Wireless Communication Channels

Prof. Giorgio M. Vitetta

Department of Engineering ‘‘Enzo Ferrari’’

University of Modena and Reggio Emilia

A.Y. 2023/2024

1
 Overview

- Phenomena characterising radio propagation

- Small scale fading and large scale fading

- Characterization of large scale fading

- Small scale fading: frequency selective channels and time selective channels

- Slow flat fading and its impact on the error performance of PAM signalling

- Diversity reception

Tech. & Syst. for Digital Communications 2

 Wireless Channels

• Wireless channels usually introduce random variations, depending on time and / or

frequency, in the amplitude and phase of transmitted signals. When these phenomena
occurs, it is usually stated the communication channel is affected by fading (briefly, we
have a fading channel).
• Fading can originate from various causes, among which the most common is
undoubtedly the presence of multiple paths (multipath) in the communication channel, i.e.
the fact that an electromagnetic signal propagates from a transmitting antenna to a
receive antenna along different paths.
• The presence of these paths is normally due to
- the phenomena of reflection, diffraction and scattering produced by objects contained in
the propagation medium;
- and / or to the fact the propagation medium is not homogeneous.

Tech. & Syst. for Digital Communications 3

 Wireless Channels

• Representation of the multipath phenomenon in a mobile communication system

Tech. & Syst. for Digital Communications 4

 Wireless Channels

• When a communication system operates in the presence of a time-dispersive channel, that is of a

channel characterized by the presence of multiple paths, the different replicas (or echoes) of the
transmitted signal collected by the receive antenna are characterized by different amplitudes and
phases; the magnitude of these differences, moreover, may depend on frequency and/or time. In
fact, the phase variation undergoing an electromagnetic wave along a given path is inversely
proportional to its wavelength and, therefore, depends linearly on its frequency. Furthermore, the
presence of changes occurring in the propagation medium and / or the presence of a relative
motion between transmitter and receiver modifies the characteristics of the communication
channel.

• In the presence of multiple paths, a radio receiver is usually unable to separate them, that is it is
unable to extract the contributions associated with different paths; for this reason, it processes the
sum of the multiple signals collected by its antenna.
• In these conditions, if the presence of channel noise is neglected, the received signal is modified, in
its spectral content (both in amplitude and in phase, an in a time varying way ), with respect to the
transmitted signal.

Tech. & Syst. for Digital Communications 5

 Wireless Channels

• In the communication scenario represented in the previous figure the presence of

multiple paths in the communication channel is usually associated with a temporal
variability in the characteristics of these paths; this results in the so-called time
selectivity, i.e. in the temporal fluctuations in the intensity of the received signal.

• This variability is due to the relative movement of the receiver with respect to the
transmitter and / or to environmental changes and, therefore, is perceived when
observing the behavior of the channel over a time interval that extends typically
from fractions of a second to several seconds.

• Commonly, to mathematically describe the behavior of a channel in a time interval

of the above mentioned duration, the adopted models are characterized by a set
of fixed delays for the echoes, each of which, however, has time-varying amplitude
and phase.

Tech. & Syst. for Digital Communications 6

 Wireless Channels

• These models do not provide a complete description of the channel, since in the
considered propagation scenario variations on a longer term also occur; these variations
may involve significant changes in the structure of the channel itself (for example, in the
number of echoes and in the relative delays) and, thus, in the structure of the transmitted
signal.

• These changes are perceived by observing the channel on a time scale of minutes to
tens of minutes or hours, and are often due to weather or solar factors; moreover, in some
cases, long-term variations include daily, seasonal, annual components, or even involve
the sunspot cycle (which has a duration of approximately 11 years).

• Usually, the phenomenon of change in the intensity of the received signal is divided into
two parts: one associated with the rapid variations and the other one related to slow
changes; these are called short term fading and long term fading, respectively.

Tech. & Syst. for Digital Communications 7

 Wireless Channels

• This choice is the result of an inaccurate, but convenient, dichotomy introduced in

the time scale adopted in the observation of communication channels. Obviously,
both types of channel fading can be modelled as continuous time random
processes.

• However, the adoption of this distinction is extremely important from a technical

viewpoint because, in the majority of channels affected by fading, only their short-
term behavior influences the choice of the modulation / coding techniques, as well
as the choice of the type of receiver (which, for example, can be equipped with
multiple antennas for diversity detection).

