Introduction into Antenna Studies

(Definition and circuit theory description. Brief historical notes. General

review of antenna geometries and arrangements. Wireless vs. cable
communication systems. The radio-frequency spectrum.)

One of the most important issues in the design of high-power transmission

systems is the matching of the antenna to the transmission line (TL) and the
generator. Matching is specified most often in terms of VSWR. Standing
waves are to be avoided because they may cause arching or discharge in the
TL. The resistive/dielectric losses are undesirable, too. They decrease the
efficiency factor of the antenna.

1. Definition and circuit theory description.

The antenna (aerial, EM radiator) is a device, which radiates or

receives electromagnetic waves.
The antenna is the transition between a guiding device (transmission line,
waveguide) and free space (or another usually unbounded medium). Its main
purpose is to convert the energy of a guided wave into the energy of a freespace wave (or vice versa) as efficiently as possible, while in the same time
the radiated power has a certain desired pattern of distribution in space.
a) transmission-line Thevenin equivalent circuit of a radiating
(transmitting) system

Rrad

Zg

Vg

IA
jX A

Zc
RL
Generator Transmission
Line

Antenna

Vg

- voltage-source generator (transmitter);

Zg

- impedance of the generator (transmitter);

Rrad

- radiation resistance (related to the radiated power as

Prad = I A2 Rrad )
RL
- loss resistance (related to conduction and dielectric losses);
jX A - antenna reactance.
Antenna impedance: Z A = ( Rrad + RL ) + jX A

b) transmission-line Thevenin equivalent circuit of a receiving antenna

system
ZA
IA
ZL

VA

Zc

Receiver Transmission
Line

Antenna

The antenna is a critical component in a wireless communication system.

A good design of the antenna can relax system requirements and improve its
overall performance.

2. Brief historical notes.

James Clerk Maxwell formulates the mathematical
model
of
electromagnetism
(classical
electrodynamics), A Treatise on Electricity and
Magnetism, 1873. He shows that light is an
electromagnetic (EM) wave, and that all EM waves
(light included) propagate through space with the
same speed, which depends on the dielectric and
the magnetic properties of the medium.
Heinrich Rudolph Hertz demonstrates in 1886 the
first wireless EM wave system: a / 2 -dipole is
excited with a spark; it radiates predominantly at
about 8 m; a spark appears in the gap of a
receiving loop. In 1890, he publishes his memoirs
on electrodynamics, simplifying the form of the
electromagnetic equations, replacing all potentials
by field strengths, and deducing Ohm's, Kirchhoff's
and Coulomb's laws.
May 7, 1895, the first wireless telegraph message is
successfully transmitted, received, and deciphered.
A brilliant Russian scientist, Alexander Popov (also
spelled Popoff, Poppov), sends a message from a
Russian Navy ship 30 miles out in sea, all the way
to his lab in St. Petersburg, Russia. The Russian
Navy declares Popov's historical accomplishment
top secret. The title Father of Radio goes to G.
Marconi.
Guglielmo Marconi (the Father of Radio) sends
signals over large distances. In 1901, he performs
the first transatlantic transmission from Poldhu in
Cornwall, England, to Newfoundland, Canada. The
receiving antenna in Newfoundland was a 200meter wire pulled and supported by a kite. The
transmitting antenna in England consisted of 50
wires, supported by two 60-meter wooden poles.

The beginning of 20th century (until WW2) marks the boom in wireantenna technology (dipoles and loops) and in wireless technology as a
whole, which is largely due to the invention of the DeForest triode tube,
used as radio-frequency generator. Radio links are possible up to UHF
(about 500 MHz) and over thousands of kilometers.
WW2 marks a new era in wireless communications and antenna
technology. The invention of new microwave generators (magnetron and
klystron) leads to the development of the microwave antennas such as
waveguide apertures, horns, reflectors, etc.

3. General review of antenna geometries and arrangements.

3.1. Single-element radiators.
A. Wire radiators (single-element)
wire antenna elements

straight-wire elements
(dipoles/monopoles)

loops

helices

There is a variety of shapes corresponding to each group. For example,

loops can be circular, square, rhombic, etc. Wire antennas are simple to make
but their dimensions are commensurable with the wavelength. This limits the
frequency range of their applicability (at most 1-2 GHz). At low frequencies,
these antennas become increasingly large.

Aperture antennas (single element)

C. Printed antennas
PRINTED PATCH RADIATORS

(a) Pyramidal horn

(b) Conical horn

Aperture antennas were

developed before and
during the WW2 together
with
the
emerging
waveguide
technology.
Waveguide transmission
lines
were
primarily
developed to transfer highpower microwave EM
signals
(centimeter
wavelengths), generated
by powerful microwave
sources
such
as
magnetrons and klystrons.
These types of antennas
are preferable in the
frequency range from 1 to
20 GHz.

(a) Rectangular patch

Modern
integrated
antennas often use
multi-layer
designs
with a feed coupled to
the radiator electromagnetically
(no
galvanic contact).

or
(b) Circular patch

(c) Printed dipole

(c) Open rectangular waveguide

The patch antennas

consist of a metallic
patch etched on a
dielectric
substrate,
which has a grounded
metallic plane at the
opposite side. They
are developed in the
beginning of 1970s.
There is great variety
of geometries and
ways of excitation.

director

Classical and not so classical shapes used to form a radiating patch:

driver
dielectric substrate

quarter-wave
transformer

top layer

reflector

bottom layer

microstrip

2.56

0.64

x2

x3
0.64

1.92

x1

top layer

x3
0.64

x1

1.92

reflector

bottom layer
(d) double-layer printed Yagi antenna with
electromagnetically-coupled feed

PRINTED SLOT RADIATORS

(a)

(b)

(c)

(d)

(e)

(f)

Slot
antennas
were
developed in the 1980s and
there is still intensive research
related to new shapes and types
of excitation. They are suited
for integration with slot-line
circuits, which are usually
designed
to
operate
at
frequencies > 10 GHz.
Both patch and slot antennas
share some common features.
They are easy and cheap to
fabricate. They are easy to
mount; they are light and
mechanically robust. They
have low cross-polarization
radiation. Their directivity is
not very high. They have
relatively high conducting and
dielectric
losses.
These
radiators are widely used in
patch/slot arrays, which are
esp. convenient for use in
spacecraft, satellites, missiles,
cars
and
other
mobile
applications.

D. Leaky-wave antennas
These are antennas derived from millimeter-wave (mm-wave) guides,
such as dielectric guides, microstrip lines, coplanar and slot lines. They are
developed for applications at frequencies > 30 GHz, infrared frequencies
included. Periodical discontinuities are introduced at the end of the guide that
lead to substantial radiation leakage (radiation from the dielectric surface).

(a)

(b)

The antennas in the mm-wave band are of big variety and are still a subject of
intensive study.

(g)

(h)

Reflector antennas

The Arecibo Observatory Antenna System

A reflector is used to concentrate the EM energy in a focal point where

the receiver/feed is located. Optical astronomers have long known that a
parabolic cylinder mirror transforms rays from a line source on its focal line
into a bundle of parallel rays. Reflectors are usually parabolic (paraboloidal).
Actually, the first use of a parabolic (cylinder) reflector was used for radio
waves by Heinrich Hertz in 1888. Rarely, corner reflectors are used.
Reflector antennas have very high gain and directivity. Typical applications:
radio telescopes, satellite telecommunications. They are not easy to fabricate
and, in their conventional technology, they are rather heavy. They are not
mechanically robust.
The largest radio telescopes:
Max Plank Institt fr Radioastronomie radio telescope, Effelsberg
(Germany), 100-m paraboloidal reflector
National Astronomy and Ionosphere Center (USA) radio telescope
in Arecibo (Puerto Rico), 1000-ft (304.8-m) spherical reflector
The Green Bank Telescope (the National Radio Astronomy
Observatory) paraboloid of aperture 100 m
TYPICAL REFLECTORS

E. Lens antennas
Lenses play a similar role to that of reflectors in reflector antennas. They
collimate divergent energy into more or less plane EM wave. Lenses are
often preferred to reflectors at higher frequencies (f > 100 GHz). They are
classified according to their shape and the material they are made of.

3.2. Antenna arrays

Antenna arrays consist of multiple (usually identical) radiating elements.
Arranging the radiating elements in arrays allows achieving unique radiation
characteristics, which cannot be obtained through a single element. The
careful choice and control of the phase shift and the amplitude of the signal
fed to each element allows the change of the radiation pattern electronically
electronic scanning. Such arrays are called phased arrays. The design and the
analysis of antenna arrays is a subject of its own, which is also related to
signal processing. Intensive research goes on nowadays, concerning smart
antennas, signal-processing antennas, tracking antennas, etc. Some
commonly met arrays are shown in the figure below.

4. Wireless vs. cable communication systems.

There are two broad categories of communication systems: those that

utilize transmission lines as interconnections (cable systems), and those that
use EM radiation with an antenna at both the transmitting and the receiving
end (wireless systems).
In areas of high density population, the cable systems are economically
preferable, especially when broadband communication is in place. Even for
narrow-band communication, such as voice telephony and low-data-rate
digital transmission, it is much simpler and cheaper to build wire networks
with twisted-pair cables, when many users are to be interconnected. Such
lines introduce an attenuation of around 2-3 dB/km at frequencies about 10
kHz. These lines are not suitable at higher frequencies because of the higher
losses and dispersion.
At higher-frequency carriers, carrying broadband signals (TV
transmission and high-data-rate digital transmission), coaxial cables are
commonly used. The loss is around 4-5 dB/km.
The least distortion and losses are offered by the optical-fiber
transmission lines, which operate at three different wavelengths: 850 nm (
2.3 dB/km), 1300 nm ( 0.25 dB/km) and 1550 nm ( 0.25 dB/km). Optical
fibers are relatively expensive and the respective transmitting/receiving
equipment is also costly. Transmission lines provide a measure of security
and noise-suppression (coaxial, optical-fiber), but they are not the best option
in many cases (long distance, wide spread over large areas, less frequency
dispersion).
A fundamental feature of all transmission lines is the exponential increase
of loss power. Thus, if the loss is 5 dB/km, then a 20-km line will have 100
dB power loss (input power is reduced by a factor of 10-10), a 40-km line will
have a 200 dB power loss. This makes it rather obvious why wireless systems
are preferred for long-range communications, and in scarcely populated
areas. In most wireless channels, the radiated power per unit area decreases
as the inverse square of the distance r between the transmitting and the
receiving point. Doubling the distance r would decrease the received power
by a factor of 4 (or 6 dB will be added). Thus, if a particular system has a 100
dB loss at r=20 km, doubling the distance will result in 106 dB loss (as
compared to 200 dB loss in a cable system). The comparison between the

coaxial-line losses and free-space attenuation at f=100 MHz is given in the

figure below.

5. The radio-frequency spectrum.

Modern personal mobile communications services
Table 1.1: General designation of frequency bands
cordless telephony

home environment
businesses PABX (Private Automatic Branch
Exchange)
PHS (Personal Handyphone System) in Japan

Northern America: PCS-1900 (Personal

digital (and analog) Communication Services)
cellular telephony Europe: GSM-900 (Global System for Mobile
Communications) and
DCS-1800 (Digital Communications Systems)
mobile data
transport

packet-switched data transfer (MOBITEX, DataTAC,

etc.)
PCS two-way data communications; paging

personal satellite
communications

INMARSAT, EUTELTRACS, Iridium, Globalstar

global navigation
systems

GPS, GLONASS, ODYSSEY

Frequency band
3-30 kHz

EM wavelength
100-10 km

Designation
Very Low Frequency
(VLF)

Services
Navigation, sonar, submarine

30-300 kHz

10-1 km

Low Frequency (LF)

Radio beacons, navigation

300-3000 kHz

1000-100 m

Medium Frequency
(MF)

AM broadcast, maritime/ coastguard radio

3-30 MHz

100-10 m

High Frequency
(HF)

Telephone, telegraph, fax; amateur

radio, ship-to-coast and ship-toaircraft communication

30-300 MHz

10-1 m

Very High Frequency

(VHF)

TV, FM broadcast, air traffic

control, police, taxicab mobile radio

300-3000 MHz

100-10 cm

Ultrahigh Frequency
(UHF)

TV, satellite, radiosonde, radar

3-30 GHz

10-1 cm

Super high Frequency

(SHF)

Airborne radar, microwave links,

satellite, land mobile
communication

30-300 GHz

10-1 mm

Extremely High
Frequency (EHF)

Radar, experimental

Table 2.1: Microwave-band designation

Frequency
500-1000 MHz
1-2 GHz
2-3 GHz
3-4 GHz
4-6 GHz
6-8 GHz
8-10 GHz
10-12.4 GHz
12.4-18 GHz
18-20 GHz
20-26.5 GHz
26.5-40 GHz

Besides, there is a variety of special application of wireless technology in

radar systems
microwave relay links
satellite systems (TV, telephony, military)
radio astronomy
biomedical engineering, etc.

