WIRELESS COMMUNICATION

MODULE II
 Path Loss and Shadowing
2

 The wireless radio channel is susceptible to noise,

interference, and other channel impediments which
change over time in unpredictable ways as a result of
user movement and environment dynamics.
 There is variation in the received signal due to path loss
and shadowing.
 Path loss
 Caused by dissipation of the power radiated by the
transmitter by effects of the propagation channel .
 Assumed to be the same at a given transmit–receive distance
(assuming no shadowing).
 Received power variation due to path loss occurs over long
distances.
 Path Loss and Shadowing…
3

 Shadowing
 Caused by obstacles between the transmitter and receiver
that attenuate signal power through absorption,
reflection, scattering, and diffraction.
 Strong attenuation blocks the signal.
 Received power variation due to shadowing occurs over
distances that are proportional to the length of the
obstructing object.
 Variations in received power due to path loss and
shadowing occur over relatively large distances, these
variations are sometimes referred to as large-scale
propagation effects.
 Radio Wave Propagation
4

Reflected Signal

Direct Signal Scattered Signal

Diffracted Signal Receiver

Transmitter
 Radio Wave Propagation…
5

 Due to multiple reflections from various objects, the

electromagnetic waves travel along different paths of
varying length.
 The interaction between these waves causes multipath
fading.
 The strengths of the waves decrease as the distance
between the transmitter and receiver increases.
 Propagation models have traditionally focused on
predicting the average received signal strength at a
given distance from the transmitter.
 Large Scale Propagation model
 Small Scale Propagation model
 Propagation Models
6

 Large Scale Propagation model:

 Predicts the mean signal strength for an arbitrary
transmitter-receiver separation distance.
 Useful in estimating the radio coverage area of a
transmitter.
 Small Scale Propagation model:
 Characterises the rapid fluctuations in signal strength over
short travel distances or short time durations
 Free Space Propagation Model
7

 Used to predict received signal strength when the

transmitter and receiver have a clear, unobstructed line-
of-sight path between them.
 Predicts that the received power decays as a function of
the transmitter- receiver separation distance raised to
some power.
 Free Space Propagation Model…
8

 The free space power received by a receiver antenna

which is separated from a radiating transmitter antenna
by a distance d, is given by the Friis free space
equation,

where
Pt - transmitted power
Pr(d) - received power as a function of T-R separation
Gt and Gr - transmitter and receiver antenna gain
 - wavelength in metres
L - system loss factor.
 Free Space Propagation Model…
9

 The gain of an antenna is related to its effective

aperture Ae, by

 The wavelength  is related to the frequency by

 The losses L is due to transmission line attenuation, filter

losses and antenna losses. L = 1 indicate no loss in the
system hardware.
 Free Space Propagation Model…
10

 The Friis equation shows that the received power falls off
as the square of the T-R separation distance.
 The path loss is defined as the difference (in dB) between
the effective transmitted power and the received power.
 The path loss is given by

 When antenna gains are excluded, i.e. antennas have

unity gain, path loss is given by
 Free Space Propagation Model…
11

 The Friis free space model is valid for values of d which

are in the far-field of the transmitting antenna.
 The far-field, or Fraunhofer region is defined as the
region beyond the far field distance df, which is given
by

where D is the largest physical linear dimension of the

antenna.
 To be in the far-field region, df >>D and df >> λ.
 Free Space Propagation Model…
12

 The equation for path loss do not hold for d = 0.

 Hence large-scale propagation models use a distance d0,
as a known received power reference point.
 The received power Pr(d), at any distance d > d0, may be
related to Pr at d0.
 d0 must be chosen such that it lies in the far-field
region and also is less than any practical distance used
in mobile communication system.
 The received power in free space at a distance greater
than d0 is given by
 Free Space Propagation Model…
13

 The received power in dBm is given by

where Pr(d0) is in Watts.

 Problems
14

Q) If a transmitter produces 50W of power, express the transmit

power in units of dBm and dBW. If 50W is applied to a unity gain
antenna with 900 MHz carrier frequency , find the received power in
dBm at a free space distance of 100m from the antenna. What is
Pr(10 km)? Assume unity gain for the receiver antenna also.

Pt = 50W ;  = c/f = (1/3) m

Pt (dBm) = 10 log(50 x 103) = 47 dBm
Pt(dBW) = 10 log(50) = 17 dBW
Pr= PtGtGr 2/(4)2d2L = 50*1*1*(1/3)2 / (4)2(100)2*1 = 3.5 x 10-6 W
Pr(dBm) = 10 log(3.5x10-6 x 103) = -24.5 dBm
Pr(10 km) = -24.5 + 20 * log(100/10000) = -64.5 dBm
 Ray Tracing
15

 A radio signal transmitted from a fixed source will

encounter multiple objects in the environment that
produce reflected, diffracted, or scattered copies of the
transmitted signal.
• These additional copies of the transmitted
signals are known as multipath signal components.
• It can be delayed in time, attenuated in power and
shifted in phase and/or frequency with respect to the
LOS signal path at the receiver.
• The multipath and transmitted signal are summed
together at the receiver, which often produces distortion
in the received signal relative to the transmitted signal.
 Ray Tracing…
16

