CHAPTER ONE

Fundamentals of Radio
Communications

The purpose of this chapter is to familiarize the reader with the basic propagation
characteristics that describe various wireless communication channels, such as
terrestrial, atmospheric, and ionospheric for VHF to the X-band. Well-known
standards in wireless communication [1–10] are introduced for the prediction of path
losses and fading effects of any radio signal in various communication links, and
finally, new possibilities that can be obtained using smart antennas are discussed.

1.1. RADIO COMMUNICATION LINK

Different radio communication links (land, land-to-air, air-to-air) covering different

atmospheric and ionospheric conditions, include several components having a
plethora of physical principles and processes, with their own independent or
correlated working characteristics and operating elements. A simple scheme of such
a radio communication link consists of a transmitter (T), a receiver (R), and a
propagation channel. The main output characteristics of such a link depend on the
conditions of radio propagation in different kinds of environments, as shown in
Figure 1.1. According to Reference [6], there are three main independent electronic
and electromagnetic design tasks related to this wireless communication network.
The first task is the transmitter antenna operation including the specification of the
electronic equipment that controls all operations within the transmitter. The second
task is to understand, model, and analyze the propagation properties of the channel
that connects the transmitting and receiving antennas. The third task concerns the
study of all operations related to the receiver.

Radio Propagation and Adaptive Antennas for Wireless Communication Links: Terrestrial, Atmospheric
and Ionospheric, by Nathan Blaunstein and Christos Christodoulou
Copyright # 2007 John Wiley & Sons, Inc.

1
2 FUNDAMENTALS OF RADIO COMMUNICATIONS

Wireless Propagation Channel

Propagation Channel

Electronic Electronic
Channel Channel

Additive Attenuation Multiplicative Additive

Absorption noise
noise (path loss) noise
(fading)

Transmitter Receiver
FIGURE 1.1. A wireless communication link scheme.

The propagation channel is influenced by the various obstructions surrounding

antennas and the existing environmental conditions. Another important question for
a personal receiver (or handheld) antenna is also the influence of the human body on
the operating characteristics of the working antenna. The various blocks that
comprise a propagation channel are shown in Figure 1.1.
Its main output characteristics depend on the conditions of radio wave
propagation in the various operational environments where such wireless
communication links are used. Next, we briefly describe the frequency spectrum,
used in terrestrial, atmospheric, and ionospheric communications, and we classify
some common parameters and characteristics of a radio signal, such as its path loss
and fading for various situations, which occur in practice.

1.2. FREQUENCY BAND FOR RADIO COMMUNICATIONS

The frequency band is a main characteristic for predicting the effectiveness of

radio communication links that we consider here. The optimal frequency band
for each propagation channel is determined and limited by the technical re-
quirements of each communication system and by the conditions of radio
propagation through each channel. First, consider the spectrum of radio frequencies
and their practical use in various communication channels [1–5].
Extremely low and very low frequencies (ELF and VLF) are frequencies below
3 kHz and from 3 kHz to 30 kHz, respectively. The VLF band corresponds to waves,
which propagate through the wave guide formed by the Earth’s surface and the
ionosphere at long distances with a low degree of attenuation (0.1–0.5 per 1000 km
[1–5]).
 NOISE IN RADIO COMMUNICATION LINKS 3

Low frequencies (LF) are frequencies from 30 kHz up to 3 MHz. In the 1950s and
1960s they were used for radio communication with ships and aircraft, but since then
they are used mainly with broadcasting stations. Because such radio waves
propagate along the ground surface, they are called ‘‘surface’’ waves [1–5].
High frequencies (HF) are those that are located in the band from 3 MHz up to
30 MHz. Signals in this spectrum propagate by means of reflections caused by the
ionospheric layers and are used for communication with aircraft and satellites, and
for long-distance land communication using broadcasting stations.
Very high frequencies (VHF) are located in the band from 30 MHz up to
300 MHz. They are usually used for TV communication, in long-range radar
systems and radio navigation systems.
Ultra high frequencies (UHF) are those that are located in the band from
300 MHz up to 3 GHz. This frequency band is very effective for wireless microwave
links, constructions of cellular systems (fixed and mobile), mobile–satellite
communication channels, medium range radars, and other applications.
In recent decades, radio waves with frequencies higher than 3 GHz (C, X,
K-bands, up to several hundred gigahertz, which in the literature are referred to as
microwaves) have begun to be widely used for constructing and performing modern
wireless communication channels.

1.3. NOISE IN RADIO COMMUNICATION LINKS

The effectiveness of each radio communication link—land, atmospheric, or

ionospheric depends on such parameters, as [5]:

– noise in the transmitter and in the receiver antennas;

– noise within the electronic equipment that communicate with both antennas;
– background and ambient noise (cosmic, atmospheric, artificial man-made, and
so forth).

