Appendix A

A Simple Introduction to Antennas

A.1 Introduction: Radiation Resistance and Radiation Patterns

An antenna is a transitional device, or transducer, that forms an interface for energy

traveling between a circuit and free space, as depicted in Fig. A.1. It is reciprocal
in the sense that it can transfer energy from the circuit to free space (transmission)
and from free space to a circuit (reception). We can represent both the circuit and
the antenna by their Thévenin equivalents as shown.
In engineering we represent the conversion of energy from electrical to some
other non-recoverable form by a resistive load, since the resistor is an element that
absorbs real power. We do the same for antennas; the radiation resistance Rr shown
in Fig. A.1 models how much power is taken from the transmitter circuit and is
radiated non-recoverably into free space. Alternately it is the source resistance of the
antenna when it receives a signal, in which case the antenna model will also include
a generator to represent the source of energy it is providing to the receiver circuit.
A reactive component Xr of the antenna model is seen in Fig. A.1. Ideally that
should be zero because it signifies that energy reflects back from the antenna to the
circuit. In the design of an antenna, one object is to make its radiation impedance
real because then there can be a smooth transition from the circuit to free space.
Ideally the radiation resistance of the antenna should match the output resistance
of the circuit, and the circuit’s output reactance should be zero, so that maximum
power transfer can occur without reflection. In practice it is often difficult to achieve
such a match so tuning circuits are sometimes employed between the circuit and the
antenna.
Now consider the construction of the antenna. If there were no antenna and the
circuit terminated in an open circuited transmission line, then theoretically all the
power from the transmitter would be reflected backwards along the line. If the end of
the line were flared, or even terminated in a dipole arrangement as shown in Fig. A.2,
then a sizable proportion of the energy traveling forward along the transmission line
from the transmitter circuit will detach and radiate into free space.
The dipole arrangement shown in Fig. A.2 is a very common form of antenna,
particularly if its length from tip to tip is equivalent to a half wavelength of the sig-
nal being radiated. We can deduce some of its properties qualitatively, particularly
in relation to the directions in which it radiates. For example, if we walked around a
107
108 A A Simple Introduction to Antennas

circuit antenna

V,I E,H

Z = R + jX

Thévenin equivalent of Thévenin equivalent

the transmitter circuit of the antenna

Zr = Rr + jXr

Fig. A.1 The antenna as an interface between a circuit and free space, along with their Thévenin
equivalent circuits; the subscript r on the antenna model stands for “radiation”

radiation into free space

dipole

Fig. A.2 Use of a flared horn

or a dipole as a transition from
a circuit to free space
A.1 Introduction: Radiation Resistance and Radiation Patterns 109

Fig. A.3 The radiation

pattern of a dipole antenna

viewed horizontally viewed vertically

vertically deployed dipole antenna in the horizontal plane it would look no different
when viewed from any angle. Therefore we would conclude that its radiating prop-
erties will not vary with angle around the dipole. If however we moved around it
vertically its aspect would change from its appearance in Fig. A.2 to a single point
when viewed directly from above or below. We can conclude therefore that its radi-
ating properties will vary with vertical angle and, indeed, it may not radiate at all in
the vertical direction. That is in fact the case for a dipole; it has a radiation pattern
or polar pattern of the form of a doughnut as depicted in Fig. A.3.
As would be imagined, the radiation pattern of an antenna can be quite com-
plicated. That leads to a number of definitions that are helpful in describing and
antenna’s properties. Figure A.4 shows a typical pattern in one plane, remembering
that the full pattern will be a three dimensional figure. It is plotted in Cartesian rather
than polar coordinates, which is often the case in practice.

main lobe
–3dB

half power beamwidth (HPBW)

first side lobe

beamwidth between
first nulls (BWFN)

back lobe

–180 –120 –60 0 60 120 180

angle from main beam

Fig. A.4 Radiation pattern of a fictional antenna plotted in Cartesian coordinates