 In ray tracing, we assume a number of reflectors with

known location and dielectric properties.
 To simplify the calculation, it is assumed that the
wavefront travels as particles.
 Then the effects of reflection, diffraction and scattering
can be solved by simple geometrical methods.
 If the transmitter, receiver and all reflectors are
immobile then the characteristic of the multipath
received signals are fixed.
 If the source or receiver is moving then the
characteristics of multipath vary with time.
 Ray Tracing…
17

 The most general ray tracing model includes all

attenuated, diffracted and scattered multipath
components.
 The two-ray model is a simple model that predicts the
signal variation from a ground reflection interfering
with the LOS path.
 This model characterizes signal propagation in isolated
area with few reflectors.
 Two-Ray Model
18

 This model considers both the direct path and a ground

reflected path between the transmitter and the receiver.
 Good for systems that use tall towers (over 50 m tall)
 Good for line-of-sight microcell systems in urban environments
 Two-Ray Model…
19

 ETOT is the electric field that results from a combination of a

direct line-of-sight path and a ground reflected path.
ETOT = ELOS + Eg
 If E0 is the free space E-field at a reference distance d0 from
the transmitter, then for d > d0, the free space propagating
electric field is given by,

 E0d0/d is the amplitude of the electric field at a distance d.

 c = 2fc is the carrier frequency of the signal.
 Two-Ray Model…
20

 Two propagating waves arrive at the receiver:

 The direct wave that travels a distance d

 The reflected wave that travels a distance d”

where  is the reflection coefficient for the ground.

 The total electric field is given by,
 Two-Ray Model…
21

 For large T-R separation, θi goes to 0 (angle of incidence to the

ground of the reflected wave) and Γ = −1.
 The total electric field equation becomes,

 The path difference, , between the line-of-sight and the

ground reflected path can be calculated as,
 Two-Ray Model…
22
 Two-Ray Model…
23

 ∆ can be expanded using a Taylor series expansion

 Two-Ray Model…
24

• This works for d >> (ht + hr)

 Two-Ray Model…
25

 Once the path difference is known, the phase difference 

between the two E-field and the time delay d between the
arrival of the two components can be calculated as,
 = 2/ = c/c
d = /c = /2fc
 Using the phase relationships
E0 d 0  
ETOT (t )  2 sin   
d  2
 2 hr ht
  0.3 rad
2 d
E0 d 0 2 hr ht k
ETOT (t )  2  2 V/m
d d d
 Simplified Path Loss Model
26

 Difficult to obtain a single model that characterizes path loss

accurately due to complexity in signal propagation.
 The simplified model for path loss as a function of distance is

K is a unitless constant dependent on the antenna

characteristics and the channel attenuation, d0 is a reference
distance in the far-field region and  is the path-loss exponent.
 Simplified Path Loss Model…
27

 The simplified model is valid only for transmission

distance d > d0, where d0 is generally assumed to be 1-
10m indoors and 10-100m outdoors.
 The value of K is given by,

 The value of  depends on the propagation environment.

 Statistical Multipath Channel
28
Models - Introduction
 In the ray tracing model, we assumed a deterministic
model to calculate the path loss.
 But in reality deterministic models are rarely available,
so a statistical characterization of multipath models
must be developed.
 If a single pulse is transmitted over a multipath
channel, the received signal will appear as a pulse
train.
 The pulses correspond to the LOS component or a
multipath component associated with a scatterer.
 Statistical Multipath Channel
29
Models – Introduction…
 The time delay spread causes significant distortion in the
received signal.
 The delay spread is the time delay between the arrival of the
first LOS component and the last received signal component.
 The multipath channel is time-varying in nature as the
transmitter or receiver could be moving.
 The location of reflectors which give rise to multipath
components whose amplitude, delay and number change over
time.
 Time-Varying Channel Impulse
30
Response
 Let the transmitted signal be,

 The corresponding received signal is the sum of the line-of-

sight component and all multipath components.

where n = 0 corresponds to the LOS component.

N(t) is the number of multipath components
rn(t) is the path length of the LOS and multipath components
n(t) is the delay given by rn(t)/c
Dn(t) is the Doppler shift and n(t) is the amplitude
 Time-Varying Channel Impulse
31
Response…
 The nth multipath component maybe associated to a
single reflector or multiple reflectors clustered together.
 Time-Varying Channel Impulse
32
Response…
 If each multipath component is due to a single reflector
then,
 The amplitude n(t) is based on the path loss and shadowing
associated with that multipath component.
 Phase change associated with delay n(t) is e-j2fc n(t)

 Doppler shift fD (t) = vcosn(t)/, where v is the receiver

n
velocity,  is the arrival angle of the received signal and
 = c/fc.
 Doppler frequency shift leads to a Doppler phase shift of

Dn = ‫ 𝑡׬‬2𝜋𝑓𝐷𝑛 𝑡 𝑑𝑡.