Now let us briefly consider each type of noise, which exists in a complete
communication system. In a wireless channel, specifically, the noise sources can be
subdivided into additive and multiplicative effects, as seen in Figure 1.1 [6,7,10].
The additive noise arises from noise generated within the receiver itself, such as
thermal noise in passive and active elements of the electronic devices, and also from
external sources such as atmospheric effects, cosmic radiation, and man-made noise.
The clear and simple explanation of the first component of additive noise is that
noise is generated within each element of the electronic communication channel due
to the random motion of the electrons within the various components of the
equipment [5]. According to the theory of thermodynamics, the noise energy can
be determined by the average background temperature, T0, as [1–5]:

EN ¼ kB T0 ð1:1Þ
4 FUNDAMENTALS OF RADIO COMMUNICATIONS

where

kB ¼ 1:38  1023 W  s  K1 ð1:2Þ

is Boltzman’s constant, and T0 ¼ 290 K ¼ 17 C. This energy is uniformly

distributed in frequency band and hence it is called ‘‘white noise.’’ The total
effective noise power at the receiver input is given by the following expression:

NF ¼ kB T0 BW F ð1:3Þ

where F is the noise figure at the receiver. The noise figure represents any
additional noise effects related to the corresponding environment, and it is
expressed as:

Te
F ¼1þ ð1:4Þ
T0

Here Te is the effective temperature, which accounts all ambient natural (weather,
cosmic noise, clouds, rain, and so forth) and man-made (industry, plants, power
engine, power stations, and so forth) effects.
The multiplicative noise arises from the various processes inside the propagation
channel and depends mostly on the directional characteristics of both terminal
antennas, on the reflection, absorption, scattering, and diffraction phenomena caused
by various natural and artificial obstructions placed between and around the
transmitter and the receiver (see Fig. 1.2). Usually, the multiplicative process in the
propagation channel is divided into three types: path loss, large-scale (or slow
fading), and short-scale (or fast fading) [7–10]. We describe these three
characteristics of the multiplicative noise separately in the following section.

1.4. MAIN PROPAGATION CHARACTERISTICS

In real communication channels, the field that forms the complicated interference
picture of received radio waves arrives via several paths simultaneously, forming a
multipath situation. Such waves combine vectorially to give an oscillating resultant
signal whose variations depend on the distribution of phases among the incoming
total signal components. The signal amplitude variations are known as the fading
effect [1–4,6–10]. Fading is basically a spatial phenomenon, but spatial signal
variations are experienced, according to the ergodic theorem [11,12], as temporal
variations by a receiver/transmitter moving through the multipath field or due to the
motion of scatterers, such as a truck, aircraft, helicopter, satellite, and so on. Thus we
can talk here about space-domain and time-domain variations of EM field in
different radio environments, as well as in the frequency domain. Hence, if we
consider mobile, mobile–aircraft or mobile–satellite communication links, we may
observe the effects of random fading in the frequency domain, that is, the
 MAIN PROPAGATION CHARACTERISTICS 5

Base station

Refle
ction

LO
S

Scattering

ction
Diffra
Scattering

Mobile
n
io
a ct
ffr
Di Diffraction
+
Scattering

FIGURE 1.2. Multipath effects caused by various natural and artificial obstructions placed
between and around the transmitting and the receiving antennas.

complicated interference picture of the received signal caused by receiver/

transmitter movements, which is defined as the ‘‘Doppler shift’’ effect [1–7,10].
Numerous theoretical and experimental investigations in such conditions have
shown that the spatial and temporal variations of signal level have a triple nature
[1–7,10]. The first one is the path loss, which can be defined as a large-scale smooth
decrease in signal strength with distance between two terminals, mainly the
transmitter and the receiver. The physical processes that cause these phenomena are
the spreading of electromagnetic waves radiated outward in space by the transmitter
antenna and the obstructing effects of any natural or man-made objects in the
vicinity of the antenna. The spatial and temporal variations of the signal path loss are
large and slow, respectively.
Large-scale (in the space domain) or slow (in the time domain) fading is the
second nature of signal variations and is caused by diffraction from the obstructions
6 FUNDAMENTALS OF RADIO COMMUNICATIONS

placed along the radio link surrounding the terminal antennas. Sometimes this
fading phenomenon is called the shadowing effect [6,7,10].
During shadow fading, the signal’s slow random variations follow either a
Gaussian distribution or a lognormal distribution if the signal fading is expressed in
decibels. The spatial scale of these slow variations depends on the dimensions of the
obstructions, that is, from several to several tens of meters. The variations of the total
EM field describe its structure within the shadow zones and are called slow-fading
signals.
The third nature of signal variations is the short-scale (in the space domain) or
fast (in the time domain) signal variations, which are caused by the mutual
interference of the wave components in the multiray field. The characteristic scale of
such waves in the space domain varies from half-wavelength to three-wavelength.
Therefore, these signals are usually called fast-fading signals.