110 A A Simple Introduction to Antennas

A.2 The Directivity and Gain of an Antenna

It is now important to describe the antenna quantitatively. We commence with the

definition of its directivity which, as the name implies, is a measure of how much it
concentrates the energy in a certain direction. This is a three dimensional concept
since an antenna, in principle, radiates in all directions. Geometric definitions for
the derivation of an expression for directivity are given in Fig. A.5, from which it
can be seen that the power available over an incremental area α at distance r from
the antenna is given by
Pt α Pt Ω
P = pα = =
4π r 2 4π
giving as the power available per unit of solid angle – also known as angular power
density
P Pt
p= = = p(θ , φ ) Wsr−1 (A.1)
Ω 4π
Equation (A.1) is an algebraic expression for the three dimensional polar pattern.
To proceed to a definition of directivity it is normalized with respect to its maximum
to give
p(θ , φ )
pn (θ , φ ) =
pmax (θ , φ )
Integrating this dimensionless, normalized quantity over three dimensions we
end up with the angular quantity

ΩA = pn (θ , φ )d θ d φ (A.2)
4π

α = r 2Ω

Pt
p=
4πr 2
r

θ
Ω (solid angle)

Fig. A.5 Coordinates and φ

definitions for calculating the
directivity of an antenna
A.2 The Directivity and Gain of an Antenna 111

which is called the solid beam angle of the antenna. It is the three dimensional angle
through which all the power of the antenna would be transmitted if its polar pattern
were uniform over that angle as depicted in Fig. A.6.
We now define the directivity of the antenna as
4π
D= (A.3)
ΩA
which can be approximated
4π
D= with angles in radians, or
θHP φHP
41253
D= with angles in degrees.
θHP φHP
The gain, G, of an antenna is closely related to its directivity. They differ only
through the efficiency, k, of the antenna, which accounts for ohmic losses in the
antenna material. Thus
G = kD 0 < k < 1
Although (A.3) allows the directivity and thus gain of an antenna to be derived
theoretically it is more usual to measure an antenna’s gain. That is done relative
to a reference antenna. One reference, even though not experimentally practical, is
the isotropic radiator, which has a directivity and thus gain of unity (see Sect. 1.2).
In principle, the gain determined from (A.3) is with respect to isotropic. It is more
usually expressed in decibels with respect to isotropic, and written as
4π
G = 10 log = 10 log 4π − 10 log(θHP φHP ) = 11 − 10 log(θHP φHP ) dBi
ΩA
assuming the antenna is ideally efficient.
It is possible to calculate the gain of the dipole antenna when its length is equal
to half a wavelength. It is, of course, also possible to construct a half wave dipole,

pn(θ,φ)

ΩA

if pn(θ,φ) = 1 Ω∈ΩA
= 0 elsewhere

∫∫ pn(θ,φ)dθ dφ ≡ ΩA
4π
then∫∫ pn(θ,φ)dθ dφ = ΩA
4π

Fig. A.6 Demonstrating the definition of the angular beamwidth of an antenna

112 A A Simple Introduction to Antennas

so that it can be used as a reference antenna when measuring the gain of another
antenna. Such a measurement is undertaken by transmitting to the antenna of in-
terest and measuring the received signal, and in fact the full radiation pattern. The
experimental antenna is then replaced by the reference dipole and the measurement
repeated so that the measurement of the experimental antenna can be normalized to
the dipole. The measured gain is then expressed as dB with respect to the half wave
dipole, written as dBλ/2
The calculated gain of the dipole is 2.16 dBi; thus we have

G dBi = G dBλ/2 + 2.16dB

A.3 The Aperture of an Antenna

Derivation of expressions for the aperture of an antenna requires a field theory anal-
ysis. For some antennas, such as a parabolic dish reflector, the aperture concept is
straightforward and, provided the dish diameter is much larger than a wavelength,
the aperture is related directly to the area presented to an incoming wave front. For
other antennas, such as linear structures and even an isotropic radiator, the aper-
ture concept is less straightforward but can, if we know its gain, be determined
from

λ2 G
A= m2 (A.4)
4π
If the antenna has a physical aperture, such as a parabolic reflector, we introduce
the concept of aperture efficiency to account for the difference between the physical
and electromagnetic apertures:

A = kaperture Aphysical 0 < kaperture < 1

A.4 Radiated Fields

A treatment of the fields radiated by an antenna requires a field theory treatment and
is beyond this coverage. It is, however, useful to examine well-known expressions
for the fields produced by a so-called short dipole because the fields generated by
other antennas can be derived from the short dipole results; it also allows us to
understand the concept of near and far fields. Figure A.7 shows the geometry of a
short dipole in which distance and direction out from the antenna is described by
the radial coordinate r.
If the short dipole is carrying a sinusoidal current