 Time-Varying Channel Impulse
33
Response…
 If the multipath component is due to a from a reflector
cluster then, two multipath components with delay 1 and 2
are resolvable if their delay difference exceeds the inverse
of the signal bandwidth.
 i.e. |1 - 2| >>Bu-1
 Multipath components that do not satisfy this resolvability
criteria cannot be separated at the receiver and are non-
resolvable.
 Non-resolvable components are combined into a single
multipath component with delay  ≈ 1 ≈ 2 and an
amplitude and phase corresponding to the sum of the
different components.
 Time-Varying Channel Impulse
34
Response…
 r(t) can be simplified by letting,

 EQN(1)
 The received signal r(t) is obtained by convolving lowpass input
signal u(t) with the lowpass time-varying channel impulse
response c(,t) and then upconverting to the carrier frequency:

 EQN(2)
 Time-Varying Channel Impulse
35
Response…
 t is the time when the impulse response is observed at the
receiver and t -  is when the impulse is launched into the
channel.
 For time-invariant channels c(,t) = c(,t + T), i.e. the response
at time t to an impulse at time t -  is equal to the response at
t+T to an impulse at time t+T- .
 Setting T = -t, we get c(,t) = c(,t – t) = c(), c() is the
standard time-invariant channel impulse response.
 From EQN(1) and EQN(2),
 Time-Varying Channel Impulse
36
Response…
 Time-Varying Channel Impulse
37
Response…
 Time-Varying Channel Impulse
38
Response…
 The system in the figure is considered at two times t1 and t2.
 At t1, there are three multipath components associated with the
received signal and with amplitude, phase, and delay triple (αi, φi,
τi), i = 1, 2, 3.
 The impulses that were launched into the channel at time t1−τi, i = 1,
2, 3, will all be received at time t1.
 The time-varying impulse response corresponding to t1 equals

 The time-varying impulse response at t2 equals

 Narrowband Fading Models
39

 The impact of multipath on the received signal depends on

whether the spread of time delays associated with the LOS
and different multipath components is large or small
relative to the inverse signal bandwidth.
 If this channel delay spread is small then the LOS and all
multipath components are typically nonresolvable, leading
to the narrowband fading model.
 The delay spread Tm for time-varying channels is usually
characterized by the rms delay spread.
 Tm << B-1 implies that the delay associated with the ith
multipath component τi ≤ Tm for all i, so u(t - i) ≈ u(t) for all i.
 Narrowband Fading Models…
40

 The equation for the received signal can be re-written as,

 The received signal differs from the transmitted signal only by

the complex scaling factor

 Let s(t) be an unmodulated carrier with random phase offset 0.

which is narrowband for any Tm.

 Narrowband Fading Models…
41

 The received signal becomes

where the in-phase and quadrature components are given by

and where the phase term

 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
Spectral Density
42

 Derive the autocorrelation and cross-correlation of the in-

phase and quadrature received signal components rI(t) and
rQ(t).
 Some assumptions made are:
 No dominant LOS component
 The amplitude αn(t), multipath delay τn(t), and Doppler frequency
fDn(t) change slowly enough to be considered constant over the time
intervals of interest.
i.e. αn(t) ≈ αn, τn(t) ≈ τn, and fDn(t) ≈ fDn .
 The Doppler phase shift φDn(t) = ‫ 𝑡׬‬2𝜋𝑓𝐷𝑛 𝑑𝑡 = 2𝜋𝑓𝐷𝑛 𝑡
 The phase of the nth multipath component becomes
φn(t) = 2πfcτn − 2πfDn t − φ0.
 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
Spectral Density
43

 Assume that the term 2πfcτn in φn(t) changes rapidly as fc

is a large value and hence 2πfcτn can go through a 360
rotation for a small change in n.

Since αn and φn are independent and φn follows a uniform

distribution in the interval [-, ] .
 Similarly it can be shown that E[rQ(t)] = 0.

 Thus E[r(t)] = 0, which means the received signal is zero

mean.
 The autocorrelation of the in-phase and quadrature
components are now considered.
 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
44
Spectral Density
 On deriving the autocorrelation of rI(t) and rQ(t), it is found
that they are uncorrelated or independent.

 The above was derived following the independence of n

and n and the uniform distribution of n.
 So rI(t) and rQ(t) are uncorrelated and independent.
 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
Spectral Density…
45

 The autocorrelation of in-phase components is derived as

below:

 Substituting
 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
Spectral Density…
46

 Since 2fcn changes rapidly due to a large value of fc and is

uniformly distributed,

since fDn = v cosn/ .

 It is observed that Ar (t, t + ) depends only on . Therefore,
I
rI(t) is a wide-sense stationary (WSS) random process.
 Using a similar derivation we can show that the quadrature

component is also WSS with autocorrelation ArQ() = ArI().

 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
Spectral Density…
47

 The cross-correlation between the in-phase and

quadrature components depends only on the time
difference τ and is given by,
 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
48
Spectral Density…
 We can now show that the received signal
r(t) =rI(t)cos(2fct) – rQ(t)sin(2fct) is also WSS with
autocorrelation given by,

 Assume a uniform scattering environment where the channel

consists of many scatterers densely packed with respect to
angle.
 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
49
Spectral Density…

• Assume N multipath components with an angle of

arrival n = n, where  = 2/N .
• Also it is assumed that each multipath component has
the same power, so E[n2] = 2Pr/N, where Pr is the
total received power.
• Then the equation for autocorrelation of rI(t) becomes

• Substituting N = 2/, we get

 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
50
Spectral Density…
 As the numbers of scatterers grow to infinity, the summation in
the previous equation becomes an integral.

where J0(x) is the Bessel function of zeroth order.