Old
VHF
L
S
S
C
C
X
X
Ku
K
K
Ka

New
C
D
E
F
G
H
I
J
J
J
K
K

Sonar (an acronym for Sound, Navigation and Ranging) is a system for underwater detection and location of objects by
acoustical echo. The first sonars, invented during World War I by British, American and French scientists, were used to
locate submarines and icebergs. Sonar is an American term dating from World War II.

Antenna Parameters

Antenna Pattern - a graphical representation of the antenna radiation

properties as a function of position (spherical coordinates).
Common Types of Antenna Patterns
Power Pattern - normalized power vs. spherical coordinate position.
Field Pattern - normalized E or H vs. spherical coordinate
position.
Antenna Field Types
Reactive field - the portion of the antenna field characterized by
standing (stationary) waves which represent stored energy.
Radiation field - the portion of the antenna field characterized by
radiating (propagating) waves which represent transmitted
energy.
Antenna Field Regions
Reactive Near Field Region - the region immediately surrounding
the antenna where the reactive field (stored energy - standing
waves) is dominant.
Near-Field (Fresnel) Region - the region between the reactive nearfield and the far-field where the radiation fields are dominant
and the field distribution is dependent on the distance from the
antenna.
Far-Field (Fraunhofer) Region - the region farthest away from the
antenna where the field distribution is essentially independent
of the distance from the antenna (propagating waves).

Antenna Field Regions

Antenna Pattern Definitions

Isotropic Pattern - an antenna pattern defined by uniform radiation
in all directions, produced by an isotropic radiator (point
source, a non-physical antenna which is the only nondirectional
antenna).
Directional Pattern - a pattern characterized by more efficient
radiation in one direction than another (all physically realizable
antennas are directional antennas).
Omnidirectional Pattern - a pattern which is uniform in a given
plane.
Principal Plane Patterns - the E-plane and H-plane patterns of a
linearly polarized antenna.
E-plane - the plane containing the electric field vector
and the direction of maximum radiation.
H-plane - the plane containing the magnetic field vector
and the direction of maximum radiation.
Antenna Pattern Parameters
Radiation Lobe - a clear peak in the radiation intensity surrounded
by regions of weaker radiation intensity.
Main Lobe (major lobe, main beam) - radiation lobe in the direction
of maximum radiation.
Minor Lobe - any radiation lobe other than the main lobe.
Side Lobe - a radiation lobe in any direction other than the
direction(s) of intended radiation.
Back Lobe - the radiation lobe opposite to the main lobe.

Half-Power Beamwidth (HPBW) - the angular width of the main

beam at the half-power points.
First Null Beamwidth (FNBW) - angular width between the first
nulls on either side of the main beam.

Antenna Pattern Parameters

(Normalized Power Pattern)

Maxwells Equations
(Instantaneous and Phasor Forms)
Maxwells Equations (instantaneous form)

Average Power Radiated by an Antenna

To determine the average power radiated by an antenna, we start with
the instantaneous Poynting vector (vector power density) defined by
(V/m A/m = W/m2)
Assume the antenna is enclosed by some surface S.
S

s
ds
- instantaneous vectors [ = (x,y,z,t), etc.]
instantaneous
scalar
t

Maxwells Equations (phasor form, time-harmonic form)

The total instantaneous radiated power rad leaving the surface S is found
by integrating the instantaneous Poynting vector over the surface.
E, H, D, B, J - phasor vectors [E=E(x,y,z), etc.]
- phasor scalar

Relation of instantaneous quantities to phasor quantities ...

(x,y,z,t) = Re{E(x,y,z)e j t}, etc.

ds = (

rad
S

) ds

ds = differential surface
s = unit vector normal to ds

ds = s ds

For time-harmonic fields, the time average instantaneous Poynting

vector (time average vector power density) is found by integrating the
instantaneous Poynting vector over one period (T) and dividing by the
period.
1

Pavg =

TT

Radiation Intensity
Radiation Intensity - radiated power per solid angle (radiated power
normalized to a unit sphere).
Prad

) dt

Pavg ds
S

In the far field, the radiation electric and magnetic fields vary as 1/r and
the direction of the vector power density (Pavg) is radially outward. If we
assume that the surface S is a sphere of radius r, then the integral for the
total time-average radiated power becomes

= Re{Ee j t}
= Re{He j t}
The instantaneous magnetic field may be rewritten as
= Re{ [ He j t + H*e

j t

]}

which gives an instantaneous Poynting vector of

Re {[E H]e j2 t + [E H *]}
~~~~~~~~~~~~~~~ ~~~~~~~

If we defined Pavg r 2 = U( , ) as the radiation intensity, then

time-harmonic
independent of time
(integrates to zero over T )

and the time-average vector power density becomes

Pavg =

Re [E
2T

H ] dt
T

where d = sin d d defines the differential solid angle. The units on the
radiation intensity are defined as watts per unit solid angle. The average
radiation intensity is found by dividing the radiation intensity by the area
of the unit sphere (4 ) which gives

H *]

= Re [E

The total time-average power radiated by the antenna (Prad) is found by

integrating the time-average power density over S.
Prad

Pavg ds = Re
S

[E
S

H *] ds

The average radiation intensity for a given antenna represents the radiation
intensity of a point source producing the same amount of radiated power
as the antenna.

Directivity

Inserting the expression for Prad into the directivity expression yields

Directivity (D) - the ratio of the radiation intensity in a given direction

from the antenna to the radiation intensity averaged over all
directions.

The maximum directivity is

The directivity of an isotropic radiator is D( , ) = 1.
The maximum directivity is defined as [D( , )]max = Do.
The directivity range for any antenna is 0 D( , ) Do.
Directivity in dB

where the term A in the previous equation is defined as the beam solid
angle and is defined by

Directivity in terms of Beam Solid Angle

We may define the radiation intensity as

where Bo is a constant and F( , ) is the radiation intensity pattern

function. The directivity then becomes

and the radiated power is

Beam Solid Angle - the solid angle through which all of the antenna
power would flow if the radiation intensity were [U( , )]max for all
angles in A.

Example (Directivity/Beam Solid Angle/Maximum Directivity)

Determine the directivity [D( , )], the beam solid angle A and the
maximum directivity [Do] of an antenna defined by F( , ) =
sin2 cos2 .

In order to find [F( , )]max, we must solve

Directivity/Beam Solid Angle Approximations

Given an antenna with one narrow major lobe and negligible radiation
in its minor lobes, the beam solid angle may be approximated by

MATLAB m-file for plotting this directivity function

where 1 and 2 are the half-power beamwidths (in radians) which are
perpendicular to each other. The maximum directivity, in this case, is
approximated by

for i=1:100
theta(i)=pi*(i-1)/99;
d(i)=7.5*((cos(theta(i)))^2)*((sin(theta(i)))^2);
end
polar(theta,d)

If the beamwidths are measured in degrees, we have

90
2
120

60

1.5

Example (Approximate Directivity)

150

30

A horn antenna with low side lobes has half-power beamwidths of

29o in both principal planes (E-plane and H-plane). Determine the
approximate directivity (dB) of the horn antenna.

0.5

180

330

210

300

240
270

Numerical Evaluation of Directivity

The maximum directivity of a given antenna may be written as

Note that the assumption of a separable radiation intensity pattern function

results in the product of two separate integrals for the radiated power. We
may employ a variety of numerical integration techniques to evaluate the
integrals. The most straightforward of these techniques is the rectangular
rule (others include the trapezoidal rule, Gaussian quadrature, etc.) If we
first consider the -dependent integral, the range of is first subdivided
into N equal intervals of length

The known function f ( ) is then evaluated at the center of each

subinterval. The center of each subinterval is defined by

The area of each rectangular sub-region is given by

where U( ) = Bo F( , ). The integrals related to the radiated power in
the denominators of the terms above may not be analytically integrable.
In this case, the integrals must be evaluated using numerical techniques.
If we assume that the dependence of the radiation intensity on and is
separable, then we may write

The radiated power integral then becomes

The overall integral is then approximated by

The approximate radiated power for antennas that are omnidirectional with
respect to [ f( ) = 1] reduces to

Using the same technique on the -dependent integral yields

For antennas which have a radiation intensity which is not separable in
and , the a two-dimensional numerical integration must be performed
which yields

Example (Numerical evaluation of directivity)

Combining the
and
dependent integration results gives the
approximate radiated power.

The approximate radiated power for antennas that are omnidirectional with
respect to [g( ) = 1] reduces to

Determine the directivity of a half-wave dipole given the radiation

intensity of

Antenna Efficiency
The maximum value of the radiation intensity for a half-wave dipole
occurs at = /2 so that

When an antenna is driven by a voltage source (generator), the total

power radiated by the antenna will not be the total power available from
the generator. The loss factors which affect the antenna efficiency can be
identified by considering the common example of a generator connected
to a transmitting antenna via a transmission line as shown below.

Zg - source impedance
ZA - antenna impedance
Zo - transmission line characteristic impedance
MATLAB m-file

Pin - total power delivered to the antenna terminals

sum=0.0;
N=input(Enter the number of segments in the theta direction)
for i=1:N
thetai=(pi/N)*(i-0.5);
sum=sum+(cos((pi/2)*cos(thetai)))^2/sin(thetai);
end
D=(2*N)/(pi*sum)

Pohmic - antenna ohmic (I2R) losses

[conduction loss + dielectric loss]

Do

1.6428

10

1.6410

15

1.6409

20

1.6409

Prad - total power radiated by the antenna

The total power delivered to the antenna terminals is less than that
available from the generator given the effects of mismatch at the source/tline connection, losses in the t-line, and mismatch at the t-line/antenna
connection. The total power delivered to the antenna terminals must equal
that lost to I2R (ohmic) losses plus that radiated by the antenna.

We may define the antenna radiation efficiency (ecd) as

connection. The reflection efficiency is easily found from transmission

line theory in terms of the reflection coefficient ( ).

which gives a measure of how efficient the antenna is at radiating the

power delivered to its terminals. The antenna radiation efficiency may be
written as a product of the conduction efficiency (ec) and the dielectric
efficiency (ed).
The total antenna efficiency then becomes
ec - conduction efficiency (conduction losses only)
ed - dielectric efficiency (dielectric losses only)
However, these individual efficiency terms are difficult to compute so that
they are typically determined by experimental measurement. This antenna
measurement yields the total antenna radiation efficiency such that the
individual terms cannot be separated.
Note that the antenna radiation efficiency does not include the
mismatch (reflection) losses at the t-line/antenna connection. This loss
factor is not included in the antenna radiation efficiency because it is not
inherent to the antenna alone. The reflection loss factor depends on the tline connected to the antenna. We can define the total antenna efficiency
(eo), which includes the losses due to mismatch as

The definition of antenna efficiency (specifically, the antenna radiation

efficiency) plays an important role in the definition of antenna gain.

Antenna Gain
The definitions of antenna directivity and antenna gain are essentially
the same except for the power terms used in the definitions.
Directivity [D( , )] - ratio of the antenna radiated power density at a
distant point to the total antenna radiated power (Prad) radiated
isotropically.
Gain [G( , )] - ratio of the antenna radiated power density at a distant
point to the total antenna input power (Pin) radiated isotropically.

eo - total antenna efficiency (all losses)

er - reflection efficiency (mismatch losses)
The reflection efficiency represents the ratio of power delivered to the
antenna terminals to the total power incident on the t-line/antenna

Thus, the antenna gain, being dependent on the total power delivered to the
antenna input terminals, accounts for the ohmic losses in the antenna while
the antenna directivity, being dependent on the total radiated power, does
not include the effect of ohmic losses.

The equations for directivity and gain are

Antenna Impedance
The complex antenna impedance is defined in terms of resistive (real)
and reactive (imaginary) components.

RA - Antenna resistance
[(dissipation ) ohmic losses + radiation]
The relationship between the directivity and gain of an antenna may be
found using the definition of the radiation efficiency of the antenna.

XA - Antenna reactance
[(energy storage) antenna near field]
We may define the antenna resistance as the sum of two resistances which
separately represent the ohmic losses and the radiation.

Rr - Antenna radiation resistance (radiation)

RL - Antenna loss resistance (ohmic loss)
Gain in dB

The typical transmitting system can be defined by a generator,

transmission line and transmitting antenna as shown below.

The generator is modeled by a complex source voltage Vg and a complex

source impedance Zg.

In some cases, the generator may be connected directly to the antenna.

The power radiated by the antenna (Pr) may be written as

Inserting the complete source and antenna impedances yields

The power dissipated as heat (PL ) may be written

The complex power associated with any element in the equivalent circuit
is given by

where the * denotes the complex conjugate. We will assume peak values
for all voltages and currents in expressing the radiated power, the power
associated with ohmic losses, and the reactive power in terms of specific
components of the antenna impedance. The peak current for the simple
series circuit shown above is

The reactive power (imaginary component of the complex power) stored

in the antenna near field (PX) is

From the equivalent circuit for the generator/antenna system, we see that
maximum power transfer occurs when

The power dissipated in the generator resistance is

Transmitting antenna system summary (maximum power transfer)

The circuit current in this case is
Power dissipated in
the generator [P/2]
Power available from
the generator [P]

Power dissipated by the

antenna [(1 ecd)(P/2)]

The power radiated by the antenna is

Power delivered to
the antenna [P/2]

Power radiated by the

antenna [ecd (P/2)]

The power dissipated in heat is

The power available from the generator source is

With an ideal transmitting antenna (ecd = 1) given maximum power

transfer, one-half of the power available from the generator is radiated by
the antenna.

The typical receiving system can be defined by a generator (receiving

antenna), transmission line and load (receiver) as shown below.

The circuit current in this case is

The power captured by the receiving antenna is

Assuming the receiving antenna is connected directly to the receiver

Some of the power captured by the receiving antenna is re-radiated
(scattered). The power scattered by the antenna (Pscat) is

The power dissipated by the receiving antenna in the form of heat is

The power delivered to the receiver is

For the receiving system, maximum power transfer occurs when

Receiving antenna system summary (maximum power transfer)

Antenna Radiation Efficiency

The radiation efficiency (ecd) of a given antenna has previously been
defined in terms of the total power radiated by the antenna (Prad) and the
total power dissipated by the antenna in the form of ohmic losses (Pohmic).

Power delivered to
the receiver [P/2]
Power dissipated by the
antenna [(1 ecd)(P/2)]

Power captured by
the antenna [P]
Power delivered to
the antenna [P/2]

Power scattered by the

antenna [ecd (P/2)]

With an ideal receiving antenna (ecd = 1) given maximum power transfer,

one-half of the power captured by the antenna is re-radiated (scattered) by
the antenna.

The total radiated power and the total

ohmic losses were determined for the
general case of a transmitting antenna
using the equivalent circuit. The total
radiated power is that dissipated in
the antenna radiation resistance (Rr).

The total ohmic losses for the antenna are those dissipated in the antenna
loss resistance (RL).

Inserting the equivalent circuit results for Prad and Pohmic into the equation
for the antenna radiation efficiency yields

Thus, the antenna radiation efficiency may be found directly from the
antenna equivalent circuit parameters.

If we define the perimeter distance of the conductor as dp, then the HF

resistance of the conductor can be written as

Antenna Loss Resistance

The antenna loss resistance (conductor and dielectric losses) for many
antennas is typically difficult to calculate. In these cases, the loss
resistance is normally measured experimentally. However, the loss
resistance of wire antennas can be calculated easily and accurately.
Assuming a conductor of length l and cross-sectional area A which carries
a uniform current density, the DC resistance is

where is the conductivity of the conductor. At high frequencies, the

current tends to crowd toward the outer surface of the conductor (skin
effect). The HF resistance can be defined in terms of the skin depth .

where

where Rs is defined as the surface resistance of the material.

For the RHF equation to be accurate, the skin depth should be a small
fraction of the conductor maximum cross-sectional dimension. In the case
of a cylindrical conductor (dp 2 a), the HF resistance is

is the permeability of the material and f is the frequency in Hz.

R

f
0

The skin depth for copper ( = 5.8107 /m,

be written as

= 4 10 7 H/m) may

RDC = 0.818 m

1 kHz

2.09 mm

10 kHz

0.661 mm

RHF = 1.60 m

100 kHz

0.209 mm

RHF = 5.07 m

1 MHz

0.0661
mm

RHF = 16.0 m

Resistance of 1 m of #10 AWG (a = 2.59 mm) copper wire.

The high frequency resistance formula assumes that the current through the
conductor is sinusoidal in time and independent of position along the
conductor [Iz(z,t) = Io cos( t)]. On most antennas, the current is not
necessarily independent of position. However, given the actual current
distribution on the antenna, an equivalent RL can be calculated.
Example (Problem 2.44) [Loss resistance calculation]
A dipole antenna consists of a circular wire of length l. Assuming the
current distribution on the wire is cosinusoidal, i.e.,

Equivalent circuit equation

(uniform current, Io - peak)
Integration of incremental
power along the antenna

Thus, the loss resistance of a dipole antenna of length l is one-half that of

a the same conductor carrying a uniform current.

Lossless Transmission Line Fundamentals

Transmission line equations (voltage and current)