Io e jω t
A.4 Radiated Fields 113

Fig. A.7 The short dipole

short
dipole r

θ
l<<λ Io
φ

and is so short that there is no distribution of current along its length at any time,
then the complete set of fields generated about the short dipole is
 
Io le j(ω t−β r) cos θ 1 1
Er = + (A.5a)
2πεo cr2 jω r3
 
Io le j(ω t−β r) sin θ jω 1 1
Eθ = + + (A.5b)
4πεo c2 r cr2 jω r3
 
Io le j(ω t−β r) sin θ jω 1
Hφ = + 2 (A.5c)
4π cr r

In other words there are transverse components (θ , φ ) of the magnetic and electric
fields. There is also a radial electric field component (r) – i.e. in the direction of
propagation. Note however it has a stronger inverse dependence on distance than
the transverse components so that if the distance is sufficiently large it disappears
and the transverse components themselves become just inverse distance dependent.
This is demonstrated by letting r go large in (A.5a–c) to give

Er = 0 (A.6a)
jω Io le j(ω t−β r) sin θ
Eθ = (A.6b)
4πεo c2 r
jω Io le j(ω t−β r) sin θ
Hφ = (A.6c)
4π cr
Thus for large distances the wave is TEM – i.e. transverse electromagnetic. Equa-
tions (A.6a–c) describe the so-called far field of the antenna. The far fields are in-
verse distance dependent and the treatment in this book, based on simple power
and power density relationships, is valid. In contrast, closer to the antenna (A.5a–c)
are needed to describe the field. That is called the near field of the antenna. The
transition from near to far field is said to occur when the inverse distance terms in
114 A A Simple Introduction to Antennas

(A.5b,c) are equal to the inverse distance squared terms, assuming that any inverse
cubic terms are then negligible. Therefore the near field/far field transition is when
 ω   1 
   
  = 2
cr r

which gives
λ
r≈ .
6
It is of interest to note from (A.6b,c) that in the far field
  √
 Eθ 
  = 1 = μo εo = Zo
 Hφ  εo c εo

the free space impedance, as would be expected.

Gain Radiation Impedance

Half Wave
Dipole λ 1.64 (2.15dBi) 73 + j42.5
2
70 + j0
if antenna slightly shorter
Quarter Wave
Monopole
λ 3.3 (5.19dBi) 36.5 + j21.3
4

Short l<<λ
Monopole 2
l
3 (4.77dBi) 10π − jX (large)
λ

Folded
Dipole
λ 1.64 (2.15dBi) 292
2

Folded
Monopole
λ
3.3 (5.19dBi) 146
4

Fig. A.8 Some common linear antennas

A.5 Some Typical Antennas 115

A.5 Some Typical Antennas

Figure A.8 shows a number of simple antennas and their characteristics. Figure A.9
shows two common, compound antennas that are built up from combinations of
active antennas, of the types shown in Fig. A.8, and passive linear elements.
The folded antennas shown in Fig. A.8 tend to have slightly broader bandwidths
than their unfolded counterparts and are often used, particularly the dipole, in more
complex structures such as the Yagi-Uda array illustrated in Fig. A.9. The short
monopole in Fig. A.8 is commonly used as an AM receiving antenna on motor
vehicles.
The log periodic antenna shown in Fig. A.9 is used when operation is necessary
over a wide band of frequencies. Although it is more complex in construction than
the Yagi, its wide operating bandwidth makes it attractive in many applications.
Figure A.10 shows a number of aperture and slot antennas, along with a bi-cone.
Aperture reflectors tend to be used when the wavelength is much smaller than the
diameter of the reflector, so they behave somewhat similar to optical reflectors of
the same type.

