Path Loss and Shadowing

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Basic Layout of a Wireless Communications System
q The transmitter produces power at the RF operating frequency and is characterized by its
output power,𝑃! , in Watts.
q A feeder cable connecting the transmitter to its antenna is characterized by its loss, 𝐿!
q The transmitting antenna provides the interface between the transmitter circuit and the
wave propagation medium and is characterized by its gain, 𝐺! .
[#$%] [#$] [#$]
q The effective isotropic radiated power (EIRP in dBm) = 𝑃! −𝐿! + 𝐺!

𝑃'
𝑃(
Radio propagation

𝑃! 𝐺! 𝐺'
EIRP = 𝑃' = EIRP×
𝐿! 𝑃( ×𝐿'

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 EIRP Example
q EIRP is the total power radiated by an isotropic antenna in a single direction.
q An isotropic antenna is a theoretical antenna that radiates equally (horizontally and
vertically and with the same intensity) in all directions.
q isotropic antenna has a gain of 1 (linear scale) or 0 dB.
[#$%] [#$] [#$]
q EIRP in dBm = 𝑃! −𝐿! + 𝐺!
q Example: Find EIRP in dBm for the following system:

Transmitter
Cable Loss
(20 dBm)
[#$]
𝐿! =7 dB EIRP = 20 +3 -7 = 16 dBm

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Received Power
q The path loss then depends on how much of this power is captured by the receiving antenna.
q 𝑆: surface power density
𝑃! 𝐺! 𝑃' = 𝐴' 𝑆
𝑆=
𝐺! , 𝐴! 4𝜋𝑟 )
Effective aperture size definition 𝐺' , 𝐴'

𝜆)
𝐴' = 𝐺 𝑃! 𝑃'
4𝜋 '
𝑆 = 𝑆(𝜃, 𝜙)

q Effective area of the receiving antenna is Ar The receiver must be

impedance matched to the antenna's
feedpoint impedance.
q Friis transmission formula
𝐴! 𝐴'
𝑃' = 𝑃!
𝑟 ) 𝜆) 𝑟=𝑑

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Friis Free Space Equation
q Then, the free space received signal strength at distance 𝑑

𝑃! 𝐺! 𝐺' 𝜆) 𝑃! 𝐺! 𝐺' 𝑐 )
𝑃' = 𝐴' 𝑆 = 𝑃' = 𝐿! = 1
4𝜋𝑑 ) 4𝜋𝑑 ) 𝑓*)
𝐿' = 1
q where 𝜆 is related to the carrier frequency by
𝑐
𝜆=
𝑓*
q 𝑓𝑐 is the carrier frequency in Hertz and 𝑐 is the speed of the light in meters/sec

By John S. Huggins - Own work, CC BY-SA 4.0,

https://commons.wikimedia.org/w/index.php?curid=62080634

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Free Space Path Loss (aka Friis’ Law)
q Recall the received power at the output of the receiving antenna
>
𝜆 𝑃!
𝑃< = 𝑃= 𝐺= 𝐺< 𝑃+ =
4𝜋𝑑 𝑃,

[FGL] [FGL] [FGM] [FGM] N

𝑃< = 𝑃= + 𝐺= + 𝐺< + 20 logIJ
OPQ
! #
q The term 4p "
is known as the free space path loss and it may be given in dB
(assume 𝐺𝑟 = 𝐺𝑡 = 1 or 0 dB)
[#$] 𝜆)
𝑃( = −10log-.
4𝜋𝑑 )
𝑐 [FG]
[#$]
𝑃( = −20 log-. 𝑃D = 20 logIJ 𝑓K + 20 logIJ 𝑑 − 147.56 (𝑑𝐵)
4𝜋𝑑𝑓*

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Free Space Propagation Example
q The Friis free space equation shows that the received power falls off as the square of the T-R
separation distances
q The received power decays with distance by 20 dB/decade
q EX: Determine the isotropic free space loss at 4 GHz for the shortest path to a
geosynchronous satellite from earth (35,863 km).
q PL=20log10(4x109)20log10(35.863x106)-147.56dB
q PL=195.6 dB
q Suppose that the antenna gain of both the satellite and ground-based antennas are 44 dB and
48 dB, respectively
q PL=+195.6-44-48=103.6 dB
q Now, assume a transmit power of 250 W at the earth station. What is the power received at
the satellite antenna?