~~~~~~~
+z directed
waves

~~~~~~~
z directed
waves

Transmitting/Receiving Systems with Transmission Lines

The power available from the generator is

The power delivered to the transmission line input terminals is

Using transmission line theory, the impedance seen looking into

the input terminals of the transmission line (Zin) is
The power associated with the generator impedance is

The resulting equivalent circuit is shown below.

Given the current and the voltage at the input to the transmission line, the
values at any point on the line can be found using the transmission line
equations.

The current and voltage at the transmission line input terminals are
The unknown coefficient Vo+ may be determined from either V(0) or I(0)
which were found in the input equivalent circuit. Using V(0) gives

MATLAB m-file (generator/t-line/load)

where

Given the coefficient Vo+, the current and voltage at the load, from the
transmission line equations are

Vg=input(Enter the complex generator voltage

);
Zg=input(Enter the complex generator impedance
);
Zo=input(Enter the lossless t-line characteristic impedance );
l=input(Enter the lossless t-line length in wavelengths
);
Zl=input(Enter the complex load impedance
);
j=0+1j;
betal=2*pi*l;
Zin=Zo*(Zl+j*Zo*tan(betal))/(Zo+j*Zl*tan(betal));
gammal=(Zl-Zo)/(Zl+Zo);
gamma0=gammal*exp(-j*2*betal);
Ig=Vg/(Zg+Zin);
Pg=0.5*Vg*conj(Ig);
V0=Ig*Zin;
P0=0.5*V0*conj(Ig);
Vcoeff=V0/(1+gamma0);
Vl=Vcoeff*exp(-j*betal)*(1+gammal);
Il=Vcoeff*exp(-j*betal)*(1-gammal)/Zo;
Pl=0.5*Vl*conj(Il);
s=(1+abs(gammal))/(1-abs(gammal));
format compact
Generator_voltage=Vg
Generator_current=Ig
Generator_power=Pg
Generator_impedance_voltage=Vg-V0
Generator_impedance_current=Ig
Generator_impedance_power=Pg-P0
T_line_input_voltage=V0
T_line_input_current=Ig
T_line_input_power=P0
T_line_input_impedance=Zin
T_line_input_reflection_coeff=gamma0
T_line_standing_wave_ratio=s
Load_voltage=Vl
Load_current=Il
Load_power=Pl
Load_reflection_coeff=gammal

Given Vg = (10+j0) V, Zg = (100+j0)

The power delivered to the load is then

The complexity of the previous equations leads to solutions which are

typically determined by computer or Smith chart.

Zo

ZL

Zin

100

75

96+j28

100

100

100

and l = 5.125 , the following results are found.

Pg

P(l)

0.1429

0.25

1.3333

0.1224

100

0.25

0.125

125

98 j22

0.1111

0.25

1.25

0.1235

75

100

72 j21

0.1429

0.2864

1.3333

0.1199

100

100

100

0.25

0.125

125

100

122+j27

0.1111

0.2219

1.25

0.1219

(0) =

(l)

Antenna Polarization
The polarization of an plane wave is defined by the figure traced by
the instantaneous electric field at a fixed observation point. The following
are the most commonly encountered polarizations assuming the wave is
approaching.

The polarization of the antenna in a given direction is defined as the

polarization of the wave radiated in that direction by the antenna. Note
that any of the previous polarization figures may be rotated by some
arbitrary angle.
Polarization loss factor
Incident wave polarization

Antenna polarization

Polarization loss factor (PLF)

PLF in dB

General Polarization Ellipse

The instantaneous components of the electric field are found by

multiplying the phasor components by e j t and taking the real part.
x

(z,t)

(z,t)

The relative positions of the instantaneous electric field components on the

general polarization ellipse defines the polarization of the plane wave.
Linear Polarization
If we define the phase shift between the two electric field
components as

we find that a phase shift of

defines a linearly polarized wave.

x

The vector electric field associated with a +z-directed plane wave can
be written in general phasor form as

(z,t)

(z,t)

Examples of linear polarization:

where Ex and Ey are complex phasors which may be defined in terms of
magnitude and phase.

Linear polarization in the x-direction ( = 0)

If Eyo = 0
If Exo = 0
Linear polarization in the y-direction ( = 90o)
Linear polarization ( = 45o)
If Exo = Eyo and n is even
Linear polarization ( = 135o)
If Exo = Eyo and n is odd

Circular Polarization

Antenna Equivalent Areas

If Exo = Eyo and

Antenna Effective Aperture (Area)

Given a receiving antenna oriented for maximum response,
polarization matched to the incident wave, and impedance matched to its
load, the resulting power delivered to the receiver (Prec) may be defined in
terms of the antenna effective aperture (Ae) as

then
x

(z,t)

(z,t)
where S is the power density of the incident wave (magnitude of the
Poynting vector) defined by

This is left-hand circular polarization.

If Exo = Eyo and
According to the equivalent circuit under matched conditions,

then
x

(z,t)

(z,t)

This is right-hand circular polarization.

Elliptical Polarization
We may solve for the antenna effective aperture which gives
Elliptical polarization follows definitions as circular polarization
except that Exo Eyo.
Exo
Exo

Eyo,
Eyo,

= (2n+)
= (2n+)

left-hand elliptical polarization

right-hand elliptical polarization

Antenna Scattering Area

Antenna Capture Area

The total power scattered by the receiving antenna is defined as the

product of the incident power density and the antenna scattering area (As).

The total power captured by the receiving antenna (power delivered

to the load + power scattered by the antenna + power dissipated in the form
of heat) is defined as the product of the incident power density and the
antenna capture area (Ac).

From the equivalent circuit, the total scattered power is

The total power captured by the antenna is

which gives
which gives

Antenna Loss Area

The total power dissipated as heat by the receiving antenna is defined
as the product of the incident power density and the antenna loss area
(AL).

From the equivalent circuit, the total dissipated power is

which gives

Note that Ac = Ae + As + AL.

Maximum Directivity and Effective Aperture

Assume the transmitting and receiving antennas are lossless and
oriented for maximum response.

These two equations yield

or

If the transmit antenna is an isotropic radiator, we will later show that

Aet, Dot - transmit antenna effective aperture and maximum directivity

Aer, Dor - receive antenna effective aperture and maximum directivity

which gives

If we assume that the total power transmitted by the transmit antenna is Pt,
the power density at the receive antenna (Wr) is
Therefore, the equivalent aperture of a lossless antenna may be defined in
terms of the maximum directivity as

The total power received by the receive antenna (Pr) is

The overall antenna efficiency (eo) may be included to account for the
ohmic losses and mismatch losses in an antenna with losses.
which gives

The effect of polarization loss can also be included to yield

If we interchange the transmit and receive antennas, the previous
equation still holds true by interchanging the respective transmit and
receive quantities (assuming a linear, isotropic medium), which gives

Friis Transmission Equation

The Friis transmission equation defines the relationship between

transmitted power and received power in an arbitrary transmit/receive
antenna system. Given arbitrarily oriented transmitting and receiving
antennas, the power density at the receiving antenna (Wr) is

account the orientation of the antenna. We may extend our previous

definition of the antenna effective aperture (obtained using the maximum
directivity) to a general effective aperture for any antenna orientation.

The total received power is then

such that the ratio of received power to transmitted power is

where Pt is the input power at the terminals of the transmit antenna and
where the transmit antenna gain and directivity for the system performance
are related by the overall efficiency

where ecdt is the radiation efficiency of the transmit antenna and t is the
reflection coefficient at the transmit antenna terminals. Notice that this
definition of the transmit antenna gain includes the mismatch losses for the
transmit system in addition to the conduction and dielectric losses. A
manufacturers specification for the antenna gain will not include the
mismatch losses.
The total received power delivered to the terminals of the receiving
antenna (Pr) is

where the effective aperture of the receiving antenna (Aer) must take into

Including the polarization losses yields

For antennas aligned for maximum response, reflection-matched and

polarization matched, the Friis transmission equation reduces to

Radar Range Equation and Radar Cross Section

The Friis transmission formula can be used to determine the radar
range equation. In order to determine the maximum range at which a given
target can be detected by radar, the type of radar system (monostatic or
bistatic) and the scattering properties of the target (radar cross section)
must be known.
Monostatic radar system - transmit and receive antennas at the
same location.

Radar cross section (RCS) - a measure of the ability of a target to reflect

(scatter) electromagnetic energy (units = m2). The area which intercepts
that amount of total power which, when scattered isotropically,
produces the same power density at the receiver as the actual target.
If we define
= radar cross section (m2)
Wi = incident power density at the target (W/m2)
Pc = equivalent power captured by the target (W)
Ws = scattered power density at the receiver (W/m2)

Bistatic radar system- transmit and receive antennas at separate

locations.

According to the definition of the target RCS, the relationship between the
incident power density at the target and the scattered power density at the
receive antenna is

The limit is usually included since we must be in the far-field of the target
for the radar cross section to yield an accurate result.

The radar cross section may be written as

Showing the conduction losses, mismatch losses and polarization losses

explicitly, the ratio of the received power to transmitted power becomes

where (Ei, Hi) are the incident electric and magnetic fields at the target and
(Es, Hs) are the scattered electric and magnetic fields at the receiver. The
incident power density at the target generated by the transmitting antenna
(Pt, Gt, Dt, eot, t, at ) is given by
where
The total power captured by the target (Pc) is

aw - polarization unit vector for the scattered waves

ar - polarization unit vector for the receive antenna
Given matched antennas aligned for maximum response and polarization
matched, the general radar range equation reduces to

The power captured by the target is scattered isotropically so that the

scattered power density at the receiver is

The power delivered to the receiving antenna load is

Example
Problem 2.65 A radar antenna, used for both transmitting and
receiving, has a gain of 150 at its operating frequency of 5 GHz. It
transmits 100 kW, and is aligned for maximum directional radiation and
reception to a target 1 km away having a cross section of 3 m2. The
received signal matches the polarization of the transmitted signal. Find the
received power.

Wire Antennas
Electrical Size of an Antenna - the physical dimensions of the antenna
defined relative to wavelength.

Infinitesimal Dipole
/50, a << )
( l

Electrically small antenna - the dimensions of the antenna are small

relative to wavelength.