Gain Radiation Impedance

Yagi Uda Array

λ λ
2 π

8L determined by driven element

λ

reflector driven directors

element
L

Log Periodic Dipole Array

~5 (7dBi)
various

lk+1
=τ
lk

Fig. A.9 Some common compound antennas; the antenna lengths and spacings for the log periodic
antenna are in the constant ratio τ as indicated for length
116 A A Simple Introduction to Antennas

slot

horn

hyperbolic
secondary
reflector

prime focus parabolic reflector cassegrain parabolic reflector

bi-cone

Fig. A.10 Aperture, slot and bi-conical antennas; the bi-cone is broad band and omni-directional
in the horizontal plane

balanced antenna

balanced
feed

transmitter unbalanced transmission line

balun

Fig. A.11 Demonstrating the use of a balun to provide the matched transition between an unbal-
ance transmission line and a balanced antenna
A.6 Baluns 117

A.6 Baluns

With the exception of monopoles, the other antennas in Figs. A.8 and A.9 require
balanced feeds. In other words they need to be fed by transmission lines that have
neither conductor at earth potential. Yet many of the feed lines in practice are coaxial
cables that clearly have one of their conductors – the braid – at earth potential.
Coaxial cables are also compatible with many transmitter output circuits that are
also unbalanced as noted by the manner in which the Thévenin equivalent is depicted
in Fig. A.1. To render the unbalanced transmission line compatible with a balanced
antenna a device referred to as a balun is employed, as illustrated in Fig. A.11. Short
for balanced-unbalanced, this device can be constructed in several forms, each of
which not only has to perform the unbalanced-to-balanced transformation but also
has to match impedances for maximum power flow and to minimise reflections.
There are many forms of balun, the simplest of which is a transformer. For narrow
band operation a simple balun can be constructed from sections of transmission line.
Appendix B
The Use of Decibels in Communications
Engineering

Logarithms have two major benefits: they readily summarise numbers that extend
over a large range and they simplify multiplication. As a consequence, the decibel
(dB), which is defined using base 10 logarithms, is widely used as a convenient
measure in many branches of engineering, but especially in communications. Al-
though it can be used with signals generally, it is principally defined in terms of
power (or power density). More precisely, the decibel (dB) is defined on the basis
of a reference power:
P
10 log10 = x dB
Pre f
We say that P is x dB larger than Pref . For example
If P = then x is
2Pref 3 dB wrt Pref
10Pref 10 dB wrt Pref
100Pref 20 dB wrt Pref
0.5Pref −3 dB wrt Pref
0.1Pref −10 dB wrt Pref

Many factors have easily constructed dB equivalents. For example

200 = 2 × 100 → 3 dB + 20 dB = 23 dB,

as a result of the additive property of logarithms.

Similarly

17 dB = 20 dB − 3 dB → 100 ÷ 2 = 50,
36 dB = 30 dB + 6 dB → 1000 × 4 = 4000.

In telecommunications, two common values of Pref are used. The dBs are then
given special symbols that imply absolute, as against relative, quantities.

119
120 B The Use of Decibels in Communications Engineering

If Pref = 1 W, then we use dBW.

If Pref = 1 mW, then we use dBm.

Thus we can see the following equivalences:

17 dBm → 50 mW 23 dBW → 200 W
3 dBm → 2 mW 10 dBW → 10 W
30 dBm → 1 W 20 dBW → 100 W
0 dBm → 1 mW 0 dBW → 1W
-20 dBm → 10μW -40 dBW → 100μW
Decibels can also be used with voltages, but the definition still rests upon power.
For example
P V2 V
10 log = 10 log10 2 = 20 log = x dB
Pre f Vre f V re f

So that if V = 2Vref , then x = 6 dB.

Appendix C
The Dielectric Constant of an Ionospheric Layer

Equation (3.1) notes that the refractive index of an ionospheric layer is given by

81N
n = 1− 2
f

We derive that expression below, following the approach of D.J. Angelakos and
T.E. Everhart, Microwave Communication, McGraw-Hill, New York, 1968.
An ionised region of the atmosphere, such as one of the layers of the ionosphere,
will be composed of free ions, electrons and neutral molecules. We assume that the
ions, because of their mass, do not respond as well to the passage of an electro-
magnetic field as the electrons and thus have little effect on it. We will therefore
concentrate our attention just on the free electrons, which we assume to be present
with density N electrons per cubic metre.
We also assume that the earth’s magnetic field has no effect, and that the colli-
sions that occur between electrons and neutral atmospheric constituent molecules
can be neglected. Those collisions are significant if we are interested in the attenua-
tion of a wave in transmission through the atmosphere; we mention that below, after
the derivation of refractive index.
The response of an individual electron of mass m to an applied electric field
|E| = Em cos ω t Vm−1 (resulting from the passage of a radio wave) is given from
Newton’s law
F = ma
If the charge on the electron is e and its velocity is v then this last expression can be
written
dv
eEm cos ω t = m
dt
which gives for the electron velocity
eEm
v= sin ω t.
ωm