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Transverse EM Wave
q Electromagnetic radiation consists of electromagnetic waves, which are synchronized
oscillations of electric and magnetic fields.
q For a transverse-electric-magnetic (TEM) plane wave traveling through a homogeneous
medium

https://commons
.wikimedia.org/
w/index.php?curi
d=2107870

𝐄 𝒓, 𝑡 = 𝐱U 𝐸. cos(𝑘𝑧 − 𝑤𝑡 + 𝜃. )
𝐲1
/ !
𝐁 Representation of the electric
B 𝒓, 𝑡 = cos(𝑘𝑧 − 𝑤𝑡 + 𝜃. ) 𝐇= field vector of a wave of
* 𝜇. circularly polarized
µà permeability electromagnetic radiation
ε à permittivity
𝜎 → conductivity
Z0 = 376.730313412(59) Ω

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Basic Propagation Mechanisms
q Reflection, diffraction, and scattering:
q Reflection occurs when a propagating electromagnetic wave
impinges upon an object
Large Flat Object
Reflection

Transmitter Receiver

q Diffraction occurs when the radio path is obstructed by a

surface that has sharp edges
q It can explain bending of waves around the corners of an obstacle

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Basic Propagation Mechanisms
D if
fra
c tio
n
t i o n
fl ec
Re

Receiver
Scattering occurs when the medium through
which the wave travels.
• consists of objects with dimensions that are
Transmitter small compared to the wavelength, or
Scattering
• the number of obstacles per unit volume is
large.

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Reflection
q When a plane wave is incident on a planar boundary between two dielectrics, it is
partially reflected and partially transmitted.
q Snell’s Law of Reflection 𝑛- sin 𝜃2 = 𝑛) sin 𝜃! Angle of Incidence=Angle of Reflection

E-Field in the plane of incidence

Ei Er reflected wave
Incident wave 𝑐
Specular Reflection 𝑛=
(mirror-like reflection) 𝜈
Hi Hr
Medium 1 𝜃" 𝜃#
e1, µ1, s1 µà permeability
ε à permittivity
e2, µ2, s2
Medium 2 Refracted
Et
(transmitted) Complex dielectric constant
wave
qt
𝜎
𝛿 =𝜖−𝑗
2𝜋𝑓!
Snell’s Law of Refraction

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Combining EF from Two Paths
q The amplitude and phase of the spherically spreading far-field electromagnetic wave
radiated by an antenna are given by:
𝐸. (𝜃) 3 )45$!678
𝐸(𝑑) = 𝑒
4𝜋𝑑
q The combined electric field from two paths operating at the same frequency is given
by:

𝑑- 𝐸-
An observer in space
𝐸)
𝑑)
𝑒 3)45%! 𝐸.- 6378 𝐸.) 6378
𝐸 = 𝐸- + 𝐸) = 𝑒 & + 𝑒 '
4𝜋 𝑑- 𝑑)

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Received Power via Two Paths
q If the second path is complete out of phase, then, the received power is then given
by
)
𝜆) 𝐺' 𝑃! 𝐺! 𝑒 6378& 𝑃! 𝐺! 𝑒 6378'
𝑃' = −
4𝜋 ) 𝑑- 𝑑)

q If the relative antenna gains are unity and the path lengths are only slightly
different: Δ𝑑 = 𝑑) − 𝑑-

𝜆) 𝐺' 𝐺! 𝑃! Δ𝑑
𝑃' = 2 1 − cos(𝑘Δ𝑑) ≪1
4𝜋𝑑 ) 𝑑-

q The nearly equal received signals from the two paths can then combine to produce a
received power that varies from zero to four times the power on one of the paths.