Electrically large antenna - the dimensions of the antenna are large
relative to wavelength.

Example Consider a dipole antenna of length L = 1m. Determine the

electrical length of the dipole at f = 3 MHz and f = 30 GHz.
We assume that the axial current along the infinitesimal dipole is
uniform. With a << , we may assume that any circumferential currents
are negligible and treat the dipole as a current filament.

f = 3 MHz
( = 100m)
Electrically small

f = 30 GHz
( = 0.01m)
Electrically large
The infinitesimal dipole with a constant current along its length is a nonphysical antenna. However, the infinitesimal dipole approximates several
physically realizable antennas.

Capacitor-plate antenna (top-hat-loaded antenna)

Inverted-L antenna

The capacitor plates can be actual conductors or simply the wire

equivalent. The fields radiated by the radial currents tend to cancel each
other in the far field so that the far fields of the capacitor plate antenna can
be approximated by the infinitesimal dipole.
Using image theory, the inverted-L antenna is equivalent to the
transmission line loaded antenna.
Transmission line loaded antenna

If we assume that L
/4, then the current along the antenna resembles
that of a half-wave dipole.

Based on the current distributions on these antennas, the far fields of the
capacitor plate antenna, the transmission line loaded antenna and the
inverted-L antenna can all be approximated by the far fields of the
infinitesimal dipole.

To determine the fields radiated by the infinitesimal dipole, we first

determine the magnetic vector potential A due to the given electric current
source J (M = 0, F = 0).

The total magnetic vector potential may then be written in vector form as

Because of the true point source nature of the infinitesimal dipole ( l

/50), the equation above for the magnetic vector potential of the
infinitesimal dipole is valid everywhere. We may use this expression for
A to determine both near fields and far fields.
The radiated fields of the infinitesimal dipole are found by
differentiating the magnetic vector potential.
The infinitesimal dipole magnetic vector potential given in the previous
equation is a rectangular coordinate vector with the magnitude defined in
terms of spherical coordinates. The rectangular coordinate vector can be
transformed into spherical coordinates using the standard coordinate
transformation.

The electric field is found using either potential theory or Maxwells

equations.
Potential Theory

Maxwells Equations (J = 0 away from the source)

Note that electric field expression in terms of potentials requires two levels
of differentiation while the Maxwells equations equation requires only one
level of differentiation. Thus, using Maxwells equations, we find

Field Regions of the Infinitesimal Dipole

We may separate the fields of the infinitesimal dipole into the three
standard regions:
Reactive near field
Radiating near field
Far field

kr << 1
kr > 1
kr >> 1

5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
0

fields radiated by an
infinitesimal dipole

10

12

Considering the bracketed terms [ ] in the radiated field expressions for the
infinitesimal dipole ...
Reactive near field (kr << 1)
Radiating near field (kr > 1)
Far field (kr >> 1)

(kr)-2 terms dominate

constant terms dominate if present
otherwise, (kr)-1 terms dominate
constant terms dominate

The vector form of the near electric field is the same as that for an
electrostatic dipole (charges +q and q separated by a distance l).

Reactive near field [ kr << 1 or r << /2 ]

When kr << 1, the terms which vary inversely with the highest power
of kr are dominant. Thus, the near field of the infinitesimal dipole is given
by

Infinitesimal dipole
near fields

Note the 90o phase difference between the electric field components and
the magnetic field component (these components are in phase quadrature)
which indicates reactive power (stored energy, not radiation). If we
investigate the Poynting vector of the dominant near field terms, we find

The Poynting vector (complex vector power density) for the infinitesimal
dipole near field is purely imaginary. An imaginary Poynting vector
corresponds to standing waves or stored energy (reactive power).

If we replace the term (Io /k) by in the near electric field terms by its
charge equivalent expression, we find

The electric field expression above is identical to that of the electrostatic

dipole except for the complex exponential term (the infinitesimal dipole
electric field oscillates). This result is related to the assumption of a
uniform current over the length of the infinitesimal dipole. The only way
for the current to be uniform, even at the ends of the wire, is for charge to
build up and decay at the ends of the dipole as the current oscillates.
The near magnetic field of the infinitesimal dipole can be shown to
be mathematically equivalent to that of a short DC current segment
multiplied by the same complex exponential term.

Radiating near field [ kr

1 or r

/2 ]

The dominant terms for the radiating near field of the infinitesimal
dipole are the terms which are constant with respect to kr for E and H
and the term proportional to (kr)-1 for Er.

Infinitesimal dipole
radiating near field

Note that E and H are now in phase which yields a Poynting vector for
these two components which is purely real (radiation). The direction of
this component of the Poynting vector is outward radially denoting the
outward radiating real power.
Far field [ kr >> 1 or r >> /2 ]
The dominant terms for the far field of the infinitesimal dipole are the
terms which are constant with respect to kr.

Infinitesimal dipole
far field

Note that the far field components of E and H are the same two
components which produced the radially-directed real-valued Poynting
vector (radiated power) for the radiating near field. Also note that there is
no radial component of E or H so that the propagating wave is a transverse
electromagnetic (TEM) wave. For very large values of r, this TEM wave
approaches a plane wave. The ratio of the far electric field to the far
magnetic field for the infinitesimal dipole yields the intrinsic impedance
of the medium.

Far Field of an Arbitrarily Oriented Infinitesimal Dipole

Given the equations for the far field of an infinitesimal dipole
oriented along the z-axis, we may generalize these equations for an
infinitesimal dipole antenna oriented in any direction. The far fields of
infinitesimal dipole oriented along the z-axis are

If we rotate the antenna by some arbitrary angle and define the new
direction of the current flow by the unit vector a , the resulting far fields
are simply a rotated version of the original equations above. In the rotated
coordinate system, we must define new angles ( , ) that correspond to the
spherical coordinate angles ( , ) in the original coordinate system. The
angle is shown below referenced to the x-axis (as is defined) but can
be referenced to any convenient axis that could represent a rotation in the
-direction.

Note that the infinitesimal far fields in the original coordinate system
depend on the spherical coordinates r and . The value of r is identical in
the two coordinates systems since it represents the distance from the
coordinate origin. However, we must determine the transformation from
to . The transformations of the far fields in the original coordinate
system to those in the rotated coordinate system can be written as

Specifically, we need the definition of sin .

trigonometric identity

According to the

we may write

Based on the definition of the dot product, the cos term may be written
as

so that

Inserting our result for the sin term yields

Example
Determine the far fields of an infinitesimal dipole oriented along the
y-axis.

Poyntings Theorem (Conservation of Power)

Poyntings theorem defines the basic principle of conservation of
power which may be applied to radiating antennas. The derivation of the
time-harmonic form of Poyntings vector begins with the following vector
identity

If we insert the Poynting vector (S = E H*) in the left hand side of the
above identity, we find

From Maxwells equations, the curl of E and H are

such that

Integrating both sides of this equation over any volume V and applying the
divergence theorem to the left hand side gives

The current density in the equation above consists of two components: the
impressed (source) current (Ji) and the conduction current (Jc).

Inserting the current expression and dividing both sides of the equation by
2 yields Poyntings theorem.

Total Power and Radiation Resistance

To determine the total complex power (radiated plus reactive)
produced by the infinitesimal dipole, we integrate the Poynting vector over
a spherical surface enclosing the antenna. We must use the complete field
expressions to determine both the radiated and reactive power. The timeaverage complex Poynting vector is

The individual terms in the above equation may be identified as

The total complex power passing through the spherical surface of radius
r is found by integrating the normal component of the Poynting vector over
the surface.

Poyntings theorem may then be written as

The terms We and Wm represent the radial electric and magnetic energy
flow through the spherical surface S.

The real and imaginary parts of the complex power are

The radiation resistance for the infinitesimal dipole is found according to

Infinitesimal dipole
radiation resistance
The total power through the sphere is

Infinitesimal Dipole Radiation Intensity and Directivity

Infinitesimal Dipole Effective Aperture and Solid Beam Angle

The radiation intensity of the infinitesimal dipole may be found by

using the previously determined total fields.

The effective aperture of the infinitesimal dipole is found from the

maximum directivity:

Infinitesimal dipole
effective aperture
The beam solid angle for the infinitesimal dipole can be found from the
maximum directivity,

or can be determined directly from the radiation intensity function.

Infinitesimal dipole
directivity function
Infinitesimal dipole
Maximum directivity

Infinitesimal dipole
beam solid angle

( /50

Short Dipole
l
/10, a << )

Note that the magnetic vector potential of the short dipole (length = l, peak
current = Io) is one half that of the equivalent infinitesimal dipole (length
l = l, current = Io).

The average current on the short dipole is one half that of the equivalent
infinitesimal dipole. Therefore, the fields produced by the short dipole are
exactly one half those produced by the equivalent infinitesimal dipole.

Short dipole
radiated fields

Short dipole
far field

Since the fields produced by the short dipole are one half those of the
equivalent infinitesimal dipole, the real power radiated by the short dipole
is one fourth that of the infinitesimal dipole. Thus, Prad for the short dipole
is

and the associated radiation resistance is

Short dipole
near fields

Short dipole
radiation resistance

The directivity function, the maximum directivity, effective area and beam
solid angle of the short dipole are all identical to the corresponding value
for the infinitesimal dipole.

Short dipole
radiating near field

Center-Fed Dipole Antenna

(a << )

field coordinates (spherical)

If we assume that the dipole antenna is driven at its center, we may