• Indeed, the aforementioned behavior influences both the structure of the received
waveform and the presence of a possible correlation between the errors occurring
in data detection.

Tech. & Syst. for Digital Communications 8

 Wireless Channels

• Long-term variations influence, however, the availability of the radio channel and,
therefore, the out of service probability of the communication system exploiting it, since,
by modifying the inner structure of the channel, such variations may make it deeply
different from that for which a receiver has been optimized in its design stage.
• In other words, in the presence of these deep changes, mantaining a minimum quality in
the a wireless link may require a signal-noise ratio (SNR) at the receive side greater than
that achievable when the maximum transmission power is used.
• When we will focus our attention on short term fading, we will assume always that it is
exclusively due to the presence of multiple time-varying paths. During our study, however,
we will never forget the fact that a short-term fading model should be always thought of as
always conditioned to the “instantaneous” values of all those parameters describing the
channel statistics on a longer term.

Tech. & Syst. for Digital Communications 9

Wireless Channels

• Our previous comments refer to time variability. Let us now qualitatively analyze
the problem of the spatial variability of the received signal in a radio channel. To
understand the essential aspects of this problem, let us analyse the following
figure, which shows the typical behavior (see the dotted curve) of the ratio,
expressed in dB , between the received power PR and the transmitted power PT
versus the distance d between the transmitter and receiver (normalized with
respect to the wavelength λ characterizing the wireless link).

Tech. & Syst. for Digital Communications 10

Wireless Channels

Tech. & Syst. for Digital Communications 11

Wireless Channels

• This figure shows the presence of rapid fluctuations in the received power of the
useful signal; these are due to the presence of multiple echoes, associated with
different paths, which at a specific point may interfere in a constructive manner,
reinforcing the received signal, or in a destructive way, so appreciably reducing the
signal amplitude.
• In fact, the effects due to the mutual interference of multiple echoes may change
significantly even if the receiver moves by a fraction of a wavelength, since small
variations in the lengths of the paths can entail substantial changes in the phases
of the received echoes.

Tech. & Syst. for Digital Communications 12

 Wireless Channels

• For this reason it is usually stated that these variations in the amplitude of the received
signal represent the so called small-scale propagation effects, or, equivalently, that
represent fading on a small scale (small-scale fading). Note, also, that from the curve
representing the ratio ( PR / PT ) dB versus the distance d, another curve, representing the
average of this ratio, can be extracted. In this case, a spatial average is computed on the
basis of a set of measurements acquired in a neighborhood of the point in which we
intend to evaluate the average itself.

• In the technical literature the average trend shown in the figure is justified by introducing
two different phenomena occurring in radio channels, known as path loss or propagation
loss, and shadowing. Both are classified as propagation effects on a large scale (large -
scale propagation effects) and, in addition, are considered as two different aspects of the
so called large-scale fading affecting radio channels.

Tech. & Syst. for Digital Communications 13

Wireless Channels

• However, the causes of these two phenomena are substantially different.

• In fact, on the one hand, the path loss is due to the spatial dispersion of the power
radiated in the transmission medium and, analytically , is characterized by a
monotonous dependence on the distance d (see the dash-dot line in the last
figure).
• The shadowing, on the other hand, is due to the presence of obstacles interposed
between the transmitter and the receiver, and the attenuation introduced by it is
modelled, if expressed in dB, as a random fluctuation, having zero mean and
superimposed to the path loss, as exemplified by the solid curve shown in the
figure.

Tech. & Syst. for Digital Communications 14

 Wireless Channels

• Another difference between these two macroscopic phenomena concerns the

diversity in the spatial scales over which their appreciable variations occur.
• It is not difficult to understand, in fact, that a significant change in path loss occurs
when there is a variation of several wavelengths in the transmitter-receiver
distance.
• On the contrary, significant fluctuations due to shadowing are perceived in
response to a variation in the distance comparable with the size of the objects
blocking radio waves.
• Usually a significant change in path loss occurs in response to a distance variation
on the order of 100-1000 m, whereas an appreciable change in channel
shadowing requires a variation on the order of 10-100 m outside buildings
(outdoor scenario) and a smaller variation inside buldings (indoor scenario).

Tech. & Syst. for Digital Communications 15

 Wireless Channels

• Over the past two decades there has been a proliferation of communications systems
able to provide to their users, through wireless terminals of small size, access to voice
and / or data services in the UHF band (300 MHz - 3 GHz).