 Similarly,

 The PSDs of rI(t) and rQ(t) are obtained by taking the Fourier
transform of their respective autocorrelation functions relative
to the delay parameter .
 Since the autocorrelations are equal so are their PSDs.
 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
51
Spectral Density…
 Narrowband Fading Models
Autocorrelation, Cross-Correlation, and Power
52
Spectral Density…

 The PSD of the received signals r(t) is obtained as

 Envelope and Power Distributions
53

 For any two Gaussian random variables X and Y, both with

mean zero and equal variance σ2, Z = (X2 + Y 2)1/2 is
Rayleigh distributed and Z2 is exponentially distributed.
 It has been observed that rI(t) and rQ(t) are both zero-mean.
 If we assume a variance of σ2 for both in-phase and
quadrature components, then the signal envelope

is Rayleigh distributed with distribution

 Envelope and Power Distributions…
54

where is the average received signal power

of the signal – that is, the received power based on path loss and
shadowing alone.
 The power distribution is obtained as,

 The received signal power is exponentially distributed with

mean 2σ2.
 Problem
55

Q) Consider a channel with Rayleigh fading and average

received power Pr = 20 dBm. Find the probability that the
received power is below 10 dBm.
ANS:

Pr = 20 dBm = 100mW

The probability that Z2 , 10dBm is given by,
 Wideband Fading Models
56

 Consider a pulse of width T is transmitted over a multipath

channel.
 If the multipath delay spread Tm << T then the multipath
components are received roughly on top of one another.

• The resulting constructive and destructive

interference causes narrowband fading of
the pulse, but there is little time spreading of
the pulse and therefore little interference with
a subsequently transmitted pulse.
 Wideband Fading Models
57

 If the multipath delay spread Tm >> T, then each of the

different multipath components can be resolved.
 These multipath components could interfere with subsequently
transmitted pulses leading to Intersymbol Interference(ISI).
 As the bandwidth of the transmitted signal increases so that Tm
≈ B-1, the approximation u(t – n(t)) ≈ u(t) will not hold.
 In this case, the received signal is a sum of all copies of the
original signal, where each copy is delayed in time by τn and
shifted in phase by φn(t).
 The signal copies will combine destructively when their phase
terms differ significantly and will distort the direct path signal
when u(t − τn) differs from u(t).
 Wideband Fading Models
58

 In the case of narrowband fading, the multipath components

combine at the receiver to yield the original transmitted signal
with amplitude and phase characterized by random processes.
 These random processes are characterized by their
autocorrelation (or PSD).
 With wideband signals, the received signal experiences
distortion due to the delay spread of the different multipath
components, so the received signal cannot be characterized by
just the amplitude and phase random processes.
 The effect of multipath on wideband signals must therefore take
into account both the multipath delay spread and the time
variations associated with the channel.
 Wideband Fading Models…
59

 The starting point for characterizing wideband channels is the

equivalent lowpass time varying channel impulse response c(τ, t).

 The Fourier transform of c(τ, t) is given by

where Sc(,) is called the deterministic scattering function of

the channel impulse response c(τ, t).
 The time varying channel impulse response c(τ, t) is random
because of the random amplitudes, phases, and delays of the
random number of multipath componentsand must be
characterized statistically.
 Wideband Fading Models…
60

 The statistical characterization of c(τ, t) is thus determined by

its autocorrelation function,

 Assuming that the channel model is WSS, the autocorrelation

becomes independent of t and is given by,

 In real environment, the channel response associated with a

multipath component of delay 1 is different from that with a
delay 2 as both are due to different scatterers.
 Such channels have uncorrelated scattering and are called
WSS with US or USWSS.
 Wideband Fading Models…
61

 Incorporating the US property we get,

 Here the two components are modeled as a single combined

multipath component with delay τ ≈ τ1 ≈ τ2 and t is the
difference in observation time.
 Ac(:t) gives the average output power associated with the
channel as a function of the multipath delay τ.
 The scattering function for random channels is defined as the
Fourier transform of Ac(τ ; t) with respect to the t parameter:
 Wideband Fading Models…
62

 The most important characteristics of the wideband channel –

including its
 Power delay profile
 Coherence bandwidth
 Doppler power spectrum
 Coherence time
are derived from the channel autocorrelation Ac(τ,  t) or the
scattering function Sc(τ, ρ).
 Power Delay Profile
63

 Also called multipath intensity profile.