assume that the current distribution is symmetrical along the antenna.

Source coordinates (rectangular)

For the center-fed dipole lying along the z-axis, x = y = 0, so that

We use the previously defined approximations for the far field magnetic
vector potential to determine the far fields of the center-fed dipole.

Transforming the z-directed vector potential to spherical coordinates gives

(Center-fed dipole far field magnetic vector potential )

The far fields of the center-fed dipole in terms of the magnetic vector
potential are

(Center-fed dipole far field electric field)

(Center-fed dipole far field magnetic field)

The time-average complex Poynting vector in the far field of the center-fed
dipole is

l = /10

l = /2

The radiation intensity function for the center-fed dipole is given by

l=

l = 3 /2

(Center-fed dipole radiation intensity function)

We may plot the normalized radiation intensity function [U( ) = BoF( )]
to determine the effect of the antenna length on its radiation pattern.

In general, we see that the directivity of the antenna increases as the length
goes from a short dipole (a fraction of a wavelength) to a full wavelength.
As the length increases above a wavelength, more lobes are introduced into
the radiation pattern.

l = /10
1
0.9

0.8

0.8

0.7

0.7

0.6

0.6
I(z) / Io

I(z) / Io

1
0.9

0.5

0.5

0.4

0.4

0.3

0.3

0.2

0.2

0.1
0
-0.05

0.1
-0.04

-0.03

-0.02

-0.01

0
z/

0.01

0.02

0.03

0.04

0
-0.25

0.05

-0.2

-0.15

-0.1

-0.05

0
z/

0.05

0.1

0.15

0.2

l = 3 /2

l=

0.8

0.6

0.7

0.4

0.6

0.2
I(z) / Io

1
0.8

I(z) / Io

1
0.9

0.5

-0.2

0.3

-0.4

0.2

-0.6

0.1

0.25

The -dependent integral in the radiated power expression cannot be

integrated analytically. However, the integral may be manipulated, using
several transformations of variables, into a form containing some
commonly encountered special functions (integrals) known as the sine
integral and cosine integral.

0.4

0
-0.5

The total real power radiated by the center-fed dipole is

l = /2

-0.8
-1
-0.4

-0.3

-0.2

-0.1

0
z/

0.1

0.2

0.3

0.4

0.5

-0.6

-0.4

-0.2

0
z/

0.2

0.4

0.6

The radiated power of the center-fed dipole becomes

The radiated power is related to the radiation resistance of the antenna by

Center-fed dipole
directivity function
which gives
The maximum directivity is
Center-fed dipole
maximum directivity
(Center-fed dipole radiation resistance)

The effective aperture is

The directivity function of the center-fed dipole is given by

Center-fed dipole
effective aperture

Center-fed dipole
Solid beam angle

Half-Wave Dipole

Center-fed
half-wave dipole
far fields

Center-fed
half-wave dipole
radiation intensity
function

Center-fed half-wave dipole

radiation resistance (in air)

Dipole Input Impedance

Center-fed
half-wave dipole
directivity
function

The input impedance of the dipole is defined as the ratio of voltage

to current at the antenna feed point.

Center-fed half-wave dipole

maximum directivity

The real and reactive time-average power delivered to the terminals of the
antenna may be written as

Center-fed half-wave dipole

effective aperture
If we assume that the antenna is lossless (RL = 0), then the real power
delivered to the input terminals equals that radiated by the antenna. Thus,

and the antenna input resistance is related to the antenna radiation

resistance by

The dipole reactance may be determined in closed form using a technique

known as the induced EMF method (Chapter 8) but requires that the radius
of the wire (a) be included. The resulting dipole reactance is

In a similar fashion, we may equate the reactive power delivered to the

antenna input terminals to that stored in the near field of the antenna.

or

The general dipole current is defined by

The current Iin is the current at the feed point of the dipole (z = 0) so that

The input resistance and reactance of the antenna are then related to the
equivalent circuit values of radiation resistance and the antenna reactance
by

(Center-fed dipole reactance)

The input resistance and reactance are plotted in Figure 8.16 (p.411) for a
dipole of radius a = 10-5 . If the dipole is 0.5 in length, the input
impedance is found to be approximately (73 + j42.5) . The first dipole
resonance (Xin = 0) occurs when the dipole length is slightly less than onehalf wavelength. The exact resonant length depends on the wire radius, but
for wires that are electrically very thin, the resonant length of the dipole is
approximately 0.48 . As the wire radius increases, the resonant length
decreases slightly [see Figure 8.17 (p.412)].

Antenna and Scatterers

Image Theory

All of the antennas considered thus far have been assumed to be

radiating in a homogeneous medium of infinite extent. When an antenna
radiates in the presence of a conductor(inhomogeneous medium), currents
are induced on the conductor which re-radiate (scatter) additional fields.
The total fields produced by an antenna in the presence of a scatterer are
the superposition of the original radiated fields (incident fields, [E inc,H inc]
those produced by the antenna in the absence of the scatterer) plus the
fields produced by the currents induced on the scatterer (scattered fields,
[E scat,H scat]).

Given an antenna radiating over a perfect conducting ground plane,

[perfect electric conductor (PEC), perfect magnetic conductor (PMC)] we
may use image theory to formulate the total fields without ever having to
determine the surface currents induced on the ground plane. Image theory
is based on the electric or magnetic field boundary condition on the surface
of the perfect conductor (the tangential electric field is zero on the surface
of a PEC, the tangential magnetic field is zero on the surface of a PMC).
Using image theory, the ground plane can be replaced by the equivalent
image current located an equal distance below the ground plane. The
original current and its image radiate in a homogeneous medium of infinite
extent and we may use the corresponding homogeneous medium equations.

Example (vertical electric dipole)

To evaluate the total fields, we must first determine the scattered

fields which depend on the currents flowing on the scatterer. The
determination of the scatterer currents typically requires a numerical
scheme (integral equation in terms of the scatterer currents or a differential
equation in the form of a boundary value problem). However, for simple
scatterer shapes, we may use image theory to simplify the problem.

Currents over a PEC

Vertical Infinitesimal Dipole Over Ground

Give a vertical infinitesimal electric dipole (z-directed) located a
distance h over a PEC ground plane, we may use image theory to
determine the overall radiated fields.

Currents over a PMC

The individual contributions to the electric field by the original dipole and
its image are

In the far field, the lines defining r, r1 and r2 become almost parallel so that

The previous expressions for r1 and r2 are necessary for the phase terms in
the dipole electric field expressions. But, for amplitude terms, we may
assume that r1 r2 r. The total field becomes

h = 0.5

The normalized power pattern for the vertical infinitesimal dipole over a
PEC ground is

h=2

h = 0.1

h = 0.25

h=

h = 10

Since the radiated fields of the infinitesimal dipole over ground are
different from those of the isolated antenna, the basic parameters of the
antenna are also different. The far fields of the infinitesimal dipole are

The time-average Poynting vector is

(Infinitesimal dipole over ground radiation resistance)

The corresponding radiation intensity function is
The directivity function of the infinitesimal dipole over ground is

The maximum value of the radiation intensity function is found at = /2.

so that the maximum directivity (at

= /2) is given by

The radiated power is found by integrating the radiation intensity function.

(Infinitesimal dipole over ground maximum directivity)

0.8

0.7

0.6

0.5

Do

Rr ()

Given an infinitesimal dipole of length l = /50, we may plot the

radiation resistance and maximum directivity as a function of the antenna
height to see the effect of the ground plane.

0.4

0.3

0.2

0.1

0
0

0.5

1.5

2.5

3.5

4.5

First, we note the relationship between Umax for the isolated dipole and the
dipole over ground.

Note that Umax for the antenna over ground is independent of the height of
the antenna over ground.

0
0

0.5

h/

For an isolated infinitesimal dipole of length

resistance is

1.5

2.5

3.5

4.5

large

h/

l = /50, the radiation

and the maximum directivity (independent of antenna length) is Do = 1.5.

Note that Rr of the infinitesimal dipole over ground approaches twice that
of Rr for an isolated dipole as h 0 (see the relationship between a
monopole antenna and its equivalent dipole antenna in the next section).
As the height is increased, the radiation resistance of the infinitesimal
dipole over ground approaches that of an isolated dipole. The directivity
of the infinitesimal dipole over ground approaches a value twice that of the
isolated dipole as h 0 and four times that of the isolated dipole as h
grows large. This follows from our definition of the total radiated power
and maximum directivity for the isolated antenna and the antenna over
ground.

Monopole

The input impedance of the equivalent antennas is given by

The input impedance of the monopole is exactly one-half that of the

equivalent dipole. Therefore, we may determine the monopole radiation
resistance for monopoles of different lengths according to the results of the
equivalent dipole.

Infinitesimal dipole
[length = l < /50]
Using image theory, the monopole antenna over a PEC ground plane
may be shown to be equivalent to a dipole antenna in a homogeneous
region. The equivalent dipole is twice the length of the monopole and is
driven with twice the antenna source voltage. These equivalent antennas
generate the same fields in the region above the ground plane.

Infinitesimal monopole
[length = l < /100]

Short dipole
[length = l, ( /50
Short monopole
[length = l, ( /100

/10)]

/20)]

Lossless half-wave dipole

[length = l = /2]
Lossless quarter-wave monopole
[length = l = /4]

The total power radiated by the monopole is one-half that of the equivalent
dipole. But, the monopole radiates into one-half the volume of the dipole
yielding equivalent fields and power densities in the upper half space.

The directivities of the two equivalent antennas are related by

Ground Effects on Antennas

At most frequencies, the conductivity of the earth is such that the
ground may be accurately approximated by a PEC. Given an antenna
located over a PEC ground plane, the radiated fields of the antenna over
ground can be determined easily using image theory. The fields radiated
by the antenna over a PEC ground excite currents on the surface of the
ground plane which re-radiate (scatter) the incident waves from the
antenna. We may also view the PEC ground plane as a perfect reflector of
the incident EM waves. The direct wave/reflected wave interpretation of
the image theory results for the infinitesimal dipole over a PEC ground is
shown below.

Infinitesimal dipole
[length = l < /50]
Infinitesimal monopole
[length = l < /100]
Lossless half-wave dipole
[length = l = /2]
Lossless quarter-wave monopole
[length = l = /4]