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Ground Reflection (Two-Ray Model)
q Free space propagation does not apply in a mobile radio environment
q Besides distance and frequency, path loss also depends on the antenna heights
q Consider the signal transmission over a smooth, reflecting, and flat plane
q Direct and reflected path distances:

Total antenna gains

for each path
𝐺9
𝐺𝑟 = 𝐺𝑐𝐺𝑑
𝐺𝑙 = 𝐺𝑎𝐺𝑏
𝐺𝑐
l

x 𝐺𝑏
x' 𝐺𝑑

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Two-Path Propagation over a Plane

q Given that 𝑑 ≫ ℎ𝑡 + ℎ- , and

:(±:)
q Using the approximation (1+y2)1/2≈ 1+y2/2 for small 𝑦 = ≪1
8
q The difference in path length (and 𝑑 = 𝑟) is
2ℎ/ ℎ-
Δ𝑑 = 𝑑. − 𝑑# ≈
𝑟
q The change in the received signal carrier phase
4𝜋ℎ! ℎ'
Δ𝜙 = 𝑘Δ𝑑 =
𝜆𝑟

q The magnitude of the received E-field which is combined from both paths

𝐸!<!9= = 𝐸82'>*! + 𝐸'>5= = 𝐸82'>*! |1 + 𝑅𝑒 637?8 |

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Two-Path Propagation over a Plane
q 𝑅 = −1 (the perfect reflection) and 𝐺 = 𝐺𝑙 = 𝐺𝑟
q The received signal power is approximately

)
𝜆 Δ𝜙
𝑃' = 𝑃! 𝐺 sin)
4𝜋𝑑 2

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Phase Variation
q Δf variation as a function of distance
q ℎ! = 50m;ℎ' = 1.5m; 𝑓* = 900MHz; c=300e3km/sec; l = 𝑐/𝑓;

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Approximate Two-Path Expression
q Approximated equation is given by

ℎ!) ℎ')
𝑃' ≈ 𝑃! 𝐺' 𝐺! A
𝑑
q Significance: the received power and path loss
become independent of frequency at large
distance
q Critical distance 𝑑𝑐 where the rapid fall-off
starts

Power varies with inverse fourth power of distance for large d.

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Diffraction
q Diffraction arises because of the way in which waves propagate; this is described by
the Huygens–Fresnel principle.
q The propagation of a wave can be visualized by considering every point on a
wavefront as a point source for a secondary radial wave.
q When waves are added together, their sum is determined by the relative phases as
well as the amplitudes of the individual waves, an effect which is often known as
wave interference.
q Diffraction demo
q https://www.youtube.com/watch?v=GgfKdVFfM28

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Knife-Edge Diffraction
q Geometry of the model
q The diffracting object is assumed to be asymptotically thin

q The diffracted signal travels distance 𝑑 + 𝑑’ resulting in phase shift of 𝜙 = 2p(𝑑 + 𝑑’)/l
q The phase shift relative to the LOS

Two buildings

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Scattered Rays
q A scattered ray with the segments s and s’ has a path loss proportional to the
product of s and s’.

Smooth Surface Rough Surface

q The received signal due to a scattered ray is given by

-./ +0+ !
,
𝜆 𝑃* 𝐺+ 𝜎𝑒 1
𝐸= 2 𝑒 -./4" *
4𝜋 . 𝑠𝑠 3

q s is the radar cross-section (RCS) of the scattering object

q RCS of a target is an effective area that intercepts the transmitted EM power and then scatters that
power isotropically back to the receiver.