1.4.1. Path Loss

The path loss is a figure of merit that determines the effectiveness of the propagation
channel in different environments. It defines variations of the signal amplitude or
field intensity along the propagation trajectory (path) from one point to another
within the communication channel. In general [1–3, 6–10], the path loss is defined as
a logarithmic difference between the amplitude or the intensity (called power) at any
two different points, r1 (the transmitter point) and r2 (the receiver point) along the
propagation path in the medium. The path loss, which is denoted by L and is
measured in decibels (dB), can be evaluated as follows as [5]:

– for a signal amplitude of A(rj ) at two points r1 and r2 along the propagation
path
A2 ðr2 Þ
L ¼ 10 log ¼ 10 log A2 ðr2 Þ  10 log A2 ðr1 Þ
A2 ðr1 Þ
¼ 20 log Aðr2 Þ  20 log Aðr1 Þ ½dB ð1:5Þ

– for a signal intensity J(rj ) at two points r1 and r2 along the propagation path

Jðr2 Þ
L ¼ 10 log ¼ 10 log Jðr2 Þ  10 log Jðr1 Þ ½dB ð1:6Þ
Jðr1 Þ

If we assume now Aðr1 Þ ¼ 1 at the transmitter, then

L ¼ 20 log AðrÞ ½dB ð1:7aÞ

and

L ¼ 10 log JðrÞ ½dB ð1:7bÞ

 MAIN PROPAGATION CHARACTERISTICS 7

For more details about how to measure the path loss, the reader is referred to
References [1–3,6–10]. As any signal passing through the propagation channel,
passes through the transmitter electronic channel and the electronic channel (see
Fig. 1.1), both electronic channels together with the environment introduce additive
or white noise into the wireless communication system. Therefore, the second main
figure of merit of radio communication channels is the signal-to-noise ratio (SNR or
S/N). In decibels this SNR can be written as:

SNR ¼ PR  NR ½dB ð1:8Þ

where PR is the signal power at the receiver and NR is the noise power at the receiver.

1.4.2. Characteristics of Multipath Propagation

Here we start with the general description of slow and fast fading.

Slow Fading. As was mentioned earlier, the slow spatial signal variations
(expressed in decibels, dB) tend to have a lognormal distribution or a Gaussian dis-
tribution (expressed in watts, W) [1–4,6–10]. The probability density function
(PDF) of the signal variations with the corresponding standard deviation, averaged
within some individual small area or over some specific time period, depends on the
nature of the terrain, of the atmospheric and ionospheric conditions. This PDF is
given by:
( )
1 ðr 
r Þ2
PDFðrÞ ¼ pffiffiffi exp  ð1:9Þ
sL 2p 2sL

Here
r ¼ hri is the mean value of the random signal level, r is the value of the
received signal strength or voltage envelope, and sL ¼ hr 2 
r 2 i is the variance or
time-average power (hri indicates the averaging operation of a variable r of the
received signal envelope).

Fast Fading. In the case of stationary receiver and transmitter (static multipath
channel), due to multiple reflections and scattering from various obstructions sur-
rounding the transmitter and receiver, the radio signals travel along different paths
of varying lengths, causing such fast deviations of the signal strength (in volts) or
power (in watts) at the receiver.
In the case of a dynamic multipath situation, either the subscribers’ antenna is in
movement or the objects surrounding the stationary antennas are moving, so the
spatial variations of the resultant signal at the receiver can be seen as temporal
variations [11,12]. The signal received by the mobile at any spatial point may consist
of a large number of signals having randomly distributed amplitudes, phases, and
angles-of-arrival, as well as different time delays. All these features change the
relative phase shifts as a function of the spatial location and, finally, cause the signal
to fade in the space domain. In a dynamic (mobile) multipath situation, the signal
8 FUNDAMENTALS OF RADIO COMMUNICATIONS

FIGURE 1.3. Geometry of the mobile link for Doppler effect estimation.