 The power delay profile represents the average power
associated with a given multipath delay.
 It is defined as the autocorrelation with t = 0: Ac() = Ac(,0).
 The average and rms delay spread are typically defined in
terms of the power delay profile Ac(τ ) as,
 Power Delay Profile…
64

Q) The power delay profile is modeled as having a one-sided

exponential distribution:

Show that the average delay spread is μTm = Tm and find the
rms delay spread.
ANS: Ac() integrates to unity.
Average delay spread is given by
 Power Delay Profile…
65

 The rms delay spread is given by,

 Power Delay Profile…
66

Q) Consider a wideband channel with multipath intensity profile

Find the mean and rms delay spreads of the channel and find
the maximum symbol rate such that a linearly modulated signal
transmitted through this channel does not experience ISI.
ANS:
Average delay spread is given by
 Power Delay Profile…
67

 The rms delay spread is given by,

 To avoid ISI, the symbol period should be large relative to the

rms delay spread. i.e. Ts >> 10Tm
 i.e. Ts = 52.5s
 Symbol Rate = 1/Ts = 1/52.5 s = 19.04 kilo symbols per
second
 Coherence Bandwidth
68

 The time-varying multipath channel can be characterized in the

frequency domain by taking the Fourier transform of c(τ, t) with
respect to τ.

 Since c(τ ; t) is WSS, its integral C(f ; t) is also WSS.

 The auto correlation of C(f;t) is given by
 Coherence Bandwidth…
69

f = f2 – f1

• Considering t = 0, Ac(f,0) = Ac(f) and is given by

 Coherence Bandwidth…
70

 Ac(f) is the Fourier Transform of the power delay profile.

 If Ac(f) = 0, it means that the channel response is independent at
frequency separation of f.
 The frequency Bc where AC(f) ≈ 0 for all f > Bc is called the
coherence bandwidth of the channel.
 It can be seen from the previous equation that if Ac() = 0 for  > T,
then Ac(f) = 0 for f > 1/T .
 The minimum frequency separation Bc for which the channel response
is roughly independent is Bc ≈ 1/T, where T is typically taken to be
the rms delay spread σTm of Ac(τ ).
 Better approximated as Bc ≈ k/σTm, where k depends on the shape
of Ac(τ ) and the precise specification of coherence bandwidth.
 Coherence Bandwidth…
71

 Can be summarized as,

 If we are transmitting a narrowband signal with bandwidth B << Bc,
then fading across the entire signal bandwidth is highly correlated,
i.e. fading is roughly equal across the entire signal bandwidth.
 This is usually referred to as flat fading.
 If the signal bandwidth B >> Bc, then the channel amplitude values
at frequencies separated by more than the coherence bandwidth
are independent.
 The channel amplitude varies widely across the signal bandwidth.
 In this case the fading is called frequency selective.
 If B ≈ Bc then channel behavior is somewhere between flat and
frequency-selective fading.
 Problem
72

Q) In indoor channels σTm ≈ 50 ns whereas in outdoor microcells σTm

≈ 30 μs. Find the maximum symbol rate Rs = 1/Ts for these
environments such that a linearly modulated signal transmitted
through them experiences negligible ISI.
ANS:
For ISI to occur, the symbol time period Ts must be greater than rms
value of delay spread σTm . i.e. Ts  10 σTm .
For indoor channels, Ts ≈ 10*50 ns = 500ns
Rs = 1/Ts = 2 Mbps
For outdoor channels, Ts ≈ 10*30 s = 300s
Rs = 1/Ts = 3.33 kbps
 Doppler Power Spectrum and
73
Channel Coherence Time
 Delay spread and coherence bandwidth do not give
information about the time varying nature of the channel.
 The time variations of the channel that arise from transmitter or
receiver motion cause a Doppler shift in the received signal.
 Doppler spread and Coherence time are the parameters that
describe the time varying nature of the channel.
 Doppler effect can be characterized by taking the Fourier
transform of AC(f ; t) relative to t:
 Doppler Power Spectrum and
74
Channel Coherence Time…
 In order to characterize Doppler at a single frequency, we set f to
zero and then define SC(ρ) = SC(0; ρ).

where Ac(t) = Ac(f = 0; t) .

 The function SC(ρ) is called the Doppler power spectrum of the
channel as the Fourier transform of an autocorrelation it gives the
PSD of the received signal.
 Ac(t) is the autocorrelation function defining how the channel
impulse response decorrelates over time.
 AC(t = T ) = 0 indicates that observations of the channel impulse
response at times separated by T are uncorrelated and therefore
independent.
 Doppler Power Spectrum and
75
Channel Coherence Time…
 The channel coherence time Tc is defined as the range of t
values over which AC(t) is approximately nonzero, and after
which the channel decorrelates.
 The maximum ρ-value for which |SC(ρ)| is greater than zero is
called the Doppler spread of the channel, denoted by BD.
 By the Fourier transform relationship between AC(t) and SC(ρ),
we have BD ≈ 1/Tc.
 Types of Small-Scale Fading
76

 Small-scale fading or simply fading is the rapid fluctuation in

the amplitude of the received signal over a short interval or
short travel distance.
 Depending on the relationship between signal
parameters(bandwidth, signal period) and channel
parameters(rms delay spread, Doppler spread), the
transmitted signal undergoes different types of fading.
 Flat Fading Vs Frequency Selective
77
Fading
 If the radio channel has constant gain over a bandwidth greater
than the bandwidth of the transmitted signal, then the received
signal undergoes flat fading.
 In flat fading, the multipath structure of the channel is such that the
spectral characteristics of the signal are preserved.
 The gain of the channel may undergo fluctuations leading to
variation in the strength of the received signal.
 Flat Fading Vs Frequency Selective
78
Fading…
 Flat fading channels are also called amplitude varying
channels or narrowband channels as the bandwidth of
the transmitted signal is narrow compared to the channel
bandwidth.
 To summarize, a signal undergoes flat fading if,
 Bs << Bc