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Wireless Channels
q Main characteristics of wireless channels
q Variations of the channel strength over time and frequency
q Large-scale fading: path loss as a function of distance and shadowing by large
objects, typically frequency independent.
q Small-scale fading: construction of destruction interferences of the multiple signal
paths,
q typically frequency dependent(spatial scale of the carrier wave length)

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Radar
q We assume a monostatic radar:
q The transmit and receive antennas are collocated

q The power density incident on the target

𝑃!
𝑆2 = 𝐺
4𝜋𝑑 ) !
q The power intercepted by the target is proportional to the
incident power density: 𝑃0 = 𝜎𝑆0
q The proportionality constant σ is the radar cross section RCS 𝜎 (m. ).

𝑃2
𝑆B =
4𝜋𝑑 )
𝐴>' 𝜎𝑆2 𝜆) 𝐺' 𝐺! 𝜎
𝑃' = 𝐴>' 𝑆B = = 𝑃!
4𝜋𝑑 ) 4𝜋 C 𝑑 A 𝑑

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Basic Propagation Mechanisms
D if
fra
c tio
n
t i o n
fl ec
Re

Receiver

Multiple signal paths

Transmitter Scattering are formed between
TX and RX

“Multipath”

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Ten-Ray Model (Dielectric Canyon)
q Developed for urban microcells

q Ray tracing over reflected paths: Two of each type of wall-reflected path, one for
each side of the street

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 General Ray Tracing (GRT)
q GRT predicts field strength and delay spread for any building configuration and
antenna placement.
q Building database
q TX-RX locations relative buildings
q Geometrical optics to trace
q the propagation of the LOS and reflected signal components, building diffracted signals,
and diffuse scattering
q Diffraction is most commonly modeled by the Fresnel knife edge diffraction model
due to its simplicity

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 An Example of Real Measurements
q The measurements and predictions for the receiving van driven along 19th St./Nash St.

Received power level (dBW)

Predict the received
local mean power
level with distance

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Power Variations
q Mean power averaged over about 10 wavelengths, itself shows fluctuations.
q Fluctuations in field strength can be better described statistically (mean, standard
deviation, or its PDF)

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Simplified Path Loss Model
q Complex analytical models or empirical measurements when tight system specifications must
be met
q Best locations for base stations
𝛾 à the path loss exponent
q Access point layouts
q However, use a simple model for general tradeoff analysis
𝑑. D
𝑃Q' (𝑑) = 𝑃! 𝐾
𝑑
𝑃Q'P(𝑑)
r(d) dBm
q dB attenuation model 𝑠1=10𝛾 dB/decade
𝑑
𝑃Q' 𝑑𝐵𝑚 = 𝑃! 𝑑𝐵𝑚 + 𝐾8E − 10𝛾 log-.
𝑑.
𝑠1
q 𝐾#$ is sometimes set to the free-space path gain
at distance 𝑑. Isotropic antennas, 𝐺# = 𝐺$ = 1
𝐺! 𝐺' 𝜆) 𝜆
𝐾8E = 10 log-. 𝐾 8E = 20 log -.
4𝜋𝑑. ) 4𝜋𝑑. 𝑑0 Log10d
EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak
 Simplified Path Loss Model vs
a Reference Distance
q The received power level can also be expressed in relation to the received
power level at a close-in reference distance 𝑑0
q Typical value for 𝑑0:
q Indoor:1m
q Outdoor: 100m to 1 km
Reference point

𝑃(, (𝑑V ) 𝑃(, (𝑑)

𝑑𝑜
D
𝑑.
𝑃Q' 𝑑 = 𝑃Q' 𝑑. 𝑑 ≥ 𝑑.
𝑑

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Typical Pathloss Exponents
q Empirically, the relation between the average received power and the distance is
! 3
determined by the expression !+
where 𝛾 is called the path loss exponent
q The typical values of 𝛾 are as:

Environment Path Loss exponent, 𝜸

Free Space 2

Urban Area 2.7 to 3.5

Suburban Area 3 to 5

Indoor (line-of-sight) 1.6 to 1.8

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Empirical Models: Okumura-Hata
q Widely used for signal prediction in urban areas for 1G
q This model is applicable for
q frequencies [150 MHz to 1500 MHz], distances of 1 km to 20 km, base station antenna
heights ranging from 30 m to 200 m
q The path loss formula at distance d and frequency fc
𝑃( 𝑑 = 𝐴 + 𝐵 log-. (𝑑) + 𝐶
q A,B, and C are factors that depend on frequency and antenna height.
𝐴 = 69.55 + 26.16 log 𝑓4 − 13.82 log ℎ5 − 𝑎 ℎ6
𝐵 = 44.9 − 6.55 log ℎ5
where 𝑓* is given in MHz and d in km.
q The function 𝑎(ℎG ) and the factor C depend on the environment:
q Small and medium-size cities:
𝑎(ℎG ) = 1.1 log 𝑓* − 0.7 ℎG − (1.56 log(𝑓* ) − 0.8)
𝐶 = 0.
EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak
 COST 231 Model
q COST 231 model is an extension of the Okumura model to 2 GHz.
q It is used by 2G and 3G standards for the outdoor case.
q COST 231 model gives
𝑃( (𝑑) = (46.3 + 33.9 log-. 𝑓I − 13.82 log-. ℎ! )
+(44.9 − 6.55 log-. ℎ! ) log-. 𝑑 − 𝐶 + 𝐶H

q where C is the correction factor for mobile antenna height in urban areas and 𝐶H is 0 dB
for medium-sized cities and suburbs and 3 dB for metropolitan areas.
q Please see the following link for more discussions on these

http://www.wiley.com/legacy/wileychi/molisch/supp2/appendices/c07_Appendices.pdf

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Large Scale Fading Components
q Shadowing: caused obstacles between TX-RX (medium distance)

d à distance

TX RX

q Goal: Characterize the variation in received signal power over distance due to path
loss and shadowing
q Based on given TX-RX distance (large distance)

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Shadow Fading
q Combined path loss and shadowing model

𝑑
𝑃' 𝑑 = 𝑃Q' (𝑑) + 𝜓8E 𝑃Q' (𝑑) = 𝑃! + 𝐾8E − 10𝛾 log-.
𝑑.
q Measurements have shown that at any value of d, path loss is random
q Distributed log-normal (normal in dB)
q Should we average in linear scale or in dB scale?
q dB scale (is there mathematical justification?)
q The distribution for log-normal shadow fading can be expressed as:

1 (𝜓8E − 𝜇K89 )
𝑓J89 𝜓8E = exp −
2𝜋 𝜎K89 2𝜎K)89

q 𝜓56 is a Gaussian distributed random variable (in dB) 𝜇7 %& = 0 with standard deviation 𝜎7 %& (also in
dB)

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Shadow Fading
𝑆L 𝑑! = 𝑒 6M ∑; 8; = 𝑒 6M8<

𝑑C
q Models attenuation from obstructions
q The attenuation through an object of depth d is
q Random due to random # of obstructions
𝑆L 𝑑 = 𝑒 6M8
q Typically follows a log-normal distribution
q dB value of power is normally distributed
q µ=0 (mean captured in path loss), 4 < 𝜎 < 12 (empirical)
q Lognormal is used to explain this model
q Decorrelated over decorrelation distance 𝑋:

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Log-Normal Shadowing Model
q In linear scale, y=Pt/Pr is log-normal distributed
𝜉 10 log-. 𝜓 )
𝑓J 𝜓 = exp − 𝜓>0
𝜓 2𝜋 𝜎K89 2𝜎K)89
𝜓8E = 10 log-. 𝜓 𝑖𝑛 𝑑𝐵
q x=10/ln10 and
q sy dB is the standard deviation of ydB (also in dB)

q The mean of y can be obtained as

𝜎K)89
𝜇K = 𝐸 Ψ = exp
2𝜉 )
𝜎K)89
10 log-. 𝜇K =
2𝜉

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Finding Path Loss Exponent from
Path Loss Measurements
q 𝛾 depends on the propagation environment
q Free space à 𝛾 = 2
q Two-ray model à 𝛾 = 4
q Example: Consider the set of empirical measurements of 𝑃- /𝑃/ given in the table
below for an indoor system at 900 MHz. Find the path loss exponent 𝛾 that minimizes
the MSE between the simplified model and the empirical dB power measurements,
assuming that 𝑑7 = 1 𝑚 and 𝐾 is determined from the free space path gain formula at
this 𝑑7 .
q Find the received power at 150 m for the simplified path loss model with this path loss
exponent and a transmit power of 1 mW (0 dBm).