fading at the mobile receiver occurs in the time domain. This temporal fading is
associated with a shift of frequency radiated by the stationary transmitter. In fact, the
time variations, or dynamic changes of the propagation path lengths, are related to
the Doppler effect, which is due to relative movements between a stationary base
station (BS) and a moving subscriber (MS).
To illustrate the effects of phase change in the time domain due to the Doppler
frequency shift (called the Doppler effect [1–4,6–10]), let us consider a mobile
moving at a constant velocity v, along the path XY, as shown in Figure 1.3. The
difference in path lengths traveled by a signal from source S to the mobile at points X
and Y is ‘ ¼ ‘ cos y ¼ v cos y, where t is the time required for the moving
receiver to travel from point X to Y along the path, and y is the angle between the
mobile direction along XY and direction to the source at the current point Y, that is,
YS. The phase change of the resultant received signal due to the difference in path
lengths is therefore

2p 2pvDt
F ¼ kD‘ ¼ ‘ cos y ¼ cos y ð1:10Þ
l l

Hence the apparent change in frequency radiated, or Doppler shift, is given by fD,
where

1  v
fD ¼ ¼ cos y ð1:11Þ
2p t l

It is important to note from Figure 1.3 that the angles y for points X and Y are the
same only when the corresponding lines XS and YS are parallel. Hence, this figure is
correct only in the limit when the terminal S is far away from the moving antenna at
points X and Y. Many authors have ignored this fact during their geometrical
explanation of the Doppler effect [1–4,10]. Because the Doppler shift is related to
the mobile velocity and the spatial angle between the direction of mobile motion and
the direction of arrival of the signal, it can be positive or negative depending on
 MAIN PROPAGATION CHARACTERISTICS 9

whether the mobile receiver is moving toward or away from the transmitter. In fact,
from Equation (1.11), if the mobile moves toward the direction of arrival of the
signal with radiated frequency fc, then the received frequency is increased, that is the
apparent frequency is fc þ fD . When the mobile moves away from the direction of
arrival of the signal then the received frequency is decreased, that is the apparent
frequency is fc  fD . The maximum Doppler shift is fDmax ¼ v=l, which, in our
futher description will simply be denoted as fm .
There are many probability distribution functions that can be used to describe the
fast fading effects, such as, Rayleigh, Suzuki, Rician, Gamma, Gamma–Gamma,
and so on. Because the Rician distribution is very general [1–4,10], as it includes
both line-of-sight (LOS) together with scattering and diffraction with non-LOS, we
briefly describe it in the following paragraph.
To estimate the contribution of each signal component, at the receiver, due to the
dominant (or LOS) and the secondary (or multipath), the Rician parameter K is
usually introduced, as a ratio between these components [1–4,10], that is,

LOS  Component power

K¼ ð1:12Þ
Multipath  Component power

The Rician PDF distribution of the signal strength or voltage envelope r can be
defined as [1–4,10]:
 2   
r r þ A2 Ar
PDFðrÞ ¼ 2 exp  I0 2 ; for A > 0; r 0 ð1:13Þ
s 2s2 s

where A denotes the peak strength or voltage of the dominant component envelope,
s is the standard deviation of signal envelope, and I0 ðÞ is the modified Bessel
function of the first kind and zero-order. According to definition (1.12), we can now
rewrite the parameter K, which was defined above as the ratio between the dominant
and the multipath component power. It is given by

A2
K¼ ð1:14Þ
2s2

Using (1.14), we can rewrite (1.13) as a function of K only, [1–3,10]:

   r pffiffiffi
r r2
PDFðxÞ ¼ exp  expðKÞI 0 2K ð1:15Þ
s2 2s2 s

For K ¼ 0, expðKÞ ¼ 1 and I0 ð0Þ ¼ 1, that is, the worst case of the fading channel.
The Rayleigh PDF, when there is no LOS signal and is equal to:
 
r r2
PDFðxÞ ¼ exp  ð1:16Þ
s2 2s2
10 FUNDAMENTALS OF RADIO COMMUNICATIONS

3.5

3.0
K = 32
2.5 K = 16

PDF 2.0

1.5 K=4

K = 0, Rayleigh
1.0

0.5

0
–30 –25 –20 –15 –10 –5 0 5 10
Strength/rms, dB
FIGURE 1.4. Rician PDF distribution versus ratio of signal to rms.

Conversely, in a situation of good clearance between two terminals with no

multipath components, that is, when K ! 1, the Rician fading approaches a
Gaussian one yielding a ‘‘Dirac-delta shaped’’ PDF described by formula (1.9)
(see Fig. 1.4). We will use these definitions in Chapter 5 for the link budget design
inside a terrestrial radio communication system.

1.4.3. Signal Presentation in Wireless Communication Channels

To understand how to describe mathematically multipath fading in communication
channels, we need to understand what kinds of signals we ‘‘deal’’ with in each
channel.