 Ts >> T
m
where Bs is the signal bandwidth, Ts is the signal time
period, Tm is the rms delay spread and Bc is the
coherence bandwidth.
 Flat Fading Vs Frequency Selective
79
Fading…
 If the gain of the radio channel is constant over a bandwidth
smaller than the bandwidth of the transmitted signal, then the
received signal undergoes frequency selective fading.
 The multipath delay spread is greater than the time period of
the transmitted signal.
Ts << Tm
 Therefore, the channel introduces Intersymbol Interference(ISI).
 In the frequency domain, certain frequency components have
greater gain than others, i.e. the channel becomes frequency
selective.
 Flat Fading Vs Frequency Selective
80
Fading…
 Frequency selective fading channels are also
called wideband channels as the bandwidth of
the signal is wider than that of the channel impulse
response.
 To summarize,

 Bs >> Bc
 Ts << T
m
 A common rule of thumb is that a channel is frequency
selective if Ts  10 Tm
 Fading Effects due to Doppler
81
Spread
 Depending on how fast the transmitted signal changes as
compared to the rate of change of the channel, a channel can
be classified as fast fading or slow fading channel.
 In a fast fading channel, the channel impulse response changes
rapidly within the symbol duration.
 i.e. the coherence time of the channel is less than the symbol
duration.
 A signal undergoes fast fading if ,
 Ts >> Tc
 Bs << BD
 Fading Effects due to Doppler
82
Spread…
 In a slow fading channel, the channel impulse response changes
at a rate much slower than the transmitted signal.
 The channel is assumed to be static for over one or more
reciprocal bandwidth intervals.
 A signal undergoes slow fading is,
 Ts << Tc

 Bs >> BD
 The velocity of the mobile and the baseband signaling
determines whether a signal undergoes fast fading or slow
fading.
 Fading Effects due to Doppler
83
Spread…
 Capacity of Wireless Channels
84

 Determining the capacity of wireless communication channels

has become very important due to the ever increasing traffic.
 The capacity determines the maximum bit rate that can be
transmitted with a small error probability.
 Early works on capacity of a communication channel was by
Shannon based on mutual information between the input
and output of a channel.
 Shannon defined channel capacity as the channel’s mutual
information maximized over all possible input distributions.
 Shannon’s coding theorem proved that a code did exist that
could achieve a data rate close to capacity with negligible
probability of error.
 Capacity in AWGN
85

 Consider a discrete-time AWGN channel with channel input

/output relationship

where x[i] is the channel input at time i, y[i] is the corresponding

channel output and n[i] is a white Gaussian noise random process.
 Assume a channel bandwidth B and received signal power P.

 The received signal-to-noise ratio (SNR) – the power in x[i]

divided by the power in n[i] – is constant and given by

 = P/N0B
where N0/2 is the PSD of the noise.
 Capacity in AWGN…
86

 The capacity of this channel is given by Shannon’s formula:

 EQN(1)
 The unit of capacity is bits per second(bps).
 For a discrete memoryless time-invariant channel with random
input x and random output y, the channel’s mutual information
is defined as

where the sum is taken over all possible input and output pairs x
∈ X and y ∈ Y and logarithm is with respect to base 2.
 Capacity in AWGN…
87

 Shannon proved that channel capacity equals the mutual

information of the channel maximized over all possible input
distributions:
  EQN(2)

 For the AWGN channel, the sum in EQN(2) becomes an

integral over continuous alphabets and the maximizing input
distribution is Gaussian, which results in the channel capacity
given by EQN(1).
 Capacity in AWGN…
88

Q) Consider a wireless channel where power falloff with distance

follows the formula Pr(d ) = Pt(d0/d )3 for d0 = 10 m. Assume the
channel has bandwidth B = 30 kHz and AWGN with noise PSD N0/2,
where N0 = 10−9 W/Hz. For a transmit power of 1W, find the
capacity of this channel for a transmit–receive distance of 100 m and
1 km.
ANS:
C = B log2(1 + );  = P/N0B
CASE1: P = P(100) = 1 (10/100)3 = 10-3 W
 = 10-3 / (10-9 * 30 *103) = 33
C = 30*103 *log2(1 + 33) = 152.6 kbps
 Capacity in AWGN…
89

Q) Consider a wireless channel where power falloff with distance

follows the formula Pr(d ) = Pt(d0/d )3 for d0 = 10 m. Assume the
channel has bandwidth B = 30 kHz and AWGN with noise PSD N0/2,
where N0 = 10−9 W/Hz. For a transmit power of 1W, find the
capacity of this channel for a transmit–receive distance of 100 m and
1 km.
ANS:
C = B log2(1 + );  = P/N0B
CASE2: P = P(103) = 1 (10/1000)3 = 10-6 W
 = 10-6 / (10-9 * 30 *103) = 0.033
C = 30*103 *log2(1 + 0.033) = 1.4 kbps
 Capacity of Flat Fading Channels
90