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 A line of best fit: Least Square Method
𝑀!"#$% 𝑑& = 10 log'( 𝐾 − 10𝛾log'( 𝑑/𝑑(
-60

-70

q Solution: We first set up the mean squared error -80

Measurements
Simplified Model
(MSE) equation for the dB power measurements as

Loss in dBm
-90
%
1 /
𝐸 𝛾 = ? 𝑀&'()*+', 𝑑" − 𝑀&-,'. 𝑑"

(dBm)
5 -100

Received powerPath
"#$

q Differentiating 𝐸[𝛾] relative to 𝛾 and setting it to zero yields -110

𝜕𝐸 𝛾
-120

=0 γ = 3.71
𝜕𝛾 -130
10
1
10
2 3
10
Distance in log10 d
𝑑' = 1𝑚, 𝐾() = −31.54dB, and Pt = 0 dBm

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 A line of best fit: Least Square Method
q To find the received power at 150m under the simplified path loss model
with γ = 3.71, and, we have

B
Pr = Pt + K − 10𝛾 log10
B0
Pr = 0 − 31.54 − 10 × 3.71 log10(150) = −112.27 dBm.

q Clearly the measurements deviate from the simplified path loss model: this
variation can be attributed to shadow fading
q Estimated error variance
j8E , 𝛾k = 𝜎kK)
𝐸 𝐾 89

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Finding 𝐾9: , 𝛾, and 𝜎9:
q If 𝐾8E is not given, then, we can express MMSE in terms of both 𝐾8E and 𝛾 so that 𝐹(𝐾56 , 𝛾)
q Example continued from the previous slide:
q Then, we need to find the partial derivatives

𝜕𝐸(𝐾,1 , 𝛾) 𝜕𝐸(𝐾,1 , 𝛾)
=0 =0
𝜕𝐾,1 𝜕𝛾
q Solving two equations simultaneously gives 𝐾56 = −31.54 dB and γ = 3.71
q Given Pt = 0 dBm, we can find the received power at 150m under the simplified path loss model:

𝑑 𝑃' =𝑃! − 31.54 − 10×3.71×log10(150)

Pr = 𝑃! +𝐾8E −10γ log-. 𝑃𝑟 = −112.27 dBm
𝑑.
q Clearly the measurements deviate from the simplified path loss model: this variation can be attributed
to shadow fading
G56 = −31.54 dB and 𝛾H = 3.71, we can get an estimate of 𝜎H7 via
q Using 𝐾 %&
𝐸 𝐾G56 , 𝛾H = 𝜎H7.
%&

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Cell Radius Prediction Without shadowing
q The signal level is same on a circle centered at the base station with radius 𝑅
q Find the distance 𝑅 such that the received signal power cannot be less than 𝑃𝑚𝑖𝑛 dBm
q The received signal power at a distance 𝑑 = 𝑅 is specified by
𝑑
𝑃Q' 𝑑 𝑑𝐵𝑚 = 𝑃! 𝑑𝐵𝑚 + 𝐾8E − 10𝛾 log-.
𝑑.