Narrowband (CW) Signals. First of all, we consider a continuous wave CW or

narrowband signals. A voice-modulated CW signal occupies a very narrow band-
width surrounding the carrier frequency fc of the radio frequency (RF) signal
(e.g., the carrier), which can be expressed as:

xðtÞ ¼ AðtÞ cos½2pfc t þ jðtÞ ð1:17Þ

where AðtÞ is the signal envelope (i.e., slowly-varied amplitude) and jðtÞ is its signal
phase. For example, for a modulated 1 GHz carrier signal by a wire signal of
bandwidth f ¼ 2fm ¼ 8 KHz, the fractional bandwidth is very narrow, that is,
8  103 Hz=1  109 Hz ¼ 8  106 or 8  104 %. Since all information in the
signal is contained within the phase and envelope-time variations, an alternative
form of a bandpass signal xðtÞ is introduced [1,2,6–10]:

yðtÞ ¼ AðtÞexpfjjðtÞg ð1:18Þ

 MAIN PROPAGATION CHARACTERISTICS 11

X(f)

∆f ∆f

–f c 0 +fc f
Real Bandpass Real Bandpass
Signal Signal

Y(f)

f
0
Baseband
Signal
FIGURE 1.5. The signal power presentation of the frequency domain. Bandpath (upper
figure) and baseband (lower figure).

which is also called the complex baseband representation of xðtÞ. By comparing

(1.17) and (1.18), we can see that the relation between the bandpass (RF) and the
complex baseband signals are related by:

xðtÞ ¼ Re½yðtÞexpðj2pfc tÞ ð1:19Þ

The relations between these two representations of the narrowband signal in the
frequency domain are shown schematically in Figure 1.5. One can see that the
complex baseband signal is a frequency shifted version of the bandpass (RF) signal
with the same spectral shape, but centered around a zero-frequency instead of the
fc [7]. Here, Xðf Þ and Yð f Þ are the Fourier transform of xðtÞ and yðtÞ, respectively
and can be presented in the following manner [1,2]:
ð
1

Yðf Þ ¼ yðtÞej2pft dt ¼ Re½Yðf Þ þ j Im½Yðf Þ ð1:20Þ

1

and
ð
1

Xðf Þ ¼ xðtÞej2pft dt ¼ Re½Xðf Þ þ j Im½Xðf Þ ð1:21Þ

1
12 FUNDAMENTALS OF RADIO COMMUNICATIONS

Substituting for xðtÞ in integral (1.21) from (1.19) gives:

ð
1

Xðf Þ ¼ Re yðtÞe j2pfc t ej2pft dt ð1:22Þ

1

Taking into account that the real part of any arbitrary complex variable w can be
presented as:

1
Re½w ¼ ½w þ w
2

where w is the complex conjugate, we can rewrite (1.22) in the following form:

ð
1
1
Xðf Þ ¼ yðtÞe j2pfc t þ y ðtÞej2pfc t ej2pft dt ð1:23Þ
2
1

After comparing expressions (1.20) and (1.23), we get

1
Xðf Þ ¼ ½Yðf  fc Þ þ Y  ðf  fc Þ ð1:24Þ
2

In other words, the spectrum of the real bandpass signal xðtÞ can be represented by
real part of that for the complex baseband signal yðtÞ with a shift of fc along the
frequency axis. It is clear that the baseband signal has its frequency content centered
around the ‘‘zero’’ frequency value.
Now we notice that the mean power of the baseband signal yðtÞ gives the same
result as the mean-square value of the real bandpass (RF) signal xðtÞ, that is,

hjyðtÞj2 i hyðtÞy ðtÞi

hPy ðtÞi ¼ ¼  hPx ðtÞi ð1:25Þ
2 2

The complex envelope yðtÞ of the received narrowband signal can be expressed
according to (1.18), within the multipath wireless channel, as a sum of phases of N
baseband individual multiray components arriving at the receiver with their
corresponding time delay, ti ; i ¼ 0; 1; 2; . . . ; N  1 [6–10]

X
N 1 X
N 1
yðtÞ ¼ ui ðtÞ ¼ Ai ðtÞ exp½ jji ðt; ti Þ ð1:26Þ
i¼0 i¼0

If we assume that during the subscriber movements through the local area of service,
the amplitude Ai time variations are small enough, whereas phases ji vary greatly
 MAIN PROPAGATION CHARACTERISTICS 13

due to changes in propagation distance between the base station and desired
subscriber, then there are great random oscillations of the total signal yðtÞ at the
receiver during its movement over a small distance. Since yðtÞ is the phase sum in
(1.26) of the individual multipath components, the instantaneous phases of the
multipath components result in large fluctuations, that is, fast fading, in the CW
signal. The average received power for such a signal over a local area of service can
be presented according to References [1–3,6–10] as:

X
N 1   N 1 X 
X  
hPCW i  A2i þ 2 Ai Aj cos ji  jj ð1:27Þ
i¼0 i¼0 i; j6¼i

Wideband (Pulse) Signals. The typical wideband or impulse signal passing

through the multipath communication channel is shown schematically in
Figure 1.6a according to [1–4]. If we divide the time-delay axis into equal segments,
usually called bins, then there will be a number of received signals, in the form
of vectors or delta functions. Each bin corresponds to a different path whose
time-of-arrival is within the bin duration, as depicted in Figure 1.6b. In this case

~
P, dB
(a)

t, µs
~
P, dB
(b)

1 2 3 4 5 6 7 8 9 10 11 t, µs

FIGURE 1.6. (a) A typical impulse signal passing through a multipath communication
channel according to [1–4]. (b) The use of binds, as vectors, for the impulse signal with
spreading.
14 FUNDAMENTALS OF RADIO COMMUNICATIONS

the time varying discrete-time impulse response can be expressed as:

( )
X
N 1
hðt; tÞ ¼ Ai ðt; tÞexp½j2pfc ti ðtÞ dðt  ti ðtÞÞ exp½jjðt; tÞ ð1:28Þ
i¼0

If the channel impulse response is assumed to be time invariant, or is at least

stationary over a short-time interval or over a small-scale displacement of the
receiver/transmitter, then the impulse response (1.28) reduces to

X
N 1
hðt; tÞ ¼ Ai ðtÞ exp½jyi dðt  ti Þ ð1:29Þ
i¼0

where yi ¼ 2pfc ti þ jðtÞ. If so, the received power delay profile for a wideband or
pulsed signal averaged over a small area can be presented simply as a sum of the
powers of the individual multipath components, where each component has a
random amplitude and phase at any time, that is,
* +
  X
N 1
2
X
N 1  
Ppulse ¼ fAi ðtÞj exp½jyi jg  A2i ð1:30Þ
i¼0 i¼0

The received power of the wideband or pulse signal does not fluctuate significantly
when the subscriber moves within a local area, because in practice, the amplitudes of
the individual multipath components do not change widely in a local area of service.
Comparison between small-scale presentations of the average power of the
narrowband (CW) and wideband (pulse) signals that is, (1.27) and (1.30), shows that
when hAi Aj i ¼ 0 or/and hcos½ji  jj i ¼ 0, the average power for CW signal and
that for pulse are equivalent. This can occur when either the path amplitudes are
uncorrelated, that is, each multipath component is independent after multiple
reflections, diffractions, and scattering from obstructions surrounding both the
receiver and the transmitter or the base station and the subscriber antenna. It can also
occur when multipath phases are independently and uniformly distributed over the
range of b0; 2pc. This property is correct for UHF/X-waveband when the multipath
components traverse differential radio paths having hundreds of wavelengths
[6–10].

1.4.4. Parameters of the Multipath Communication Channel

So the question that is remains to be answered which kind of fading occurs in a given
wireless channel.

Time Dispersion Parameters. First some important parameters for a wideband

(pulse) signal passing through a wireless channel, can be determined, for a certain
threshold level X (in dB) of the channel under consideration, from the signal
 MAIN PROPAGATION CHARACTERISTICS 15

power delay profile, such as mean excess delay, rms delay spread and excess delay
spread.
The mean excess delay is the first moment of the power delay profile of the pulse
signal and is defined as:
X
N 1 X
N 1
A2i ti Pðti Þti
i¼0 i¼0
h ti ¼ ¼ ð1:31Þ
X
N 1 X
N 1
A2i Pðti Þ
i¼0 i¼0

The rms delay spread is the square root of the second central moment of the power
delay profile and is defined as
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
st ¼ ht2 i  hti2 ð1:32Þ

where
X
N1 X
N1
A2i t2i Pðti Þt2i
 2 i¼0 i¼0
t ¼ ¼ ð1:33Þ
X
N 1 X
N 1
A2i Pðti Þ
i¼0 i¼0

These delays are measured relative to the first detectable signal arriving at the
receiver at t0 ¼ 0. We must note that these parameters are defined from a single
power delay profile, which was obtained after temporal or local (small-scale) spatial
averaging of measured impulse response of the channel [1–3,7–10].