121
122 C The Dielectric Constant of an Ionospheric Layer

This movement of electrons gives rise to a conduction current described by the trans-
verse areal current density

jcond = veN Am−2

e2 NEm
= sin ω t (C.1)
ωm
There will also be a displacement areal current density as a result of the dielectric
behaviour of the medium, found from
dD dE
jdis = =ε (C.2)
dt dt
in which D is the electric displacement vector and ε is the permittivity of the
medium.
For a plasma as dilute as an ionospheric layer ε = εo , so that

d |E|
|jdis | = εo = −εo ω Em sin ω t
dt
Thus the magnitude of the total current in the layer induced by the passage of a
radio wave is

e2 N
|jtotal | = |jcond + jdis | = ( − εo ω )Em sin ω t (C.3)
ωm
Even though there are free electrons present, we now regard the layer as behaving
entirely as a dielectric with permittivity ε – i.e. as though there were no free elec-
trons. The displacement current density is then given just by (C.2). If we call that
an effective displacement current density and equate it to the actual current density
given by (C.3) we have using (C.1)
 
jeffective,dis  = −εω Em sin ω t = |jcond + jdis |

so that

e2 N
−εω = ( − εo ω )
ωm
or
2
ε = εo (1 − ω p ω 2 ) (C.4)
in which
 1/2
e2 N
ωp = (C.5)
mεo
is called the plasma frequency of the region with electron density N.
C The Dielectric Constant of an Ionospheric Layer 123

Since

e = 1.6 × 10−19 C
m = 9.11 × 10−31 kg
εo = 8.85 pFm−1

then
ω p2 = 3175N
Thus (C.4) becomes
3175N 81N
ε = εo (1 − ) = εo (1 − 2 ) (C.6)
ω2 f
from which we recognise that the equivalent dielectric constant (or relative permit-
tivity) of the region is
81N
εrel = (1 − )
f2
Thus the refractive index of a region of the ionosphere of electron density N and
frequency f is

81N
n = 1− 2 (C.7)
f
Recall that this expression, and (C.6), was derived by ignoring losses resulting
from electron-neutral collisions. If they were included Rohan1 shows that (C.6)
would be
 
e2 N
ε = εo 1 − (C.8)
mεo (ω 2 + ν 2 )
in which ν is the collision frequency of the electrons and neutrals. While the colli-
sion frequency is very high in the lower atmosphere because of the neutral density, it
is of the order of 1000 or less at the height of the upper ionospheric layers. As a con-
sequence, at the sorts of frequencies normally associated with sky wave propagation
ω 2  ν 2 , so that (C.8) reduces to (C.6).
Electron-neutral collisions give rise to losses in the ionosphere, particularly at
the lower levels; their effect can be characterised by an equivalent conductivity from
which an attenuation constant can be derived. Again, following Rohan, the conduc-
tivity of a region of ionisation is

e2 N ν
σ=
m(ω 2 + ν 2 )

1 P. Rohan, Introduction to Electromagnetic Wave Propagation, Artech, Boston, 1991.

124 C The Dielectric Constant of an Ionospheric Layer

The attenuation constant is then

60πσ 60π e2 N ν
α= =
n nm(ω 2 + ν 2 )

in which n is refractive index. Thus the attenuation of a layer decreases with an

increase in operating frequency.
To obtain an idea of the levels of attenuation likely to be encountered by a wave
travelling through the D region (above its critical frequency) suppose we choose
typical values of ν = 107 s−1 , N = 108 electrons m−3 and f = 1 MHz (ie the AM
broadcast band). After substituting we have

α = 3.8 × 10−5 Npm−1

= 0.33 dBkm−1

Thus if the layer were equivalently 25 km thick at that effective electron den-
sity then the total attenuation at 1 MHz would be 8.25 dB at vertical incidence and
considerably more at oblique incidence. In the evening such a high level of attenu-
ation does not occur because of the absence of the D region. As a consequence it is
possible to receive distant AM stations in the evening that are not available during
daylight hours.
Index

