STATISTICAL MULTIPATH MODEL

• Typical wireless channel is time-varying system with parameters/MPCs (i.e.

amplitude, phase, Doppler shift, delay) being random and liable to change with time.
• Leads to time-varying channel impulse response
• Due to channel’s complexity & variability makes it difficult to obtain accurate
deterministic channel models. Hence multipath channels must be characterized
statistically e.g. attenuation due to signal path obstructions (shadowing).
• Statistical models also used to characterize constructive and destructive interference
for a large number of MPCs (small-scale fading).
• Statistical models are most accurate in environments with fairly regular
geometries and Uniform dielectric properties e.g. outdoor environments.

Figure: A single reflector and a cluster of closely-spaced reflectors. The reflector cluster
generates multipath components that are typically resolvable in wideband channels and non-
resolvable in narrowband channels.

• Since channel is linear but may not be time invariant, it will be

modelled as linear time varying channel.
 TIME-VARYING CHANNEL IMPULSE RESPONSE

• Assume a transmitted signal

OR

u(t) => complex baseband signal of s(t) with bandwidth Bu,

fc (or f)=> carrier frequency &

ϕ0 => random initial phase (uniformly distributed between 0 and 2π).

 Received signal is (sum of LOS path & all resolvable MPCs)

N(t) => random # of multipaths at time t,

n is their index, and

Gn(t)=> antenna gain,

Rn(t) - reflection coefficient,

dn(t) – path length, and

τn(t) = dn(t)/c the delay corresponding to each MPC.

fn(t) =fD cosθn(t) => doppler shift,

fD = v/λ is doppler frequency and

θn(t) is angle-of-arrival of MPC relative to mobile’s direction of motion.

 • It is useful to define

ϕn(t) = phase due to time delay & doppler in nth MPC,

αn(t) = amplitude of nth MPC

Therefore

• With αn(t), τn(t), & ϕn(t) changing over time, they are characterized as
random processes which are assumed to be both stationary & ergodic.

• r(t) is also a stationary and ergodic random process

• αn(t) is a function of path loss & shadowing while ϕn(t) depends on delay & Doppler,
hence the two random processes are assumed to be independent.
• Received signal may be considered as convolution of baseband input signal u(t) with
the equivalent lowpass time-varying channel impulse response c(τ,t) of the channel and
then up-converting to carrier frequency

where

time-varying linear filter to model the channel

t is time when impulse response is observed

t-τ is time when impulse put into the channel

τ is how long ago impulse was put into the channel for the current observation path delay
for MPC currently observed

 Since αn(t) depends on the path-loss it varies slowly with time.

 Phase ϕn(t) varies rapidly since f + ∆fn(t) is quite large. This leads to rapid variation in
signal strength due to constructive and destructive addition of multipath rays.
 Depending on whether spread of time delays associated with LOS & MPCs is small or
large relative to inverse signal bandwidth, leading to either narrowband or wideband
fading models respectively
 Delay spread equals time delay between arrival of first received signal component (LOS
or multipath) and the last received signal component associated with a single transmitted
pulse.

 For a random process Tm is a random variable given as

Tm= stddev(τ1,. . . , τN(t))

 Channel delay spread commonly characterized by average delay spread and rms delay
spread

NARROWBAND FADING MODELS

 Suppose the delay spread Tm of a channel is small relative to the inverse baseband
signal bandwidth Bu of the transmitted signal; that is, suppose Tm<< B u−1. We refer to
the fading in this scenario as narrowband fading which is a function of both the signal
bandwidth Bu and the channel delay spread Tm.

• When average delay spread ETm is small relative to inverse bandwidth of signal (ETm
≪ 1/Bu), it can be assumed that u(t − τn(t)) ≈ u(t), for all MPCs (for all n and t) i.e. the
signals “overlap"

• Hence no signal distortion (due to little time spreading in received signal)

• Therefore

• Compared to s(t) this shows that channel behaves as time-varying random coefficient
(complex scale factor) g(t) (i.e. as a random process). i.e. signal suffers only scaling by
a complex factor g(t)
• Assuming that u(t) = 1 (tone transmission) then g(t) = r(t)
• Hence

• In-phase & quadrature components

• If N(t) is large & αn(t) & ϕn(t) are independent & identically distributed (i.i.d.)
and zero mean then Central Limit Theorem (CLT) is used to approximate rI (t) and
rQ(t)
• gI (t) and gQ(t) are zero mean jointly Gaussian distributed if ϕn(t) is uniformly
distributed so they can be completely characterized by the mean, autocorrelation,
cross-correlation.