Coherence Bandwidth. The power delay profile in the time domain and the
power spectral response in the frequency domain are related through the Fourier
transform. Hence, to describe a multipath channel in full, both the delay spread
parameters in the time domain, and the coherence bandwidth in the frequency
domain are used. As mentioned earlier the coherence bandwidth is the statistical
measure of the frequency range over which the channel is considered ‘‘flat.’’ In
other words, this is a frequency range over which two frequency signals are
strongly amplitude correlated. This parameter, actually, describes the time disper-
sive nature of the channel in a small-scale (local) area. Depending on the degree
of amplitude correlation of two frequency separated signals, there are different
definitions for this parameter.
The first definition is the coherence bandwidth, Bc , which describes a band-
width over which the frequency correlation function is above 0.9 or 90%, and it is
given by:
Bc  0:02s1
t ð1:34Þ
16 FUNDAMENTALS OF RADIO COMMUNICATIONS

The second definition is the coherence bandwidth, Bc , which describes a

bandwidth over which the frequency correlation function is above 0.5 or 50%, or:

Bc  0:2s1
t ð1:35Þ

There is not any single exact relationship between coherence bandwidth and rms delay
spread, and equations (1.34) and (1.35) are only approximate equations [1–6,7–10].

Doppler Spread and Coherence Time. To obtain information about the time
varying nature of the channel caused by movements, from either the transmitter/
receiver or scatterers located around them, new parameters, such as the Doppler
spread and the coherence time, are usually introduced to describe the time variation
phenomena of the channel in a small-scale region. The Doppler spread BD is defined
as a range of frequencies over which the received Doppler spectrum is essentially
nonzero. It shows the spectral spreading caused by the time rate of change of the
mobile radio channel due to the relative motions of vehicles (and scatterers around
them) with respect to the base station. According to [1–4,7–10], the Doppler spread
BD depends on the Doppler shift fD and on the angle a between the direction of
motion of any vehicle and the direction of arrival of the reflected and/or scattered
waves (see Fig. 1.3). If we deal with the complex baseband signal presentation,
then we can introduce the following criterion: If the baseband signal bandwidth is
greater than the Doppler spread BD , the effects of Doppler shift are negligible at the
receiver.
Coherence time Tc is the time domain dual of Doppler spread, and it is used
to characterize the time varying nature of the frequency dispersiveness of the
channel in time coordinates. The relationship between these two-channel charac-
teristics is:
1 l
Tc  ¼ ð1:36Þ
fm v

We can also define the coherence time according to [1–4,7–10] as the time duration
over which two multipath components of receiving signal have a strong potential for
amplitude correlation. One can also define the coherence time as the time over
which the correlation function of two various signals in the time domain is above 0.5
(or 50%). Then according to [7,10] we get

9 9l l
Tc  ¼ ¼ 0:18 ð1:37Þ
16pfm 16pv v

This definition is approximate and can be improved for modern digital

communication channels by combining Equations (1.36) and (1.37) as the geometric
mean between the two, this yields
0:423 l
Tc  ¼ 0:423 ð1:38Þ
fm v
 MAIN PROPAGATION CHARACTERISTICS 17

Ts Bc

t f
Tc Bs

FIGURE 1.7. Comparison between signal and channel parameters.

The definition of coherence time implies that two signals arriving at the receiver
with a time separation greater than Tc are affected differently by the channel.

1.4.5. Types of Fading in Multipath Communication Channels

Let us now summarize the effects of fading, which may occur in static or dynamic
multipath communication channels.

Static Channel. In this case multipath fading is purely spatial and leads to con-
structive or destructive interference at various points in space, at any given instant
in time, depending on the relative phases of the arriving signals. Furthermore, fading
in the frequency domain does not change because the two antennas are stationary.
The signal parameters, such as the signal bandwidth, Bs, the time of duration, Ts,
with respect to the coherent time, Bc, and the coherent bandwidth, Tc, of the
channel are shown in Figure 1.7. There are two types of fading that occur in the static
channels:

A. Flat slow fading (FSF) (see Fig. 1.8), where the following relations between
signal parameters of the signal and a channel are valid [7–10]:
0:02
Tc  Ts ; 0 ffi BD  Bs ; st  Ts ; Bc   Bs ð1:39Þ
st
Here all harmonics of the total signal are coherent.
B. Flat fast fading (FFF) (see Fig. 1.9), where the following relations between
the parameters of a channel and the signal are valid [7–10]:

Tc  Ts ; 0 ffi BD  Bs ; st 4Ts ; Bc  B s ð1:40Þ

Bs

t f
Ts s t
Bc

FIGURE 1.8. Relations between parameters for flat slow fading.

18 FUNDAMENTALS OF RADIO COMMUNICATIONS

Bc

t f
Ts st
Bs

FIGURE 1.9. Relations between parameters for flat fast fading.

