Shadowing

„ Also called slow-fading

Log-normal Shadowing „ Accounts for random variations in
received power observed over distances
comparable to the widths of buildings
Instructor: M.A. Ingram
„ Extra transmit power (a fading margin)
ECE4823 must be provided to compensate for
these fades

Local Average Power

Measurements Scanning in 3D
The antenna
„ Take power measurements in Watts as the antenna is moved in
a on the order of a few wavelengths
„ After the cart is stabilized,
„ Average these measurements to give a local average power the XYZ actuators move an Z
measurement
antenna in 3 D to make a set
> λ/2
local power measurements Y X

Velocity of antenna
„ One local average power is
computed for each cart
location
5

Z
Points where power measurements are made
0
40

Y
30 50
45
40
20 35
30

X
25
10 20
15
10
0 5
0

Scan volume

Same-Distance Measurements Repeat for Multiple Distances

„ Local averages are made for many „ Similar collections of average powers
different locations, keeping the same are made for other Tx-Rx distances
transmitter-receiver distance
„ These local averages will vary randomly
with location
Example Tx-Rx
locations within
a floor of a building

1
Likelihood of Coverage Likelihood of Coverage, cont’d
„ At a certain distance, d, what is the „ Since only the path loss L(d) is random,
probability that the local average the probability can be expressed as a
received power is below a certain probability involving path loss:
threshold γ?
P( Pr (d ) < γ ) P( Pr (d ) < γ ) = P( L(d ) > β )
Pt Gt Gr
Pr (d ) =
Lt L(d ) Lr maximum tolerable path loss
distance dependence shown explicitly

Typical Macrocell
Characteristics Path Loss Assumptions
[not real data]
-60 . „ The mean loss in dB follows the power law :
.... ..

.. ... ... d 
...... . . ..

-70 ......... ...

.

... . ..
. . . ..

L(d ) = L(d o ) + 10n log10  

..
Total Path Loss [-dB]

.. . .
. . .. .............. .. .  do 
........... . . . ... ...
-80 . .... .... . . .
. .. . .. ......... ... . . . The set of average
The measured loss in dB varies about this
. .. ... ... .. .
. . ..
. . ...... . .. . . . .
. .. ....
. .

losses measured „
..

-90
.... ... . .... ........... ... ..
.
. . ..

at about 4km mean according to a zero-mean Gaussian RV,

. .
. . . .. .
. . . . ....

[10log10(4)=6]
.

... ..... . . .. . Xσ, with standard deviation σ

.. .. ........... ..
-100
.. .. ..
.
.

.... ... .. . . ...

. .. d 
.... .. ... . .
-110
L(d ) = L(d o ) + 10n log10   + X σ
.

-120  do 
0 1 2 3 4 5 6 7 9 10

10log10(Distance from Base Station [km])

Typical Data Characteristics Probability Calculation

[not real data]
-60 . „ Since L(d) is Gaussian, we need to
.... ..

.. ... ... know how to calculate probability

...... . . ..

-70 ......... ...

.

... . ..
Best Fit
. . . ..

..
Total Path Loss [-dB]

. .. . .
. . .. .. Path Loss
involving Gaussian RVs
........... . . . ... ...
-80 . .... .... . ............ .... .. . Exponent:
n=4
. .. ... ... ..
... .. . .
.
. . ..
. . ...... . .. . . . .
. .. ....

P ( L( d ) > β )
. .. . .. .. . . .
. .
..

-90
.... ... . .... ........... ... ..
.
. . ..

. .. . . .. .
. . . . ....
.

σ is usually
... ..... . . .. .
.. .. ........... ..
-100
.. .. ..
.

5-12 dB for
.

.... ... .. . . ...

. .. mobile comm

.... .. ... . .
-110
.

-120
0 1 2 3 4 5 6 7 9 10

10log10(Distance from Base Station [km])

2
Q Function The Problem with Q
„ If X is a Gaussian RV with mean α and „ The integrand of Q has no
standard deviation σ, then antiderivative
 b −α  „ Q is found tabulated in books
P ( X > b) = Q 
 σ  „ Q can be calculated using numerical
where Q is a function defined as integration
+∞
1  x2 
Q( z ) =
2π ∫ exp −
z
dx
2 

What is Log-normal
Shadowing? Inverse Q Problems
„ If Y is a Gaussian RV and Z is defined „ Sometimes, the probability is specified and
such that Y=logZ, then Z is a log- we must find one of the parameters in the
normal RV argument of Q
 b −α 
„ Shadowing is log-normal shadowing P ( X > b) = Q 
 σ 
when the path loss in dB is Gaussian;
Suppose the value of P( X > b) is given, along
this means that the path loss expressed „
with values of b and α. Solve for σ
as a ratio is log-normal
„ Must look up the argument of Q that gives
the specified value.

Example Inverse Q Problem Boundary Coverage

„ Suppose the mean of the local average received powers at a „ Suppose that a cell has radius R and γ is the
certain distance is -30dBm, that the standard deviation of
shadow fading is 9 dB, and that the observed received power is minimum acceptable received power level
above the threshold 95% of the time. What is the threshold „ Then P( Pr ( R) > γ ) is the “likelihood of
power?  b − (−30) 
P( Pr > b) = Q  = 0.95 coverage” at the boundary of the cell
 9 
„ Q is usually tabulated for arguments of 0.5 and less, so we must „ P( Pr ( R) > γ ) is also the “fraction of time” that a
use the fact that
Q(z ) = 1 − Q(− z ) mobile’s signal is acceptable at a distance R
from the transmitter, assuming the car moves
„ The argument of Q that yields 0.05 is about 1.65
around that circle
b + 30
− = 1.65, and b = −44.85 dBm
9

3
 Percentage of Useful Service
Area Integral Evaluated
„ By integrating these probabilities over „ Assuming log-normal shadowing and the
all the circles within a disk, one can power path loss model, the fraction of useful
compute the fraction of the area within service area is
1  1 − 2ab    1 − ab  
the cell that will have acceptable power U (γ ) = 1 − erf (a) + exp  1 − erf  
 b  
2
2  b 
levels
1
2π R where
U (γ ) =
πR 2 ∫ ∫ P( P (r ) > γ ) rdrdθ
r
γ − Pr ( R) 10n log10 e
0 0 a= and b=
σ 2 σ 2

The Error Function erf and Q

„ erf(x) is another form of the Gaussian integral „ The erf function and Q are related:
(like Q(x))
„ erf(x) has odd symmetry, with extreme values
±1. erf1(x) erf ( z ) = 1 − 2Q ( 2z)
1
x
2
∫e
−t 2
erf ( x ) = dt
π 0
x
-1

„ Note that some authors may define erf

differently

When the Average Boundary

Power is Acceptable Summary
„ Local averages in dB of received power (or
Suppose
„
path loss) tend to be Gaussian when the
Pr (R ) = γ .
Then we may
ensemble is all Tx-Rx locations with the same
use this graph distance in the same type of environment
from [Rappaport „ The mean local average path loss follows the
’96] to figure
the percent standard power model (proportional to
useful service 10logdn)
area
„ Can use Q or erf to calculate the likelihood of
boundary coverage or the percent of useful
service area

4
References
„ [Rapp, ’96] T.S. Rappaport, Wireless
Communications, Prentice Hall, 1996
„ [Saunders,`99] Simon R. Saunders,
Antennas and Propagation for Wireless
Communication Systems, John Wiley
and Sons, LTD, 1999.