• In this scenario we cannot forget the systems for cordless telephony, the cellular
systems for mobile communications and the wireless local area networks (WLAN 's).

• The design of all these modern systems requires an accurate understanding of the radio
channel, both in its large-scale and a small scale aspects, in different propagation
environments, some outside buildings (outdoor), others inside them (indoor) .

Tech. & Syst. for Digital Communications 16

 Wireless Channels

• Cellular system

Antenna

Coverage
Area

Cell

Tech. & Syst. for Digital Communications 17

Wireless Channels

Tech. & Syst. for Digital Communications 18

 Wireless Channels

• In the first and second generation of cellular telephone systems the coverage area of each base
station (BS) has been divided into cells or macrocells, each with a radius typically ranging from 1 to
10 km. Each cell is covered by a BS, which uses a specific set of channels, to ensure
communications of full duplex type. Base stations are equipped with antennas placed in an higher
position than the surrounding objects and radiating a power usually between 1 and 10 W.

• In such a scenario substantially heterogeneous environments can be found; these can be

characterized by very different distributions in the amplitudes and delays of the echoes generated
by the surrounding environment. For this reason, communication system standards, in order to
ensure identical test conditions for radio receivers implemented by different companies, provide an
accurate description of the statistical properties characterizing different propagation environments.

Tech. & Syst. for Digital Communications 19

 Wireless Channels

• The global system for mobile communication (GSM) standard, for example, provides statistical
models for three different propagation environments, known as

- rural area (RA )

- hilly terrain (HT)

- typical urban (TU)

• Finally, it is worth pointing out that in a macrocellular propagation environment a line of sight (LOS)
component is often absent and this makes the prediction of propagation losses extremely difficult.

Tech. & Syst. for Digital Communications 20

 Wireless Channels
• Recently, there has been a strong interest in small cells.
• Generally speaking, small cells are low-powered radio access nodes, with a range of a
few meters to a mile in diameter. There are three types of small cells, and ranging from
smallest to largest they are called femtocells, picocells and microcells. As a class, they
are considered “small” compared to a mobile macrocell, which can have a range of up to
30 km.
• The various types of small cells have various applications.
• Femtocells are typically user-installed to improve coverage area within a small vicinity,
such as home office or a dead zone within a building. Femtocells can be obtained through
your mobile operator or purchased from a reseller. Unlike picocells and microcells,
femtocells are designed to support only a handful of users and are only able of handling a
few simultaneous calls.
• Picocells offer greater capacities and coverage areas, supporting up to 100 users over a
range of less than 100 m. Picocells are frequently deployed indoor to improve poor
wireless and cellular coverage within a building, such as an office floor or retail space.

Tech. & Syst. for Digital Communications 21

Wireless Channels

• Microcells are difficult to precisely distinguish from picocells, but their coverage
area is the prime delineator. Microcells can cover areas less than a mile in
diameter and use power control to limit this radius. Microcells can be deployed
temporarily in anticipation of high-traffic within a limited area, such as a sporting
event, but are also installed as a permanent feature of mobile cellular networks.

Tech. & Syst. for Digital Communications 22

 Wireless Channels

• Indoor environments

• The development of cordless systems and WLAN’s motivates the intense research in the last 25
years devoted to the study of the propagation and the modeling of the radio channel within
buildings for various uses (offices, warehouses, industrial buildings , etc.), i.e., in a variety of indoor
environments.

• In the technical literature dealing with the problem of radio coverage achievable within buildings
two different scenarios are often considered: in the first scenario the transmitter is located on the
roof of a building other than that in which the receiver is positioned (outdoor-to-indoor scenario),
whereas in the second one both the transmitter and the receiver are located in the same building .

Tech. & Syst. for Digital Communications 23

 Wireless Channels

• Generally speaking, the prediction of propagation losses in an indoor environment is by

far more complicated than the same problem in an outdoor scenario. If, for instance, we
need to predict the electric field intensity inside an office building, a number of factors,
such as the separating walls (which exhibit a frequency dependent behavior), the
presence of multiple floors (if the transmitter is located on a floor different from that of the
receiver), the furniture, the presence of metal pipes and ventilation ducts, must be taken
into account.