q We need to satisfy: 𝑃F- 𝑅 ≥ 𝑃60G

q Solving the above equation for the radius 𝑅, we obtain

𝑅 ≤ 𝑑. ×10..-(V</D)

q Where 𝑃R = −𝑃G2S + 𝑃! + 𝐾8E

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Outage Probability under Shadowing
q Combining effects of path loss and shadowing have important implications for
wireless system design
q Target minimum received power level, 𝑃𝑚𝑖𝑛 (𝑑)
q The path loss in dB at a distance 𝑑 (> 𝑑0) from the base station is
𝑑
𝑃' 𝑑 = 𝑃! + 𝐾8E − 10𝛾 log-. + 𝜓8E
𝑑.
q where RV 𝜓8E characterizes the effect of shadowing
q In general, 𝜓8E can follow any probability distribution

q Define the outage probability 𝑃𝑜𝑢𝑡(𝑃𝑚𝑖𝑛, 𝑑) under path loss and shadowing to be the
probability that the received power at a distance d, 𝑃𝑟(𝑑) falls below 𝑃𝑚𝑖𝑛

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Outage Probability with Log-Normal Shadowing
q The received power at 𝑑(> 𝑑0) w.r.t. the loss at d0 is given by
𝑑
𝑃',8EG 𝑑 = 𝑃!,8EG + 𝐾8E − 10𝛾 log-. + 𝜓8E
𝑑.

q We can determine the outage probability that the received signal power at location 𝑑 (>
𝑑0) is below the threshold 𝑃𝑚𝑖𝑛,

𝑃<]! 𝑑, 𝑃G2S = 𝑃𝑟𝑜𝑏 𝑃',8EG 𝑑 ≤ 𝑃G2S,8EG

𝑃G2S,8EG − 𝑃',8EG (𝑑)
𝑃<]! 𝑑, 𝑃G2S = 1 − 𝑄
𝜎K89
_ 1 𝑧
1 ' 𝑄 𝑧 = erfc
𝑄 𝑧 =𝑝 𝑥>𝑧 =u 𝑒 6` /) 𝑑𝑦 2 2
2𝜋
^
𝑄 −𝑧 = 1 − 𝑄(𝑧) 𝑧≥0

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Cell Coverage Area
q Next, we want to find the probability of the event
“𝑃r(𝑟) ≥ 𝑃𝑚𝑖𝑛” over a circular area 𝐴 centered at the base station with radius R.

q The probability is given by

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Coverage Area Prediction
q 𝐴 = 𝜋𝑅2 and 𝑑𝐴 = 𝑟𝑑q𝑑𝑟

10𝛾 log-. (𝑒)

𝑏=
𝜎K89

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Multislope Model
q Piecewise linear model

𝛾*

𝛾+
𝛾,

q The dual slope model is characterized by a constant path loss factor K and a path loss
exponent 𝛾- above reference distance d0 up to some critical distance d1, after which point
power falls off with path loss exponent 𝛾) :

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Indoor Propagation Models
q The indoor radio channel differs from the traditional mobile radio channel in two
aspects:
q the distances covered are much smaller
q the variability of the environment is much greater for a small range of T-R separation
distances
q Propagation within buildings is strongly influenced by
q the layout of the building
q the construction materials
q the building type
q Indoor radio propagation is dominated by the same mechanisms as outdoor:
q reflection
q diffraction
q scattering

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Partition Losses between Floors
q The losses between the floors of a building are determined by
q the external dimensions
q materials of the building
q type of construction used to create the floors
q external surrounding
q the number of windows
q presence of tinting

q Partition losses between floors are called floor attenuation factor (FAF)

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Attenuation Factor Model
q A model described by Seidel which includes the effect of building type as well as the
variations caused by obstacles.
q Provides flexibility
q Reduces the standard deviation between measured and predicted path loss to around
4 dB, as compared to 13 dB in log-distance model
q The attenuation factor model:
dI dJ

𝑃' = 𝑃! − 𝑃( 𝑑 − x 𝐹𝐴𝐹2 − x 𝑃𝐴𝐹2 + 𝜓8E

2c- 2c-

q 𝑃𝐿(𝑑) represents path loss for the “same floor” measurement

q FAF: floor attenuation factor
q PAF: partition attenuation factor

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Indoor Measurements
q Scatter plot of path loss as a function of distance in a building

g=6 g=5

g=4

g=3

g=2

g=1

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Summary
q The fundamental limitations of the mobile radio channel on the wireless
communications systems is studied
q The propagation models of radio waves are discussed
q Free space propagation model
q Simplified power falloff model
q Outdoor propagation models
q Indoor propagation models
q Log-distance path loss model with shadowing
q Outage Probability and Cell radius prediction with and without shadowing