Autocorrelation, Cross Correlation

• Assuming αn(t), τn(t) & fDn(t) are changing slowly & considered constant over time
intervals of interest:
• αn(t) ≈ αn, τn(t) ≈ τn, & fDn(t)≈ fDn. Assumptionvalid for models without
dominant LOS
 component
• With these assumptions, one can verify that
 E[rQ(t)] = 0
E[r(t)] = 0
• If ϕn(t) uniform, in-phase/quad components are zero-mean Gaussian processes, indep.,
and stationary.

Also

i.e. autocorrelation does not depend on t and thus rI(t) and rQ(t) are wide-sense stationary
(WSS) random processes.

• i.e. cross-correlation does not depend on t

• Auto and cross correlation depends on AoAs of
MPCs From these results received signal is also WSS

with autocorrelation
 WIDEBAND FADING MODELS
• Multipath delay spread is greater than pulse width hence individual MPCs resolvable
• True when time difference between components exceeds signal bandwidth
• MPCs interfere with subsequently transmitted pulses; causing ISI
• This distortion can be mitigated through equalization, multicarrier modulation, and
spread spectrum.

• For a wideband input signal the received signal is given by

• where the time-varying channel is

• Note now that since we cannot set u(t−τn(t)) ≈ u(t), it may be possible to resolve the
different MPCs.
• As in narrowband case, rI(t) and rQ(t) components of c(τ, t) are independent Gaussian
processes with same autocorrelation, a mean of zero, and a cross-correlation of zero.
• Autocorrelation function of c(τ, t) is given by
 • In practice, most channels are WSS, such that joint statistics of a channel measured at
two different times t & t+∆t depends only on time difference ∆t.
• Hence modelling channel as WSS, autocorrelation is independent of t

• Moreover, in practice channel response associated with one MPC of delay τ1 is

uncorrelated with response associated with a MPC at a different delay τ2 ≠ τ1, since the
two components are caused by different scatterers (i.e. uncorrelated scattering (US)).
• Hence channel is WSS-US

• Ac(τ; ∆t) = average output power associated with the channel as a function of multipath
delay τ = τ1 = τ2 and difference ∆t in observation time.
• Ac(τ, 0) = Ac(τ) is called multipath intensity profile or power delay profile and it
represents average power of the channel response at time τ to an impulse at time zero
i.e. power distributed over different paths
Multipath Intensity Profile
 For Ac(τ, ∆t = 0)= Ac(τ)

 Determines average delay spread or mean of TM and rms delay spread or standard
deviation of TM
 Approximate max delay of significant MPC.
 PARAMETERS OF MOBILE MULTIPATH CHANNELS

 Power delay profiles are averaging instantaneous power delay profile over a local
area to determine an average small-scale power delay profile. Parameters of
mobile multipath channels includes
Time dispersion parameters
Coherence bandwidth
Doppler spread & Coherence time
 Parameters that describes the time dispersive nature of the channel in a local area are
Delay spread
Coherence bandwidth
 Parameters that describes the time varying nature of the channel in a small
scale region are
Doppler spread
Coherence time

DELAY SPREAD AND COHERENCE

BANDWIDTH
 In multipath phenomenon, the time difference between the arrival moment of
the first multipathcomponent and the last one is called delay spread.

Some parameters are used to quantify the multipath channel.

These parameters are used to compare different multipath channels and to

develop design guidelinesfor wireless systems.

The following multipath parameters are used to quantify the time dispersive
properties of wide bandmultipath channels:

The rms delay spread and mean excess delay are defined from a single power delay
profile which is thetemporal or spatial average of consecutive impulse response
over a local area.

a. Mean excess delay

The mean excess delay is the first moment of the power

delay profile (PDP).It is expressed as

b. RMS delay spread

 This is the most important single measure for the delay times extent of a
multipath delay channel.
 This parameter calculates the standard deviation value of the delay
of reflections.
 The standard deviation value will be weighted proportional to the energy in
the reflected waves. The rms delay spread is the square root of the second
central moment of the power delay profile.
 c. Maximum excess delay

 The maximum excess delay (X dB) of the power delay profile is defined to be
the time delay duringwhich multipath energy falls to X dB below the
maximum.