 We assume a discrete-time channel with time-varying gain

√g[i], 0 ≤ g[i], and AWGN n[i].
 The channel power gain is g[i] and is independent of the
channel input.
 The channel gain g[i] can change at each time i, either as an
independent and identically distributed (i.i.d.) process or with
some correlation over time.
 Capacity of Flat Fading
91
Channels…
 LetP denote the average transmit signal power, N0/2 the
noise PSD of n[i], and B the received signal bandwidth.
 The instantaneous received SNR is then

and its expected value over time is given by

 SinceP/N0B is a constant, the distribution of g[i] determines

the distribution of γ[i].
 Capacity of Flat Fading
92
Channels…
 An input message w is sent from the transmitter to the receiver.
 The receiver reconstructs an estimate wˆ of the transmitted
message w from the received signal.
 The message is encoded into the codeword x, which is
transmitted over the time-varying channel as x[i] at time i.
 The channel gain g[i], also called the channel side information
(CSI), changes during the transmission of the codeword.
 The capacity of this channel depends on what is known about
g[i] at the transmitter and receiver.
 Capacity of Flat Fading
93
Channels…
 Three different scenarios are considered here:
 Channel distribution information (CDI): The distribution of g[i] is
known to the transmitter and receiver.
 Receiver CSI: The value of g[i] is known to the receiver at time i,
and both the transmitter and receiver know the distribution of
g[i].
 Transmitter and receiver CSI: The value of g[i] is known to the
transmitter and receiver at time i, and both the transmitter and
receiver know the distribution of g[i].
 Transmitter and receiver CSI allow the transmitter to
adapt both its power and rate to the channel gain at time
i, leading to the highest capacity of the three scenarios.
 Channel Side Information at
94
Receiver
 Consider the case where the CSI g[i] is known to the receiver
at time i.
 It means that γ [i] is known to the receiver at time i.
 It is also assumed that both the transmitter and receiver know
the distribution of g[i].
 There are two channel capacity definitions that are relevant to
system design:
 Shannon capacity, also called ergodic capacity
 Capacity with outage.
 Shannon (Ergodic) Capacity
95

 Shannon capacity defines the maximum data rate that can

be sent over the channel with a small error probability.
 For Shannon capacity the rate transmitted over the channel is
constant: the transmitter cannot adapt its transmission strategy
relative to the CSI.
 Poor channel states typically reduce Shannon capacity because
the transmission strategy does not incorporate the effect of
these poor states.
 The ergodic (Shannon) capacity in a fading channel is
therefore the expected value of the capacity, taken over all
realizations of the channel.
 Shannon (Ergodic) Capacity…
96

 Shannon capacity of a fading channel with receiver CSI for an

average power constraintP is given by,

 The formula for Shannon capacity is a probabilistic average

and is equal to Shannon capacity for an AWGN channel with
SNR γ averaged over the distribution of γ.
 Hence Shannon capacity is also called Ergodic capacity.
 Shannon (Ergodic) Capacity…
97

where is the average SNR on the channel.

 Thus, the Shannon capacity of a fading channel with receiver

CSI only is less than the Shannon capacity of an AWGN

channel with the same average SNR.
 Or it can be said that fading reduces Shannon capacity when

only the receiver has CSI.

 Shannon (Ergodic) Capacity…
98

Q) Consider a flat fading channel with i.i.d. channel gain √g[i], which
can take on three possible values: √g1 = .05 with probability p1 = .1,
√g2 = .5 with probability p2 = .5, and √g3 = 1 with probability p3 =
.4. The transmit power is 10 mW, the noise power spectral density N0/2
has N0 = 10−9 W/Hz, and the channel bandwidth is 30 kHz. Assume the
receiver has knowledge of the instantaneous value of g[i] but the
transmitter does not. Find the Shannon capacity of this channel and
compare with the capacity of an AWGN channel with the same
average SNR.
ANS:
The channel has three possible received SNRs:
1 = Ptg1/N0B = 0.01 * (0.05)2/(10-9 * 30*103) = 0.8333 =
-0.79dB
 Shannon (Ergodic) Capacity…
99

ANS:
2 = Ptg2/N0B = 0.01*(0.5)2/(10-9*30*103) = 83.33 = 19.2 dB
3 = Ptg3/N0B = 0.01*(1)2/(10-9*30*103) = 333.33 = 25 dB
Shannon capacity is given by

The average SNR is given by  = (0.1*0.8333) + (0.5*83.33) +

(0.4*333.33) = 175.08
Capacity of AWGN channel with this SNR is C = Blog2(1 + 175.08)
= 223.8 kbps
 Capacity With Outage
100

 The Outage capacity is defined as the maximum rate that can

be maintained in all channel states with some probability of
outage.
 Applies to slowly varying channels, where the instantaneous
SNR γ is constant over a large number of transmissions (a
transmission burst) and then changes to a new value based on
the fading distribution.
 Allows bits sent over a given transmission burst to be decoded
at the end of the burst with some probability that these bits
will be decoded incorrectly.
 The transmitter fixes a minimum received SNR γmin and encodes
for a data rate C = B log2(1+γmin).
 Capacity With Outage…
101

 The data is correctly received if the instantaneous received SNR is

greater than or equal to γmin.
 If the received SNR is below γmin then the bits received over that
transmission burst cannot be decoded correctly and the receiver
declares an outage.
 The probability of outage is therefore

 The average rate correctly received over many transmission bursts is

since data is only correctly received on 1 − Pout transmissions.