Dynamic Channel. There are two different types of fading also that occur in a
dynamic (mobile) channel:

A. Frequency selective fast fading (FSFF) (see Fig. 1.10), when fast fading
depends on the frequency. In this case following relations between the
parameters of a channel and the signal are valid [7–10]:

Tc  Ts ; BD  Bs ; st  T s ; Bc  B s ð1:41Þ

B. Frequency selective slow fading (FSSF) (see Fig. 1.11), when slow fading
depends on the frequency. Therefore, the following relations between the
parameters of a channel and the signal are valid [7–10]:

Tc > Ts ; BD < Bs ; st < Ts ; Bc  Bs ð1:42Þ

Using these relationships between the parameters of the signal and of a channel, we
can a priori define the type of fading which may occur in a wireless communication
link (see Fig. 1.12).

noise

Tc Bs

t f
Ts BD = 2 f m

noise

Bc

t f
Ts st Bs

FIGURE 1.10. Relations between parameters for frequency selective fast fading.
 MAIN PROPAGATION CHARACTERISTICS 19

Ts BD

t f
Tc Bs

Bs

t f
Ts st Bc

FIGURE 1.11. Relations between parameters for frequency selective slow fading.

st Time Domain

s t > TS FFSF FSSF

(Frequency Selective (Frequency Selective
Fast Fading) Slow Fading)
TS

s t < TS FFF FSF

(Flat Fast Fading) (Flat Slow Fading)

TC < TS TS TC > TS TC

Frequency Domain
BC

BC > BS FSF FFF

(Flat Slow Fading) (Flat Fast Fading)

BS

FSSF FFSF
BC < BS (Frequency Selective (Frequency Selective
Slow Fading) Fast Fading)

BD < BS BS BD > BS BD

FIGURE 1.12. Common picture of different kinds of fading, depending on the relations
between the signal and the channel main parameters.
20 FUNDAMENTALS OF RADIO COMMUNICATIONS

1.5. PROBLEMS IN ADAPTIVE ANTENNAS APPLICATION

The main problem with land communication links is estimating the ratio between the
coherent and multipath components of the total signal. That is, the Ricean parameter
K, to predict the effects of multiplicative noise in the channel of each subscriber
located in different conditions in the terrestrial environment. This is shown in
Figure 1.13 for various subscribers numbered by i ¼ 1, 2, 3, . . . .
However, even a detailed prediction of the radio propagation situation for each
subscriber cannot completely resolve all issues of effective service and increase
quality of data stream sent to each user. For this purpose, in future generations of
wireless systems, adaptive or smart antenna systems are employed to reduce
interference and decrease bit error rate (BER). This topic will be covered in detail in
Chapter 8. We present schematically the concept of adaptive (smart) antennas in
Figure 1.14.

K6

Shopping centre
K4
T
House
K5

Building 1 Building 1

House

House
K3
K2
K1 Building 2
House

Park
House
Tree

Building 1
Tree Factory
Tree

FIGURE 1.13. Scheme of various scenarios in urban communication channel.

 BIBLIOGRAPHY 21

K6
R
K4
T
K5

K3
K1 R
R
K2
R

FIGURE 1.14. A scheme for using adaptive antennas for each user located in different
conditions in a service area.

Even with smart antennas (see Chapter 8), we cannot totally cancel the effects of
the environment, especially in urban areas, due to the spread of the antenna beam
(see Fig. 1.14). Chapters 5 and 10 will focus on terrain effects where a thorough
analysis of these effects on the design of wireless system will be presented.

BIBLIOGRAPHY

[1] Jakes, W. C., Microwave Mobile Communications, New York, John Wiley and Son, 1974.
[2] Steele, R., Mobile Radio Communication, IEEE Press, 1992.
[3] Stuber, G. L., Principles of Mobile Communications, Boston-London, Kluwer Academic
Publishers, 1996.
[4] Lee, W. Y. C., Mobile Cellular Telecommunications Systems, McGraw Hill, New York,
1989.
[5] Blaunstein, N., Radio Propagation in Cellular Networks. Artech Houses, Boston-
London, 1999.
[6] Blaunstein, N., and J. B. Andersen, Multipath Phenomena in Cellular Networks, Artech
Houses, Boston-London, 2002.
[7] Saunders, S. R., Antennas and Propagation for Wireless Communication Systems,
John Wiley & Sons, New York, 1999.
[8] Bertoni, H. L., Radio Propagation for Modern Wireless Systems, Prentice Hall PTR,
New Jersey, 2000.
[9] Blaunstein, N., Chapter 12, Wireless Communication Systems, Hanbook of Engineering
Electromagnetics, Edited by Rajeev Bansal, Marcel Dekker, NY, 2004.
[10] Rappaport, T. S., Wireless Communications, New York, Prentice Hall PTR, 1996.
[11] Leon-Garcia, A., Probability and Random Processes for Electrical Engineering,
New York: Addison-Wesley Publishing Company, 1994.
[12] Stark, H., and Woods J. W., Probability, Random Processes, and Estimation Theory for
Engineers, New Jersey, Prentice Hall, 1994.