~~~~~~~~~
direct wave

~~~~~~~~~~
reflected wave

At lower frequencies (approximately 100 MHz and below), the

electric fields associated with the incident wave may penetrate into the
lossy ground, exciting currents in the ground which produce ohmic losses.
These losses reduce the radiation efficiency of the antenna. They also
effect the radiation pattern of the antenna since the incident waves are not
perfectly reflected by the ground plane. Image theory can still be used for
the lossy ground case, although the magnitude of the reflected wave must
be reduced from that found in the PEC ground case. The strength of the
image antenna in the lossy ground case can be found by multiplying the
strength of the image antenna in the PEC ground case by the appropriate
plane wave reflection coefficient for the proper polarization ( V).

If we plot the radiation pattern of the vertical dipole over ground for
cases of a PEC ground and a lossy ground, we find that the elevation plane
pattern for the lossy ground case is tilted upward such that the radiation
maximum does not occur on the ground plane but at some angle tilted
upward from the ground plane (see Figure 4.28, p. 183). This alignment
of the radiation maximum may or may not cause a problem depending on
the application. However, if both the transmit and receive antennas are
located close to a lossy ground, then a very inefficient system will result.
The antenna over lossy ground can be made to behave more like an
antenna over perfect ground by constructing a ground plane beneath the
antenna. At low frequencies, a solid conducting sheet is impractical
because of its size. However, a system of wires known as a radial ground
system can significantly enhance the performance of the antenna over lossy
ground.

Monopole with a radial ground system

The radial wires provide a return path for the currents produced within the
lossy ground. Broadcast AM transmitting antennas typically use a radial

ground system with 120 quarter wavelength radial wires (3o spacing).
The reflection coefficient scheme can also be applied to horizontal
antennas above a lossy ground plane. The proper reflection coefficient
must be used based on the orientation of the electric field (parallel or
perpendicular polarization).
The Effect of Earth Curvature
Antennas on spacecraft and aircraft in flight see the same effect that
antennas located close to the ground experience except that the height of
the antenna over the conducting ground means that the shape of the ground
(curvature of the earth) can have a significant effect on the scattered field.
In cases like these, the curvature of the reflecting ground must be
accounted for to yield accurate values for the reflected waves.
Antennas in Wireless Communications
Wire antennas such as dipoles and monopoles are used extensively
in wireless communications applications. The base stations in wireless
communications are most often arrays (Ch. 6) of dipoles. Hand-held units
such as cell phones typically use monopoles. Monopoles are simple, small,
cheap, efficient, easy to match, omnidirectional (according to their
orientation) and relatively broadband antennas. The equations for the
performance of a monopole antenna presented in this chapter have
assumed that the antenna is located over an infinite ground plane. The
monopole on the hand-held unit is not driven relative to the earth ground
but rather (a.) the conducting case of the unit or (b.) the circuit board of the
unit. The resonant frequency and input impedance of the hand-held
monopole are not greatly different than that of the monopole over a infinite
ground plane. The pattern of the hand-held unit monopole is different than
that of the monopole over an infinite ground plane due to the different
distribution of currents. Other antennas used on hand-held units are loops
(Ch. 5), microstrip (patch) antennas (Ch. 14) and the planar inverted F
antenna (PIFA). In wireless applications, the antenna can be designed to

perform in a typical scenario, but we cannot account for all scatterer

geometries which we may encounter (power lines, buildings, etc.). Thus,
the scattered signals from nearby conductors can have an adverse effect on
the system performance. The detrimental effect of these unwanted
scattered signals is commonly referred to as multipath.

Antenna Arrays
Antennas with a given radiation pattern may be arranged in a pattern
(line, circle, plane, etc.) to yield a different radiation pattern.

Given an antenna array of identical elements, the radiation pattern of the

antenna array may be found according to the pattern multiplication
theorem.
Pattern multiplication theorem

Antenna array - a configuration of multiple antennas (elements)

arranged to achieve a given radiation pattern.
Linear array - antenna elements arranged along a straight line.
Circular array - antenna elements arranged around a circular
ring.

Array element pattern - the pattern of the individual array element.

Array factor - a function dependent only on the geometry of the array
and the excitation (amplitude, phase) of the elements.
Example (Pattern multiplication - infinitesimal dipole over ground)

Planar array - antenna elements arranged over some planar

surface (example - rectangular array).
Conformal array - antenna elements arranged to conform to
some non-planar surface (such as an aircraft skin).
There are several array design variables which can be changed to achieve
the overall array pattern design.
Array Design Variables
1.
2.
3.
4.
5.

General array shape (linear, circular, planar, etc.).

Element spacing.
Element excitation amplitude.
Element excitation phase.
Patterns of array elements.

Phased array - an array of identical elements which achieves a given

pattern through the control of the element excitation phasing.
Phased arrays can be used to steer the main beam of the
antenna without physically moving the antenna.

The far field of this two element array was found using image theory to be

element pattern

array factor

N-Element Linear Array

The far fields of the individual array elements are

The array factor AF is independent of the antenna type assuming all

of the elements are identical. Thus, isotropic radiators may be utilized in
the derivation of the array factor to simplify the algebra. The field of an
isotropic radiator located at the origin may be written as (assuming 2polarization)

We assume that the elements of the array are uniformly-spaced with a

separation distance d.

In the far field of the array

The overall array far field is found using superposition.

(Array factor for a uniformly-spaced N-element linear array)

The current magnitudes the array elements are assumed to be equal and the
current on the array element located at the origin is used as the phase
reference (zero phase).

Uniform N-Element Linear Array

(uniform spacing, uniform amplitude, linear phase progression)

The array factor then becomes

A uniform array is defined by uniformly-spaced identical elements

of equal magnitude with a linearly progressive phase from element to
element.

Inserting this linear phase progression into the formula for the general Nelement array gives

Below are plots of the array factor AF vs. the array phase function R as the
number of elements in the array is increased. Note that these are not
plots of AF vs. the elevation angle 2.

The function R is defined as the array phase function and is a function of

the element spacing, phase shift, frequency and elevation angle. If the
array factor is multiplied by e jR, the result is

Subtracting the array factor from the equation above gives

The complex exponential term in the last expression of the above equation
represents the phase shift of the array phase center relative to the origin.
If the position of the array is shifted so that the center of the array is
located at the origin, this phase term goes away.

Some general characteristics of the array factor AF with respect to R:

(1) [AF ]max = N at R = 0 (main lobe).
(2) Total number of lobes = N!1 (one main lobe, N!2 sidelobes).
(3) Main lobe width = 4B/N, minor lobe width = 2B/N

The array factor may be normalized so that the maximum value for any
value of N is unity. The normalized array factor is

Broadside and End-fire Arrays

The phasing of the uniform linear array elements may be chosen such
that the main lobe of the array pattern lies along the array axis (end-fire
array) or normal to the array axis (broadside array).
End-fire array
Broadside array

The nulls of the array function are found by determining the zeros of the
numerator term where the denominator is not simultaneously zero.

main lobe at 2 = 0o or 2 = 180o

main lobe at 2 = 90o

The maximum of the array factor occurs when the array phase function is
zero.

For a broadside array, in order for the above equation to be satisfied with
2 = 90o, the phase angle " must be zero. In other words, all elements of the
array must be driven with the same phase. With " = 0o, the normalized
array factor reduces to
The peaks of the array function are found by determining the zeros of the
numerator term where the denominator is simultaneously zero.
Normalized array function
Broadside array, " = 0o
The m = 0 term,

represents the angle which makes R = 0 (main lobe).

Consider a 5-element broadside array (" = 0o) as the element spacing is

varied. In general, as the element spacing is increased, the main lobe
beamwidth is decreased. However, grating lobes (maxima in directions
other than the main lobe direction) are introduced when the element
spacing is greater than or equal to one wavelength. If the array pattern
design requires that no grating lobes be present, then the array element
spacing should be chosen to be less than one wavelength.

If we consider the broadside array factor as a function of the number of

array elements, we find that, in general, the main beam is sharpened as the
number of elements increases. Below are plots of AF for a broadside array
(" = 0o) with elements separated by d = 0.258 for N = 2, 5, 10 and 20.

Using the pattern multiplication theorem, the overall array pattern is

obtained by multiplying the element pattern by the array factor. As an
example, consider an broadside array (" = 0o) of seven short vertical
dipoles spaced 0.58 apart along the z-axis.

90

90

120

60

120

60

0.8
0.6
150

30

0.4

0.5

150

30

0.2

The normalized element field pattern for the infinitesimal dipole is

180

180

element pattern

The array factor for the seven element array is

330

240

300
270

The overall normalized array pattern is

array factor

210

210

330

240

300
270

If we consider the same array with horizontal (x-directed) short dipoles, the
resulting normalized element field pattern is

Since the element pattern depends on the angle N, we must choose a value
of N to plot the pattern. If we choose N = 0o, the element pattern becomes

and the array pattern is given by

If we plot the array pattern for N = 90o, we find that the element
pattern is unity and the array pattern is the same as the array factor. Thus,
the main beam of the array of x-directed short dipoles lies along the y-axis.
The nulls of the array element pattern along the x-axis prevent the array
from radiating efficiently in that broadside direction. End-fire arrays
may be designed to focus the main beam of the array factor along the array
axis in either the 2=0o or 2=180o directions. Given that the maximum of
the array factor occurs when
in order for the above equation to be satisfied with 2 = 0o, the phase angle
" must be
For 2 = 180o, the phase angle " must be
which gives

The normalized array factor for an end-fire array reduces to

Normalized array function

End-fire array, " = Kkd

Consider a 5-element end-fire array (2 = 0o) as the element spacing is

varied. Note that the phase angle " must change as the spacing changes
in order to keep the main beam of the array function in the same direction.

If the corresponding positive phase angles are chosen, the array factor plots
are mirror images of the above plots (about 2 = 90o ). Note that the endfire array grating lobes are introduced for element spacings of d $ 0.58.

7-element array end-fire array, vertical short dipoles (d = 0.258, " = !90o)

90

90

120

60

60

0.8

The normalized array factor for the 7-element end-fire array is

0.8

0.6

0.6

150

30

150

30

0.4

0.4

0.2

0.2

180

180

0
array factor

element pattern
330

210

The overall array field pattern is

240

120

330

210

300

240

300

270

270
90

0.5
60

120
0.4
0.3

30

150
0.2
0.1
180

0
array pattern
210

330

240

300
270

7-element end-fire array, x-directed horizontal short dipoles

(d = 0.258, " = !90o)
The overall array pattern in the N = 0o plane is

Hansen-Woodyard End-fire Array

The Hansen-Woodyard end-fire array is a special array designed for
maximum directivity.
Ordinary end-fire array
Hansen-Woodyard end-fire array

" = kd

" = (kd + *)

In order to increase the directivity in a closely-spaced electrically long endfire array, Hansen and Woodyard analyzed the patterns and found that a
additional phase shift of

90

90

120

60

60

increased the directivity of the array over that of the ordinary end-fire array