 Maximum excess delay is sometimes called the excess delay spread.

 The maximum excess delay defines the temporal extent of multipath that is
above a particularthreshold.
 In all cases, must be specified with a threshold that relates the multipath noise
floor to maximumreceived multipath component.

Figure 1.13:
Measured multipath
power delay
profiles
COHERENCE BANDWIDTH

The maximum frequency separation for which the signals are still strongly
correlated is calledcoherence bandwidth (Bc).
Coherence bandwidth, Bc, is derived from the rms delay spread.
Coherence bandwidth is a statistical measure of the range of frequencies over
which the channel isconsidered to be flat .

If the signal bandwidth larger than Bc is transmitted through the channel, it will
subject to Frequencyselective distortion.
The channel will be referred as a frequency selective channel.
If the signal bandwidth is larger than Bc , it will experience amplitude attenuation
only with nodistortion.
This channel will be referred as a frequency non-selective fading channel.

Doppler spread and Coherence time are parameters which describe

DOPPLER SPREAD AND COHERENCE TIME
the time dispersivenature of the channel in a local area.
Also, describes the time varying nature of the channel in a small-scale
region.
They do not give information about the time varying nature of the
channel in large scaleregion.
Doppler spread ( BD)
Doppler spread BD is a measure of the spectral broadening caused by
the time rate ofchange of the mobile radio channel
Doppler spread BD is defined as the range of frequencies over
 which the receivedDoppler spectrum is essentially non-zero.

 It describes the time varying nature frequency dispersiveness of the

channel
 If the baseband signal bandwidth is much greater than BD, the effects of
Doppler spread are negligible at the receiver. This is a slow fading
channel.
Coherence time (TC)
Coherence time is the time over which two signals are having
strong potential for amplitudecorrelation.

Coherence time is the range of time over which similar fading occurs.
The Doppler spread and coherence time are inversely proportional to
one another.

Thus, if the transmitter, receiver, or the intermediate objects move very fast, the
Doppler spread is largeand the coherence time is small, i.e., the channel changes
fast.
That is,
 If the reciprocal bandwidth of the baseband signal is greater than the coherence time of
the channel, then the channel will change during the transmission of the baseband
message, thus causing distortion at the receiver.

TYPES OF SMALL SCALE

The type of fading in the signal propagating through a mobile radio
channel depends onthe nature of the transmitted signal with respect
to the characteristics of the channel.
Depending on the relation between the signal parameters and
channel parameters,different transmitted signals will undergo
different types if fading.
Eg: Signal parameters Bandwidth, symbol period, etc
Channel parameters
 RMS delay and Doppler spread.
The time dispersion and frequency dispersion mechanisms in a mobile
radio channel leadto four possible effects.
The four different types of fading are

Figure 1.14: Types of small scale fading

 Time dispersion due to multipath causes the transmitted signal to undergo
either flat or frequency selective fading.

Figure 1.15: Types of small scale fading (Based on Multipath time delay spread)

a.Flat Fading:

This form of multipath fading affects all the frequencies (almost

equally) in the given channel. Flat multipath fading changes the amplitude
and rising & falling time of the signal.
If mobile radio channel has a constant gain and linear phase response
over a bandwidth that isgreater than the bandwidth of the transmitted
signal, then the received signal will undergo flatfading.
Channel impulse response

Time characteristics
Received signal changes with time due to fluctuations in the gain of the
channel caused bymultipath.
Spectral characteristics
Multipath structure of the channel is such that the spectral characteristics
of the transmittedsignal are preserved at the receiver.
Channel characteristics

Figure 1.16: Flat fading channel characteristics

The amplitude and channel gain varies with time, but the spectrum of the
transmission ispreserved.
Flat fading channels are also known as amplitude varying channels (or)
narrowband channels ( since, the bandwidth of the applied signal is narrow
than the channel bandwidth).
Flat fading channels cause deep fades and requires 20 or 30 dB more
transmitter power toachieve low bit error rates.
Time varying statistics are like Rayleigh flat fading.
Rayleigh distribution channels are used to measure variations of amplitude.
A signal undergoes flat fading if