• In addition, measurement campaigns carried out in these environments have shown the
presence of many echoes; this makes the spatial fluctuations in the intensity of the
received field more rapid and, therefore, more difficult to predict than in an outdoor
scenario.

Tech. & Syst. for Digital Communications 24

Large Scale Propagation Models

• Received power in free space conditions (far field region)

2
 l 
PR  d  = PT   GT GR ,
 4d 

Transmission Receive
power Transmitter- antenna gain
receiver Transmit
distance antenna gain
l = link wavelength

Tech. & Syst. for Digital Communications 25

Large Scale Propagation Models

• If in a wireless transmission we define the quantity

PT
Ld   ,
PR  d 

called propagation loss or path loss, and representing the power attenuation due
to the transmission medium at a distance d from the transmit antenna, the
equation appearing in the previous slide can be rewritten as

2
1  4d 
L fs  d  =   .
GT GR  l 

Free space
Tech. & Syst. for Digital Communications 26
Large Scale Propagation Models

GT  GR  1 (isotropic antennas)
L fs  l dB  22
d l
L fs  2l dB  22  6  28

distance is doubled (6 dB increase in power attenuation)

• The power PR  d  can be also expressed as:

2
d 
PR  d   PR  d 0   0 
d 

reference distance (in the far field region)

Tech. & Syst. for Digital Communications 27
 Large Scale Propagation Models

• Then, we have that

2
d 
L fs  d  = L fs  d 0    ,
 d0 
which, if power attenuations are expressed in dB, can be rewritten as

d 
L fs  d dB = L fs  d 0 dB  10 nlog 10   ,
 d0 
n2 (path loss exponent)
where

L fs  d dB  10log 10L fs  d 

Tech. & Syst. for Digital Communications 28

Large Scale Propagation Models

• The validity of the results just described requires the presence of an perfectly
uniform and non-absorbing atmosphere, and an earth’s surface at an infinite
distance or with a negligible reflection coefficient.
• In practice, when assessing path loss, the effects of the atmosphere, of the
earth’s surface and, possibly, of the ionosphere should be also taken into account.
If the presence of these effects is not neglected, the mathematical laws just shown
often turn out to be quite optimistic, as exemplified by the study of the scenario
described in example shown below.

Tech. & Syst. for Digital Communications 29

Large Scale Propagation Models

• Radio link between a base station and a mobile terminal. Two paths (a direct path
and a reflected path) are clearly visible. The earth’s surface is supposed to be flat,
independently of the value of the distance d (plane earth model)

Tech. & Syst. for Digital Communications 30

 Large Scale Propagation Models

• Some useful indications on the attenuation introduced by a communication channel with

multiple paths due to the presence of the earth's surface can be acquired by analyzing the
scenario represented in the last figure.
• In this case the wireless link between a base station (BS), positioned at a given height
with respect to ground level, and a mobile terminal, located at a given height and which is
located at distance d from the base station is considered.
• It is also assumed that the radio signal sent by one of the two devices to the other one
(for example, from the base station to the mobile terminal) propagates along two different
paths, i.e. a direct path, and a path generated by the reflection of the earth’s surface,
represented as a flat surface (for this reason, in the technical literature this model is
usually referred to as the plane earth model). The signals associated with the two paths,
being characterized by different lengths, can interfere constructively or destructively.

Tech. & Syst. for Digital Communications 31

Large Scale Propagation Models

• At small distances from the base station, distance variations comparable to λ and
due to a movement of the mobile terminal produce significant fluctuations in the
path loss; these variations can be related to the presence of a sinusoidal
component in the expression of the loss. If spatial averaging is accomplished to
cancel these fluctuations (which represent small scale fading), the expression

PR  d   l 
2

= 2  , n  2 (path loss exponent)

PT  4d 

is obtained (mathematical details are omitted for simplicity).

Tech. & Syst. for Digital Communications 32

 Large Scale Propagation Models

• If the distance is large ( d  hBS  hMT ), the reciprocal of the path loss can be
approximately expressed as

PR  d 
2
h h  n  4 (path loss exponent)
  BS 2MT  ,
PT  d 

• From this result it can be easily inferred that, at a large distance from the base
station, the received power is inversely proportional not to the square of the
distance, but to its fourth power.