given an element spacing of

0.8

0.6
150

0.6

30

150

0.4

30
0.4

0.2

0.2

180

180

0
array factor

element pattern
210

330

240

120

0.8

330

210

300

240

300

270
90

270
1

120

60
0.8

For very long arrays (N - large), the element spacing in the HansenWoodyard end-fire array approaches one-quarter wavelength. The HansenWoodyard design shown here does not necessarily produce the maximum
directivity for a given linear array but does produce a directivity larger than
that of the ordinary end-fire array [by a factor of approximately 1.79 (2.5
dB)]. The Hansen-Woodyard end-fire array design can be summarized as

0.6
150

30
0.4
0.2

180

array pattern
210

330

240

300
270

where the upper sign produces a maximum in the 2 = 0o direction and the
lower sign produces a maximum in the 2 = 180o direction. The HansenWoodyard end-fire design increases the directivity of the array at the
expense of higher sidelobe levels.

Non-Uniformly Excited, Equally-Spaced Arrays

Given a two element array with equal current amplitudes and
spacing, the array factor is

The excitation coefficients for the binomial array are given by Pascals
triangle.

For a broadside array (" = 0o) with element spacing d less than one-half
wavelength, the array factor has no sidelobes. An array formed by taking
the product of two arrays of this type gives

This array factor, being the square of an array factor with no sidelobes, also
has no sidelobes. Mathematically, the array factor above represents a 3element equally-spaced array driven by current amplitudes with ratios of
1:2:1. In a similar fashion, equivalent arrays with more elements may be
formed.
The binomial array has the special property that the array factor has no
sidelobes for element spacings of 8/2 or less. Sidelobes are introduced for
element spacings larger than 8/2.
N = 5, d = 0.58
The current coefficients of the resulting N-element array take the form of
a binomial series. The array is known as a binomial array.

Binomial
array

N = 10, d = 0.58

Array Factor - Uniform Spacing, Nonuniform Amplitude

Consider an array of isotropic elements positioned symmetrically
along the z-axis (total number of elements = P). The array factor for this
array will be determined assuming that all elements are excited with the
same current phase (N = 0o for simplicity) but nonuniform current
amplitudes. The amplitude distribution assumed to be symmetric about the
origin.
P = 2M + 1 (Odd)

P = 2M (Even)

P = 2M + 1 (Odd)

P = 2M (Even)

P = 2M + 1 (Odd)

where

Through the transformation of x = cos u, the terms may be written as a set

of polynomials [Chebyshev polynomials - Tn(x)].

P = 2M (Even)

Note that the array factors are coefficients multiplied by cosines with
arguments that are integer multiples of u. Using trigonometric identities,
these cosine functions can be written as powers of u.

Using properties of the Chebyshev polynomials, we may design arrays with

specific sidelobe characteristics. Namely, we may design arrays with all
sidelobes at some prescribed level.

Chebyshev Polynomials

Dolph-Chebyshev Array Design Procedure

(1.) Select the appropriate AF for the total number of elements (P).

(2.) Replace each cos(mu) term in the array factor by its expansion in
terms of powers of cos(u).
(3.) For the required main lobe to side lobe ratio (Ro), find xo such that

Properties of Chebyshev Polynomials

1.
2.
3.
4.
5.

Even ordered Chebyshev polynomials are even functions.

Odd ordered Chebyshev polynomials are odd functions.
The magnitude of any Chebyshev polynomial is unity or less in the
range of !1 # x #1.
Tn (1) = 1 for all Chebyshev polynomials.
All zeros (roots) of the Chebshev polynomials lie within the range of
!1 # x #1.

Using the properties of Chebyshev polynomials, we may design arrays with

all sidelobes at a prescribed level below the main beam (Dolph-Chebyshev
array). The order of the Chebyshev polynomial should be one less than the
total number of elements in the array (P!1).

(4.) Substitute cos(u) = x/xo into the array factor of step 2. This
substitution normalizes the array factor sidelobes to a peak of unity.
(5.) Equate the array factor of step 4 to TP-1(x) and determine the array
coefficients.
Example
Design a 5-element Dolph-Chebyshev array with d = 0.58 and
sidelobes which are 20 dB below the main beam.
(1.) P = 5, M = 2

(2.)

(3.)

(4.)

(5.) Equate coefficients and solve for a1, a2, and a3.

Equation Section 1
MICROSTRIP ANTENNAS PART I
(Introduction, Construction and geometry, Feeding techniques, Substrate
properties, Loss calculation)

2. Construction and Geometry

Generally the MSA are thin metallic patches of various shapes etched on
dielectric substrates of thickness h, which usually is from 0.003o to 0.05o.
the substrate is usually grounded at the opposite side.

1. Introduction
Microstrip Antenna (MSA) received considerable attention in the 1970s,
although the first designs and theoretical models appeared in the 1950s.
They are suitable for aircraft, satellite and missile applications where size,
weight, ease of installation, mechanical reliability and cost are important. The
MSA are low profile, mechanically robust, inexpensive to manufacture,
compatible with MMIC designs and relatively light and compact.
With respect to radiation properties they are versatile in terms of resonant
frequencies, polarization, pattern and impedance. They allow the use of
additional tuning elements like pins or varactor diodes between the patch and
the ground plane.
Disadvantages:
relatively low efficiency (due to dielectric and conductor losses)
low power
spurious feed radiation (surface waves, strips, etc)
narrow frequency bandwidth (at most a couple of percent)
relatively high level of cross polarization radiation
MSA are applicable in the GHz range (f > 0.5 GHz) for lower
frequencies their dimensions are too large.

The dimensions of the patch are usually in the range from o/3 to o/2. The
dielectric constant of the substrate r is usually in the range from 2.2 to 12.
The most common designs use relatively thick substrates with lower r
because they provide better efficiency and larger bandwidth. On the other
hand, this implies larger dimensions of the antennas. The choice of the
substrate is very much limited by the microwave circuit coupled to the
antenna, which has to be built on the same board. The microwave circuit
together with the antenna is usually manufactured by photo-etching
technology.

Types of microstrip radiators

(c) Microstrip traveling wave antennas

(a) Single radiating patches

Comb MTWA

Meander Line Type MTWA

Rectangular Loop Type MTWA

(b) Single slot radiator

Franklin Type MTWA

The open end of the long TEM line is terminated in a matched resistive
load

The feeding microstrip line is beneath (etched on the other side of the
substrate) see dash-line
3

(d) Microstrip antenna arrays

3. Feeding Methods
1) Microstrip feed easy to fabricate, simple to match by controlling the
inset position and relatively simple to model. However, as the substrate
thickness increases, surface waves and spurious feed radiation increase.

2) Coaxial probe feed easy to fabricate, low spurious radiation; difficult

to model accurately; narrow bandwidth of impedance matching.

Some more examples of microstrip and coaxial probe feeds:

3) Aperture coupling (no contact), microstrip feed line and radiating patch
are on both sides of the ground plane, the coupling aperture is in ground
plane low spurious radiation, easy to model; difficult to match,
narrow bandwidth.

STRIP FEEDS

COAX FEEDS

4) Proximity coupling (no contact), microstrip feed line and radiating

patch are on the same side of the ground plane largest bandwidth (up
to 13%), relatively simple to model, has low spurious radiation.

4. Surface Waves
Surface waves can be excited at the dielectric-to-air interface. Surface
waves give rise to end-fire radiation. In addition they can lead to unwanted
coupling between array elements
The phase velocity of the surface waves is strongly dependent on the
dielectric constant r and thickness h of the substrate.
The excitation of surface waves in a dielectric slab backed by a ground
plane has been well studied (Collin, Field Theory of Guided Waves). The
lowest-order TM mode, TM0, has no cut-off frequency. The cut-off
frequency for the higher-order modes (TMn and TEn) are given by
nc
, n = 1, 2,!,
fc( n ) =
(1)
4h r 1
where c is the speed of light. The cut-off frequencies for the TEn modes are
given by the odd n = 1, 3, 5,, and the cut-off frequencies for the TMn
modes are given by even n. For the TE1 mode the calculated values of h / c(1)

are [ c(1) = c / f c(1) , h / c(1) = n /(4 r 1) ]:

a) 0.217 for duroid (r = 2.32)
b) 0.0833 for alumina (r = 10)
Thus, the lowest-order TE1 mode is excited at 41 GHz for 0.16 cm thick
duroid substrate, and at about 39 GHz for 0.0635 cm thick alumina substrate.
The substrate thickness has to be chosen so that the ratio h / 0 is well below
h / c(1) (here 0 is the free-space wavelength at the operating frequency), i.e.,
[3]
c
.
(2)
h<
4 fu r 1
where fu is the highest frequency in the band of operation. Note that h better
be chosen as high as possible, under the constraint of (2), so that maximum
efficiency is achieved. Of course, the chosen height h has to conform to the
commercially available substrates, too. Another practical formula for h is
given in [2]:
0.3c
h
.
(3)
2 fu r
9

The TM0 mode has no cut-off frequency and is always present to some
extent. The surface TM0 wave excitation becomes appreciable when h/ >
0.09 (r 2.3) and when W/ > 0.03 (r 10). Generally, to suppress the TM0
mode, the dielectric constant should be lower and the substrate height should
be smaller. Unfortunately, decreasing r increases the antenna size, while
decreasing h leads to lower antenna efficiency and frequency band.
5. Criteria for Substrate Selection
a) Possibility for surface-wave excitation
b) Effects of dispersion on the dielectric constant and low tangent of the
substrate
c) Magnitude of copper loss and dielectric loss
d) Anisotropy in the substrate
e) Effects of temperature, humidity, and aging
f) Mechanical
requirements:
conformability,
machinability,
solderability, weight, elasticity, etc.
g) Cost
The first 3 factors are of special concern in the millimeter-wave range (f > 30
GHz).

ELECTRICAL PROPERTIES OF COMMONLY USED SUBSTRATE MATERIALS

FOR MICROSTRIP ANTENNAS
Material

Unreinforced PTFE, Cuflon

Reinforced PTFE, RT Duroid 5880
Fused Quartz
96% Alumina
99.5% Alumina
Sapphire
Semi-Insulating GaAs

Dielectric
Constant
2.1
2.20 (1.5%)
3.78
9.40 (5%)
9.80 (5%)
9.4, 1.6
12.9

Loss
Tangent
0.0004
0.0009
0.0001
0.0010
0.0001
0.0001
0.0020

10

NONELECTRICAL PROPERTIES OF COMMONLY USED SUBSTRATE MATERIALS

FOR MICROSTRIP ANTENNAS
Properties
PTFE
Fused
Alumina Sapphire GaAs
Quartz
temperature
range (C)

-55 260

< +1100

< +1600

-24 370

-55 260

thermal
conductivity
(W/cmK)

0.0026

0.017

0.35 to
0.37

0.42

0.46

coefficient of 16.0 to 108.0

thermal
expansion
(ppm/K)

0.55

6.30 to
6.40

6.00

5.70

+350.0 to
temperature
coefficient of +480.0
dielectric
constant
(ppm/K)

+13.0

+136.0

+110 to
+140

Very
poor
Good

Very
poor
Good

Poor

Poor

Good

Good

minimum
thickness
(mil)

Machinability Good
Solderability

Good

Dimensional
Stability

Good
Poor for
unreinforced,
very good for
others

Excellent Good

Good

Estimated
Cost
(US $/cm2)

1.1

36

3000

560

6. Dispersion effects in substrate

The dependence of the dielectric constant r and the loss tangent on
frequency is referred to as frequency dispersion. For frequencies up to 100
GHz (the typical range for microstrip antennas), the dispersion of r is
practically negligible. The losses, however, display noticeable changes. In
general, the loss tangent increases with frequency, while r decreases.
7. Dielectric loss and copper loss
The loss in the feed lines and the patches themselves are usually computed
with formulas, which were first derived for microstrip transmission lines.

a) Dielectric loss (in dB per unit length, length measured in the units used
for 0 )
z ( f ) 1
r
eff
tan

(4)
d = 27.3
( r 1)
0
r ( f )
eff

b) Copper loss (in dB per unit length)

2

32

W
h ,
1.38 Rs
1
for
2

hZ 0
h
W

32 +

c =

s Z 0 reff ( f ) W 0.667
R
h , for W 1
6.1 105
+

W
h
h

h
+ 1.444

11

(5)

12

In the above equations:

reff ( f ) is the effective dielectric constant (generally, dispersive).

reff

Its quasistatic (low frequency) expression [2] is

1/ 2
+ 1 1
h
r
r
+

+
1
12
,
for W / h > 1

W
2
2
(6)
(0) =
1/ 2
2
W
W
r + 1 + r 1 1 + 12 h
+ 0.04 1 , for
1

2
W
h
h
2

Alternative expression for the quasi-static approximation of reff can

be found in [5].
The quasi-static expressions need a dispersion correction for
frequencies higher than 8 GHz. One possible correction is based on
an empirical formula for the dispersive phase velocity in a microstrip
line [5]. We first compute a normalized frequency (normalized with
respect to the cut-off of the TE1 mode:
4h r 1
f
(7)
f n = (1) =
0
fc
Then, the dispersive phase velocity is calculated as
f n2 reff (0) + r
1
vp =

(8)
f n2 +1
0 reff (0)
Finally,

reff ( f ) = (c / v p ) 2

(9)

For alternative formulas, refer to [5].

Z0 is the characteristic impedance of the microstrip line (generally,