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Multipath
q In many wireless applications, the receiving antenna may collect energy from more
than one path.
q The antenna will combine the signals from the several paths incident on it.
q The amplitude and relative phase of each signal are important.
q The resulting signal is the phasor sum of the signals that the antenna will present to
the receiver.

complex channel gain = 𝛼𝑒 3e = x 𝛼2 𝑒 3fK

2

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

SUPPLEMENTARY SLIDES

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Fading Margin
q Fading margin is needed to guarantee certain outage probability

𝑃𝐷𝐹(𝑃',8EG ) Fading Margin [dB]

𝑃<]! = 𝑃𝑟𝑜𝑏 𝑃',8EG − 𝐹𝑀8E < 𝑃G2S,8E

V),89M,NO

𝑃',8EG

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Transmit and Receive Signal Models
q Frequency range: 0.3-3GHz (UHF) and 3-30 GHz (SHF)
q Real signals
q Complex signal models are used for analytical simplicity
q We model the transmitted signal as

q u(t)=x(t)+jy(t) is a complex baseband signal

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Diffracted Signal
q The knife-edge diffraction path loss L(n) can only be approximated with respect to n

q The knife-edge diffraction model yields the following formula for the received
diffracted signal:

q where t=(d+d’)/c is the delay

q n is called the Fresnel-Kirchoff diffraction parameter

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Wireless Channel Models
q How to model wireless channels
q Solve the electromagnetic field equations to find the field impinging on the receiver
antenna.
q This is too complex to solve (exact)
q Use statistic characteristics of the channel and built corresponding models
q Large –scale models
q Based on an arbitrary transmitter-receiver (T-R) separation
q Large models predict the average received signal strength at a given distance.
q Small-scale models
q Characterize rapid fluctuations of the signal strength at short travel distances or short
time durations

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Empirical Path Loss Models
q Radio propagation models are derived using a combination
of analytical and empirical methods
q The empirical approach is based on fitting curves or
analytical expressions that recreate a set of measured data
q Analytical models characterize Pr(d)/Pt as a function of
distance, so path loss is well defined
q In contrast, empirical measurements of Pr(d)/Pt as a
function of distance include the effects of path loss,
shadowing, and multipath

Principle of shadowing

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Local Mean Attenuation
q In order to remove multipath effects, empirical measurements for path loss typically average their
received power measurements in the vicinity of a given receiver location
q The LMA in a given environment, like a city, depends on the specific location of the transmitter and
receiver corresponding to the LMA measurement.
q This average path loss is called the local mean attenuation (LMA) at distance d, and generally
decreases with d due to free space path loss and signal obstructions.
q Thus, the empirical path loss PL(d)=Pr(d)/Pt for a given environment (e.g. a city, suburban area, or
office building) is defined as the average of the LMA measurements at distance d, averaged over all
available measurements in the given environment.

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Rain Fading
q Attenuation arising from significant rain intensity, measured in mm/h becomes
considerable for frequencies above 5 GHz. R is the rainfall rate measured in
millimeters per hour.

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Gaseous Absorption
q The atmosphere of the Earth introduces losses as a result of energy absorption by
atmospheric gases

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak

 Path Loss, Shadowing, and Multipath
q Path Loss, shadowing, and multipath versus distance

q Combined path loss and shadowing model

𝑑
𝑃' 𝑑𝐵𝑚 = 𝑃! 𝑑𝐵𝑚 + 𝐾8E − 10𝛾 log-. + 𝜓8E
𝑑.

EESC6390 Intro. to Wireless Comm. Systems Prof. Murat Torlak