Angular power density, 110 Balun, 117

Antenna Band designators, 7
aperture, 2, 112 microwave, 8
aperture efficiency, 112 Brewster angle, 99
back lobe, 109
beamwidth between first nulls, 109 Cellular radio systems, 79
bi-cone, 116 Co-channel interference, 80
cassegrain reflector, 116 Conductivity, 15, 19
directivity, 110 Critical frequency, 25, 30
folded dipole, 114
folded monopole, 114 Dielectric constant, 5, 88
gain, 2, 111 complex, 20, 89
half power beamwidth, 109 effect of moisture content, 94
half wave dipole, 107, 114 Diffraction, 45
horn, 116 gain, 47
Log periodic array, 115 Direct ray, 8, 39
Diversity, 72
main lobe, 109
antenna, 74
polar pattern, 109
frequency, 72
prime focus parabolic reflector, 116
space, 73
quarter wave monopole, 114
time, 73
radiation impedance, 107 Ducting, 52
radiation pattern, 109
radiation resistance, 107 Earth gravitational constant, 76
short dipole, 112 Effective earth radius, 44, 51
short monopole, 13, 14, 114 Effective isotropically
side lobe, 109 radiated power, 12
slot, 116 Electron density, 23
Yagi-Uda array, 115 Escape ray, 29
Atmosphere Extra-ordinary ray, 35
modified refractive index, 52
refractive index, 48 Fade margin, 71
standard, 48 Far field, 113
Atmospheric attenuation, 54 Free space path loss, 7
Attenuation constant, 88 Frequency re-use, 11, 79
in waveguide, 103 Fresnel zone, 48
Automatic gain control, 84 Friis’ Noise Formula, 63

125
126 Index

G-to-T ratio, 78 Noise temperature, 59

Geostationary orbit, 75 cascaded two ports, 63
Ground current, 9, 14 equivalent input noise
Ground reflected ray, 8, 39 temperature, 60
Group velocity, 33, 90 equivalent output noise temperature, 61
two port, 59
Huygens’ principle, 45
Omega-beta diagram, 90
Impedance of free space, 4, 97 Optimum working
Inverse distance law, 4 frequency, 36
Ionogram, 30
Ionosonde, 30 Passive reflectors, 73
Ionosphere, 8 Permeability
D region, 18, 24 absolute, 5
daytime, 24 relative, 5, 87
E layer, 24 Permittivity
F layer, 25 absolute, 5
F1 layer, 24 complex, 94
F2 layer, 24 relative, 5, 123
night time, 25 Phase constant, 88
refractive index, 26 in waveguide, 105
sporadic E layer, 25 Phase velocity, 32, 89
Ionospheric sounding, 30 Plasma frequency, 122
Ionospheric wave, 10, 23 Polarisation
Isotropic radiator, 2 circular, 3
dielectric, 92
Lambertian scattering, 100 elliptical, 3
Loss horizontal, 3
atmospheric absorption, 7 parallel, 97
dielectric, 102 perpendicular, 97
diffraction, 7, 46 vertical, 3
free space path loss, 7 Power density, 2
rainfall, 7 Poynting vector, 4
refraction, 7 Propagation constant, 19, 88
Loss tangent, 94
Quasi-conductor, 91
Maximum usable frequency, 35
MUF factor, 35 Radar, 12
Multi-path, 39, 72 primary, 12
secondary, 12
Near field, 113 Radar cross section, 12, 74
Neper, 88 Rainfall attenuation, 53
Noise, 57 Rayleigh criterion, 99
additive, 57 Rayleigh distribution, 84
atmospheric, 72 Receiver figure of merit, 78
environmental, 58 Reflection coefficient, 40, 97, 98, 105
galactic, 72 Refraction
Johnson, 58 atmospheric, 17, 45, 48
multiplicative, 57 Refractive index, 5
shot, 58 Rician distribution, 84
thermal, 58
voltage, 66 Shadow zone, 45
Noise bandwidth, 59 Signal to noise ratio, 57, 61
Noise figure, 61 Skin depth, 21
Index 127

Skip distance, 35 T factors, 35

Sky wave, 10, 23 Transmission coefficient, 97
range, 36 Troposcattering, 45
Space wave, 8, 39 Tunnels, 102
atmospheric attenuation, 54
Virtual height, 23, 31
ducting, 52
field strength, 41
Wave impedance, 97
rainfall effects, 53 Waveguide, 103
range, 43 circular, 103
refraction, 48 cut-off frequency, 104
Surface wave, 10, 13 evanescent attenuation, 104
attenuation factor, 14 rectangular, 103