BS Bc and TS
Where
Bs Signal Bandwidth.
Bc Coherence bandwidth
Ts Reciprocal bandwidth (symbol period)
rms delay spread
b.Frequency Selective Fading

Assume the channel possesses a constant-gain and linear phase response

over a bandwidth that is smaller than the bandwidth of transmitted signal,
then the channel creates frequency selective fading on the received signal.
Channel impulse response

Time characteristics
For this, the received signal includes multiple versions of the transmitted
waveform whichare attenuated (delayed) and delayed in time. Hence the
received signal is distorted.
Spectral characteristics
Frequency selective fading is due to time dispersion of the transmitted
symbols within thechannel. Thus the channel induces intersymbol
interference (ISI).
The received signal spectrum has greater gains than others.

Channel characteristics

Figure 1.17: Frequency selective fading channel characteristics

  The gain is different for different frequency components.
Frequency selective fading channels are also known as wideband channels
since the bandwidth of the signal s(t) is wider than the bandwidth of the
channel impulse response
Selective fading lowers less fading effects [Time varying distortions]
Fading models are used to measure frequency.
For Frequency selective fading, the spectrum S(f) of the transmitted signal
bandwidth (BS)which is greater than the coherence bandwidth (BC)
That is, a signal undergoes frequency selective fading if

FADING EFFECTS DUE TO DOPPLER SPREAD

Depending on how rapidly the transmitted baseband signal changes as

compared to the rate ofchange of the channel, a channel may be classified as
follows.

Figure 1.18: Types of small scale fading (Based on Doppler spread)

a. Fast Fading / Time Selective Fading

In a fast fading channel, the channel impulse response changes rapidly

within the symbolduration.
That is, the coherence time of the channel is smaller than the symbol period
of the transmittedsignal.
This causes Frequency dispersion (also called time selective fading) due to
Doppler spreading,which leads to signal distortion.
In the frequency domain, signal distortion due to fast fading increases with
increasing Dopplerspread relative to the bandwidth of the transmitted
signal. Therefore, a signal undergoes fast fading if

 It should be noted that when a channel is specified as a fast or slow fading

channel, it does notspecify whether the channel is flat fading or frequency
selective in nature.
Fast fading only deals with the rate of change of the channel due to motion.
In flat fading channel, the impulse response of the flat fading channel is a
delta function (no timedelay).
Hence, a flat fading, fast fading channel is a channel in which amplitude of
the delta functionvarying faster than the rate of change of the transmitted
baseband signal.
In case of a frequency selective, fast fading channel, the amplitudes, phases,
and time delays ofany one of the multipath components vary faster than the
rate of change of the transmitted signal. In practice, fast fading only occurs
for very low data rates.
b. Slow Fading

 In a slow fading channel, the channel impulse response changes

at a rate much slower than thetransmitted baseband signal s(t).
 In this case, the channel is assumed to be static over one or
several reciprocal bandwidth intervals.
 In the frequency domain, the Doppler spread of the channel is
much less than the bandwidth of thebaseband signals.
 Therefore, a signal undergoes slow fading if

 The relation between the various multipath parameters and the type of
fading experienced by the signalare summarized in Figure.

Figure 1.19: Matrix illustrating type of fading experienced by

a signal as a function of
(a) symbol period
(b) baseband signal bandwidth
 LARGE-SCALE PROPAGATION
Redicts mean signal strength for an
arbitrary transmitter-receiver (T-R)
separation distance are useful in
estimating the radio coverage area of
a transmitter is called large-scale
propagation
As the mobile moves away from
transmitter over large distances, the
local average received signal
will gradually decrease
The local average signal is computed
by large scale propagation models
typically (Computed by averaging
signal measurements over
measurement track)
cellular=> 1 GHz-2 GHz band power
movement from 1m to 10m.
SMALL-SCALE PROPAGATION
Rapid fluctuations of the received
signal strength over very short travel
distance/short duration are called
Small-scale propagation.
As the mobile moves away from
transmitter oversmall distances, , the
received signal may fluctuate, giving
rise to small scale fading
The received power may vary from
[30/40 dB] when the receiver is
moved by fraction of wavelength.