Tech. & Syst. for Digital Communications 33

 Large Scale Propagation Models

• In the technical literature it is usually stated that the change from a dependence on d 2
to that on d 4 occurs at a certain breakpoint, whose distance, at high frequencies,
can be approximated as

hBS hMT
dbr  4
l

Tech. & Syst. for Digital Communications 34

Large Scale Propagation Models

hMT  4l PR  d 
2
 l 
 2 
PT  4 d 
hBS  30l

PR  d 
2
h h 
  BS 2MT 
PT  d 

Tech. & Syst. for Digital Communications 35

Large Scale Propagation Models

• Finally it is important to point out that this model, despite its relevance, is not in agreement
with experimental measurements (acquired in real world scenarios) in the following
aspects:
a)The value n = 2 has been actually encountered at a small distance from the transmit
antenna, but at larger distances values of n different from (the theoretical value) 4 and,
more precisely, ranging from 1.5 to 5.5 have been found (the real value of n depends on
various significant features of the surrounding environment);
b) The breakpoint estimated on the basis of theoretical considerations does not
correspond to that found experimentally; for instance, experimental data acquired in urban
scenarios in which the plane earth model previously shown is applied for the prediction of
the path loss in LOS conditions have evidenced that the real breakpoint distance is
usually smaller than its theoretical counterpart.
c) Experimental data have evidenced the presence of a second breakpoint, beyond
which n takes on values larger than 6.

Tech. & Syst. for Digital Communications 36

 Large Scale Propagation Models

• As far as the last point is concerned, it is also important to note that in certain cases this
result can be motivated taking into consideration the presence of a radio horizon, which
has been completely ignored in the derivation of the model illustrated above.

• The mathematical results illustrated in the considered example cannot be applied to other
scenarios since they depend on the geometrical properties of the propagation scenario.

• It is worth pointing out that the complexity of the propagation mechanisms of

electromagnetic signals makes it difficult to derive a single mathematical model able to
accurately assess the path loss in different scenarios. In a specific environment this
parameter can be assessed either using suitable software applications (implementing
complicated mathematical methods) or resorting to a measurement campaign.

Tech. & Syst. for Digital Communications 37

 Large Scale Propagation Models

• These tools can be used in those cases in which, addressing the problem of system
design, certain specifications must be precisely met. This happens, for instance, when we
need to accurately position the base stations of a mobile phone system in a geographic
region in which a given radio coverage must be guaranteed.

• If we need to assess , however, the adequacy of various technical solutions in system

design, simple models are really useful. In these cases the average path loss in dB at a
distance d is usually estimated through the formula

d 
L  d dB  10log 10L  d  = L  d 0 dB  10nlog 10   ,
 d0 
where d 0 is the close-in reference distance and n is the so-called path loss exponent.
This formula represents a generalization of the similar expression referring to free space
propagation (and which is obtained from the last one selecting n = 2).

Tech. & Syst. for Digital Communications 38

 Large Scale Propagation Models

• We observe, first of all, that, like in the case related to free space, the distance d 0 in
the last expression is associated with a circumference lying in the far field zone of the
transmitting antenna. This value, in addition, is normally selected to be small with
respect to the link lengths typical of the considered communication systems. Typical
values of d 0 are 1 m, 100 m and 1 km for an indoor scenario, outdoor microcells
and outdoor macrocells, respectively.

• The use of the previous formula requires the knowledge of the loss L(d 0 ) dB and the
exponent n. The first quantity can be acquired experimentally or estimated by
assuming that the law of propagation in free space applies at a distance d 0 ; as far
as the second parameter is concerned, however, the estimates of n acquired in
different scenarios through measurement campaigns can be exploited.

Tech. & Syst. for Digital Communications 39

 Large Scale Propagation Models

• The typical ranges of the parameter n in different environments are summarised in

the following table.

• They refer to different (indoor an outdoor) measurement scenarios, different

antenna heights and a transmission frequency equal to 0.9 GHz o 1.9 GHz (the
path loss appreciably depends on the trasmission frequency).

• From these data it can be easily inferred that the environments of indoor type are
characterized by a wide range of values for the parameter n, which, moreover, can
also take on a value smaller than 2. The significant width of the interval is justified
by observing that, in such environments, various objects, the materials used for
floors and for walls, the layout and the size of rooms and the number of floors can
contribute to attenuation; the possible presence of an exponent n < 2 is due,
however, to the presence of guided propagation (waveguiding effect).

Tech. & Syst. for Digital Communications 40