss.

• Terrain profile varies from a simple curved to highly mountainous

profile. In addition, the presence of trees, buildings and other
obstacles must also be taken into account.

15
Outdoor Propagation Models

• A number of propagation models are available to predict path loss

over irregular terrain.

• While all these models aim to predict signal strength at a particular

receiving point or in a specific local area called a sector, the methods
vary widely in their approach, complexity and accuracy.

16
Outdoor Propagation Models

• Some of the commonly used outdoor propagation models are:

• Longley-Rice Model: The model is applicable to point to point

communication systems in the frequency range from 40Mhz to 100
GHz, over different kind of terrain.

o The median transmission loss is predicated using the path geometry of

the terrain profile and the refractivity of the troposphere.

17
Outdoor Propagation Models

o Signal Strength within a radio horizon is predicted using a two ray ground
reflection model.

o Diffraction losses over isolated obstacles are estimated using the Fresnel
Kirchhoff knife edge models.

• The model is also available as a computer program to calculate the

large scale median transmission loss for frequencies of 20MHz to
10GHz.

18
Outdoor Propagation Models

• The model operates in two modes:

• When a detailed path profile is available, the path specific parameters can
easily be determined and the prediction is called a point to point mode
prediction.

• If detailed path profile is not available, the method assumes different

techniques to estimate the path specific parameters, and such a
prediction is called an area mode prediction.

19
Outdoor Propagation Models
• Durkin’s Model: The Durkin’s path loss simulator consists of two
parts:

• The first part assumes that the receiving antenna receives all its
energy along the radial and, therefore, experiences no multipath
propagation while having the information about the proposed service
area.

• The second part of the simulation, calculates the expected path loss
around the radial.

20
Outdoor Propagation Models
• Okumura Model: Used for signal prediction in urban
areas and is applicable for frequencies in the range from
150-1920MHz and distances up to 100km.

• To determine path loss, the free space path loss between

the points of interest is first determined by using the
following formula:
L 50(dB) = LF +Amu – G (hte)-G (hre)-Garea

Where L50 is the 50th percentile (median) value of

propagation path loss, LF is the free space propagation
loss, A is the median attenuation relative to free space and
the two gains are related to transmitter and receiver, while
area gain is due to the type of environment. 21
Outdoor Propagation Models

• For Okumura mode, the transmitting gain is less than 1000m and is
greater than 3m, and is given by

• G(hte)= 20 log (hte/200)

• G(hre)= 10 log (hre/3)

22
Outdoor Propagation Models: Exercise

• Find the median path using okumura’s model for d=50km, hte=100m,
hre=10m and Amu= 43dB in a suburban environment. If the base
station transmitter radiates and EIRP of 1KW at a carrier frequency of
900Mhz, find the power at the receiver (assuming a unity gain
receiving antenna and area gain of 9dB).

• Another model known as HATA model is the correction of the

Okumura Model.

23
Indoor Propagation Models

• There is a great deal of interest in characterizing Radio Propagation

inside buildings. The Indoor radio channels differs from the
traditional mobile channel in two aspects:

• The distances covered are much smaller.

• The variability in the environment is much greater for a much smaller

range of T-R separation distance.

24
Indoor Propagation Models

• Indoor propagation is dominated by the same mechanisms as

outdoor; reflection, diffraction and scattering. However, the
conditions are much variable depending on whether interior doors
are open or closed inside a building.

• Antennas mounted at desk level in a partitioned office receive vastly

different signals than those mounted on the ceiling.

25
Indoor Propagation Models

• Indoor channels may be classified as LOS or OBS. Some of the key models
are:

• Partition Losses: Buildings have a wide variety of partitions and obstacles

which form the internal and external structure. In general, they are
classified as hard partition and soft partition. Different loss measured
during obstruction of common building material are given in table 4.3
(Rappaport).

• Log distance path loss model has been shown to be quite accurate in
modeling indoor path loss….

26
Indoor Propagation Models
Signal penetration into Buildings: RF penetration has been found to
be a function of frequency as well as height within the building.

• At lower floors of a building, the urban clutter induces greater

attenuation and reduces the level of penetration.

• Ericsson multiple breakpoint model and attenuation factor model are

some other models used to model indoor propagation.

27
Example 2.5-Indoor Propagation
• Suppose, in an office building, a 2.4 GHz transmitter located at a work
station is separated from the network access node (receiver) by a
distance of 35m. The transmission must pass through 5m of an office,
through a plaster board wall and then through an open area. The
propagation is modelled as a free space for the first 5m and with a
loss exponent of 3.1 for the remainder of the distance. The plaster
board wall causes 6 dB attenuation of the signal. The Isotropic
transmitter radiates 20 dBm. Can the link be closed if the receiver has
a sensitivity of -75 dBm.

28
Problem 2.7
• Using the same model as in Example 2.5, predict the path loss for the
site geometry shown in Figure 2.12. Assume that the walls cause an
attenuation of 5dB and the floors 10dB.

03 m R

Open area of 20 m
T 04 m

29
Link Calculations
• In designing a system for reliable communications, we must perform
link budget calculation to ensure that sufficient power is available at
the receiver to close the link and meet the SNR requirement.

• Recall that the phrase “Closing the link” refers to the requirement
that the received power is greater than the receiver sensitivity.
PR (dB) PT (dB) GT (dB) GR (dB) LP (dB)

• Right hand side provides enough power at the receiver to detect the
transmitted information.

30
Free Space Link Budget
• We can also represent the link budget equation as
PT
PR GTGR
Lp

• However, the above equation does not include the effect of noise.

• We know that here are many sources of noise. However, to start with we
assume that the dominant effect is receiver noise and that this noise is
characterized by noise spectral density No.

31
Free Space Link Budget
• To include the effect of noise, we divide each side of the Friis
equation by NO.
PR PT
GTGR
No LpNO

• Recall that No is characterized by thermal noise, which is given by

No = K . T (W/Hz)

32
Free Space Link Budget
• To include the effect of noise, we divide each side of the Friis equation by
NO.
PR C G
EIRP Lp K
No N0 T

• In the above equation, C/No is the received carrier to noise density ratio
(dB-Hz).

• EIRP is GTPT is the equivalent isotropic radiated power of the transmitter in

dBW and G/T is the ratio of receiver antenna gain to noise temperature in
dB/K
33
Free Space Link Budget

PR C G
EIRP Lp K
No N0 T

• The C/No ratio is one of a number of equivalent ways of expressing the SNR.

• EIRP is GTPT is the equivalent isotropic radiated power of the transmitter in

dBW and G/T is the ratio of receiver antenna gain to noise temperature in
dB/K

34
Exercise
• Consider the design of a radio-controlled model airplane with a
maximum range of 300 m. The receiver requires a C/No ratio of 47
dB-Hz. Due to poor isolation from the aircraft engine, the receiver has
a noise temperature of 320 Kelvins. What EIRP would have to be
transmitted to achieve the maximum range? Assume line-of-sight
transmission at 45 MHz, and assume that transmit and receive
antennas have gains of -3 dB relative to an isotropic antenna.

35
Link Budget

• If the C/No ratio is less than the required one, then reliable
communication is not possible.

• If it is greater than the required one, the link has a margin, which
allows for a tolerance under unknown or sudden transmission
impairments.

36