 The value of γmin is a design parameter based on the acceptable

outage probability.
 Capacity With Outage…
102

Q) Assume a channel with a bandwidth of 30 kHz and three possible

received SNRs: γ1 = .8333 with p(γ1) = .1, γ2 = 83.33 with p(γ2 ) = .5,
and γ3 = 333.33 with p(γ3) = .4. Find the capacity versus outage for
this channel, and find the average rate correctly received for outage
probabilities Pout < .1, Pout = .1, and Pout = .6.
ANS:
 When Pout < .1, can be decoded correctly in all channel states.
γmin = γ1 = .8333 ; C = B log2(1+ γmin) = 30000 log2(1.833) = 26.23
kbps
 When .1  Pout < .6

γmin = γ2 = 83.33 ; C = B log2(1+ γmin) = 30000 log2(84.33) = 191.94 kbps

 When .6 ≤ Pout < 1,
γmin = γ3 = 333.33 ; C = B log2(1+ γmin) = 30000 log2(333.33) = 251.55 kbps
 Capacity With Outage…
103

Q) Assume a channel with a bandwidth of 30 kHz and three possible

received SNRs: γ1 = .8333 with p(γ1) = .1, γ2 = 83.33 with p(γ2 ) = .5,
and γ3 = 333.33 with p(γ3) = .4. Find the capacity versus outage for
this channel, and find the average rate correctly received for outage
probabilities Pout < .1, Pout = .1, and Pout = .6.
ANS:
 For Pout < .1, data transmitted at rates close to capacity C = 26.23
kbps are always correctly received.
 For Pout = .1, the rate correctly received is (1−.1)191940 = 172.75
kbps.
 For Pout = .6, the rate correctly received is (1− .6)251550 = 125.78
kbps.
 Problems
104

Q) Consider an AWGN channel with bandwidth 50 MHz, received

signal power 10 mW, and noise PSD N0/2 where N0 = 2 · 10−9
W/Hz. How much does capacity increase by doubling the received
power? How much does capacity increase by doubling the channel
bandwidth?
ANS:
C = B log2(1 + )
Case1:  = P/N0B = 10*10-3/(2*10-9*50*106) = 0.1; C = 6.875Mbps

Case2:  = P/N0B = 2*10*10-3/(2*10-9*50*106) = 0.2; C = 13.15Mbps

Case3:  = P/N0B = 10*10-3/(2*10-9*2*50*106) = 0.05; C = 7Mbps

 Problems
105

Q) Consider two users simultaneously transmitting to a single receiver

in an AWGN channel. This is a typical scenario in a cellular system with
multiple users sending signals to a base station. Assume the users have
equal received power of 10 mW and total noise at the receiver in the
bandwidth of interest of 0.1 mW. The channel bandwidth for each
user is 20 MHz. Suppose that the receiver decodes user 1’s signal first.
In this decoding, user 2’s signal acts as noise (assume it has the same
statistics as AWGN). What is the capacity of user 1’s channel with this
additional interference noise?
ANS: C = B log2(1 + )
SNR =  = 10mW/(10.1 mW) = 0.99
C = 20*106 log2(1.99) = 19.85 Mbps
 Problems
106

Q) Consider a flat fading channel of bandwidth 20 MHz and where,

for a fixed transmit powerP, the received SNR is one of six values: γ1
= 20 dB, γ2 = 15 dB, γ3 = 10 dB, γ4 = 5 dB, γ5 = 0 dB, and γ6 = −5
dB. The probabilities associated with each state are p1 = p6 = .1, p2 =
p4 = .15, and p3 = p5 = .25. Assume that only the receiver has CSI.
Find the Shannon capacity of this channel.
ANS: γ1 = 20 dB = 100; γ2 = 15 dB = 31.62; γ3 = 10 dB = 10
γ4 = 5dB = 3.16; γ5 = 0 dB = 1; γ6 = -5 dB = 0.31
= 2.8831 * B = 57.66 Mbps
 Problems
107

Q) Consider a transmitter which radiates a sinusoidal carrier

frequency of 1850 MHz. For a vehicle moving at 26.82m/s compute
the received carrier frequency if the mobile is moving i) directly
towards the transmitter ii) directly away from the transmitter iii) in a
direction which is perpendicular to the direction of arrival of the
transmitted signal.
ANS:
fc= 1850MHz;  = c/fc = 0.162 m
Case1: f = fc + fd ; fd = v/ cos = 26.82/0.162
f = 1850. 00016MHz
Case2: f = fc – fd = 1849.999MHz
Case3:  = 90. So there is no doppler shift.
108

THANK YOU!!!