dispersive):

120 reff
W

, for
1
W
W
h

+
1.393
+
0.667ln
+
1.444

Z0 = h
(10)
h

60
W
W
8h

ln + 0.25 ,
for
1
h
h
W
reff

is a constant dependent on the strip thickness t:

h 1.25t 1.25 4 W
W
1
+

ln
1 + 1 +
, for
W

h 2
t
W
=
1 + h 1 1.25t + 1.25 ln 2t , for W 1

W W

h 2
t

W is the effective strip width:

W ' 1.25t
W' 1
4 W
1 + ln

+
, for

h
h 2
t
W' h
=
h W ' 1.25t
W' 1
2h
+
1 + ln t , for h 2
h

Rs is the effective surface resistivity of the conductor:

2
2
Rs = Rs 1+ arctan 1.4 ,

(11)

(12)

(13)

where Rs = f / is the high-frequency surface resistivity of the

conductor. Rs relates to the skin-depth as Rs = ( ) 1 . For a
uniform surface current distribution over a conducting rod of length l
and perimeter of its cross-section P, the resultant resistance is
Rhf = Rs l / P , .
(14)
13

14

Finally, the total loss is the sum of the conduction and dielectric losses:
t = d + c
(15)

SURFACE RESISTIVITY AND SKIN-DEPTH OF COMMONLY USED CONDUCTORS

Rs [Ohm/square x10-7f]

Metal

Silver
Copper
Gold
Aluminum

Ag
Cu
Au
Al

2.5
2.6
3.0
3.3

=6.1x10 S/m
= 5.8x10-7S/m
=4.1x10-7S/m
=3.5 x10-7S/m
-7

Skin-depth at 2 GHz [m]

1.4
1.5
1.7
1.9

15

Some Additional References

[1] D. M. Pozar and D. H. Schaubert, eds., Microstrip Antennas, IEEE Press,

1995 (a collection of significant manuscripts on microstrip antennas).
[2] R.A. Sainati, CAD of Microstrip Antennas for Wireless Applications,
Artech, 1996 (comes with some CAD freeware).
[3] P. Bhartia, K. V. S. Rao, and R. S. Tomar, Millimeter-wave Microstrip
and Printed Circuit Antennas, Artech, 1991.
[4] J.-F. Zrcher and F. E. Gardiol, Broadband Patch Antennas, Artech,
1995.
[5] K.C. Gupta et al., Microstrip Lines and Slotlines, 2nd ed., Artech, 1996.

16

MICROSTRIP ANTENNAS PART II

(Transmission-line model. Design procedure for a rectangular patch. Cavity
model for a rectangular patch.)
1. Transmission line model the rectangular patch
The TL model is the simplest of all, representing the rectangular patch as
a parallel-plate transmission line connecting two radiating slots (apertures),
each of width W and height h:

on Leff , not L. A sketch of the E-field distribution for the first (dominant)
resonant mode, n =1, is shown in the figure below.

Leff > L
x

patch (side view)

G
E001

ground

A. Computing the effective patch length

(
L
= 0.412
h
(

+ 0.3 + 0.264
h
.
W

+ 0.8
reff 0.258
h

For the computation of reff , see Lecture #20.

In the figure above, z is the direction of propagation of the transmission line.

The TL model is not accurate and lacks versatility. However, it gives a
relatively good physical insight into the nature of the patch antenna, and the
field distribution for all TM00 n modes.
The slots represent very high-impedance terminations from both sides of
the transmission line (almost an open circuit). Thus, we expect this structure
to have highly resonant characteristics depending crucially on its length along
z, L. The resonant length of the patch, however, is not exactly equal to the
physical length due to the fringing effect. The fringing effect makes the
effective electrical length of the patch longer than its physical length,
Leff > L . Thus, the resonance condition ( n ) Leff = n / 2 , n=1, 2,!, depends
1

reff

(21.1)

Leff = L + 2L

(21.2)

C. The patch width W

B. Resonant frequency of the dominant TM001 mode

Leff =

0
2

v
2 f (001)

c
c
=
f r(001) =
(001)
2 r f
2 Leff r

(21.3)

The resonant frequency of a patch depends strongly on L, therefore, the

exact calculation of Leff is necessary to predict the antenna resonance:
c
(21.4)
f r(001) =
2 reff ( L + 2L )

The field of the TM001 mode does not depend on the x and y coordinates but it
strongly depends on the z coordinate, along which a standing wave is formed.
The figure below shows the field distribution along z when the patch is in
resonance.

Ex  cos
z
Leff

2
c
2
=
(21.5)
2 f r 0 0 r + 1 2 f r r + 1
Expression (21.5) makes the width W equal to half-wavelength resonant
mode along z. It leads to good radiation efficiencies and acceptable
dimensions. Thus, the patch can be viewed as a continuous planar source
consisting of infinite number of infinitesimally thin half-wavelength dipoles.
W=

D. Equivalent circuit of the patch

The dominant TM001 mode has a uniform field distribution along the yaxis at the slots formed at the front and end edges of the patch. The
equivalent conductance and susceptance can be obtained from the theory of
uniform apertures (see chapter Aperture Antennas in any antenna
textbook).
2
W
1 2 h
h
1
1
G=
< ,
(21.6)
, for
0 10
1200 24 0

2 h
W
h
1
1 0.636ln
B=
< .
(21.7)
, for

1200
10

0
0

The limitation (h / 0 ) < 0.1 is necessary since a uniform field distribution

along the x-axis is also assumed. The patch has two radiating slots (see figure
below).

x
y
z = Leff / 2

z = Leff

1
t#
o
sl
G
E (001)

2
t#
o
sl W

The equivalent circuit of a slot is constructed as a parallel R-C circuit, using

the values computed by (21.6) and (21.7):

constant (see Lecture 20). For each slot, G represents the radiation loss
and B = C represents the capacitance associated with the fringe effect.
E. Resonant input resistance

G =1/ R represents the radiation losses, while B = jC is the equivalent

susceptance, which represents the capacitive nature of the slot.
More accurate values for the conductance G can be obtained through the
cavity model:
I
G=
,
(21.8)
120 2
where

When the patch is resonant, the susceptances of both slots cancel out at
the feed point regardless of the position of the feed along the patch. Thus, the
input admittance is always purely real. This real value, however, strongly
depends on the feed position along z. This is easily shown through the Smith
chart for the admittance transformation through a transmission line.
At the feed point, the impedance of each slot is transformed by the
respective transmission line representing a portion of the patch:
Leff g / 2
G

k0W

sin 2 cos
sin 3 d = 2 + cos X + X S ( X ) + sin X , (21.9)
I =
i
X
cos

L1

L2

Yin

Y1 Y2

sin y
dy .
y
0
The equivalent circuit representing the whole patch in the TM001 mode
includes the two radiating slots as parallel R-C circuits and the patch
connecting them as a transmission line whose characteristics are computed in
the same way as those of a microstrip transmission line.
and X = k0W, k0 = 0 0 . Si denotes the sine integral, Si ( x) =

Z c , g = 0 0 reff

Leff g / 2

Yin = Y1 + Y2
(21.10)
The admittance transformation is given by
YL + jY0 tan( g L)
Yin =Yc
=YL L= , Yc = Z c1
(21.11)
g
Yc + jYL tan( g L)
if the line is loss-free. Below, the Smith charts illustrate the slot-impedance
transformations and their addition, which produces a real normalized
admittance, in three cases: (1) the patch is fed at one edge ( L1 = 0 , L2 = L ),
(2) the patch is fed at the center ( L1 = L2 = L / 2 ), and (3) the patch is fed at a
distance (feed inset) z0 = 0.165 .

Here, Z c is the characteristic impedance of the line, and g is its phase

constant. When losses are not neglected, we must include also the attenuation
5

Y2

slot #2
Y2

slot #1 Y
1

Y1
feed-point at the middle of cavity

Y2

Yin =Y1+Y2 = 215 = 30

L1 = L2 = L / 2 0.23

L 0.46

slot #2 Y
2

slot #1 Y1

feed-point at the edge of cavity

Yin =Y1 +Y2 = 20.09 = 0.18

L 0.46

line to slot #2

L2 = L 0.46

The edge feed and the inset feed are illustrated below.

L1 = z0 0.165
Y1

feed-point at inset z 0

L2 = L z0 (0.46 0.165) = 0.315

Y2

line to slot #2

Yin =Y1+Y2 = 20.5 =1

L 0.46

slot #2

z0

Yin

slot #2 Y
2

slot #1 Y1

line to slot #1

slot #1

Yin

The two slots are separated by an electrical distance of 180D . However,

because of the fringe effect the physical length L of slightly less than /2.
The reduction of the length is not much. Typically, it is 0.48 L 0.49.
Ideally, the resonant input impedance of the patch for the dominant TM001
mode is entirely resistive and equal to half the transformed resistance of each
slot:
1
1
Zin =
=
= Rin .
(21.12)
Yin 2G1

In reality, there is some mutual influence between the two slots, described by
a mutual conductance and it should be included for more accurate
calculations:
1
,
(21.13)
Rin =
2(G1 G12 )
where the + sign relates to the odd modes, while the sign relates to the
even modes. Normally, G12  G1 .
For most patch antennas fed at the edge, Rin is greater than the
characteristic impedance Zc of the microstrip feed line (typically Zc = 50 to
75 ). That is why, the inset-feed technique is widely used to achieve
impedance match.
10

The figure below illustrates the normalized input impedance of a 1-D (along
the y axis) loss-free open-ended transmission-line, whose behavior is very
close to that of the dominant mode of the patch.

2. Designing a rectangular patch using the transmission line model

Input data: r of substrate, h, fr

1) Calculate W using (21.5).
2) Calculate reff using (21.5) and equation (6) from Lecture 20.
3) Calculate the extension L due to fringing effect using (21.1).
4) Calculate actual (physical) length of the patch using

1
L = 0 2L or L =
2L .
(21.16)
2
2 f r reff 0 0

5) Calculate radiating slot admittance using (21.6) and (21.7).

6) Calculate resonant input resistance at patch edge using (21.12) or
(21.13) with G1 = G from (21.6).
7) If Rin calculated in step 6 is too large, calculate the inset distance z0
using (21.14) or (21.15).

3. Cavity model for the rectangular patch

Using modal expansion, the input resistance for the inset-feed at z = z0 is

given approximately by
2 G12 + B12
1
B
2
sin 2 z0 1 sin
Rin =
z0 . (21.14)
cos z0 +
2
2 ( G1 G12 )
Yc
L
L Yc
L
Here, G1 and B1 are calculated using (21.6) and (21.7). For most feeding
microstrips, G1 / Yc  1 and B1 / Yc  1. Then,
1


(21.15)
Rin =
cos 2 y0 = Rin( z = 0 ) cos 2 y0 .
L
2 ( G1 G12 )

L
Notice that the inset feeding technique for impedance match of the microstrip
antennas is essentially identical to the off-center or asymmetrical feeding
techniques for dipoles. In both cases, a position is sought along a resonant
structure, where the current magnitude has the desired value.

The TL model is very limited in its description of the real processes

x
taking place when a patch is excited. It takes into account only the TM 00
n
modes where the energy propagates only in the longitudinal z direction. The
field distribution along the x and y axes is assumed uniform. It is true that the
x
is prevalent but the performance of the patch is very much
dominant TM 001
affected by higher-order modes, too.
The cavity model is a more general model of the patch which imposes
open-end conditions at the side edges of the patch. It represents the patch as a
dielectric-loaded cavity with:
- electrical walls (above and below), and
- magnetic walls (around the perimeter of the patch.
The magnetic wall isG a wall at which
G
n H = 0 (the H field is purely normal)
G
G
n E = 0 (the E field is purely tangential)
It is analogous to the open end termination in the theory of transmission lines.
If we treat the microstrip antenna only as a cavity, we could not represent
radiation because an ideal loss-free cavity does not radiate and its input

11

12

Fig 14.14, pp 735, C. Balanis

impedance is purely reactive. To account for the radiation, a loss mechanism

has to be introduced. This is done by introducing an effective loss tangent,
eff.
The thickness of the substrate is very small. The waves generated and
propagating beneath the patch undergo considerable reflection at the edges of
the patch. Only a very small fraction of them is being radiated. GThus, the
antenna is quite inefficient. The cavity model assumes that the E field is
purely tangential to the slots formed between the ground plane and the patch
edges (magnetic walls). Moreover, it considers only TM x modes, i.e., modes
with no Hx component. These assumptions are, basically, very much true.

1 2 X 1 2Y 1 2 Z
+
+
= k 2
X x 2 Y y 2 Z z 2
2X
2Y
2Z
2
2
+
=
0,
+
=
0,
+ k z2 Z = 0
k
X
k
Y
x
y
2
2
2
x
y
z
The characteristic equation is
k x2 + k y2 + k z2 = k 2 .
The solutions of (21.22) are harmonic functions:
X ( x)= Anc cos(k xn x)+ Ans sin(k xn x)

(21.21)
(21.22)

(21.23)

Y ( y )= Bnc cos(k yn y )+ Bns sin(k yn y )

(21.24)

Z ( z )= Cnc cos(k zn z )+Cns sin(k zn z )

n

When the functions in (21.24) are substituted in (21.20), they give the
general solution of (21.18). The particular solution of (21.18) depends on the
boundary conditions.
In our case, there are electric walls at x = 0 and x = h . There, the
. Having in
tangential E-field components must vanish, i.e., E y = Ez = 0
x = 0, h

The TMx modes are fully described by a single scalar function Ax the xcomponent of the magnetic vector potential:
G
A = Ax x .
(21.17)
In a homogeneous source-free medium, Ax satisfies the wave equation:
2 Ax + k 2 Ax = 0 .
(21.18)
For regular shapes (like the rectangular cavity), it is advantageous to use the
separation of variables:
2 Ax 2 Ax 2 Ax
+
+
+ k 2 Ax = 0
(21.19)
2
2
2
x
y
z
Ax = X ( x)Y ( y ) Z ( z )
(21.20)
X
Y
Z
+ XZ 2 + XY 2 = k 2 XYZ
2
x
x
x
2

YZ

13

mind that

1 2 Ax
1 2 Ax
1 2 Az
2
Ex =
2 + k Ax , E y =

, Ez =

,(21.25)
j x
j xy
j xz

we set Ax at the top and bottom walls as

Ax
= 0.
(21.26)
x x =0,h
At all side walls, we set a vanishing normal derivative for Ax :
Ax
z

= 0,
z = 0, L

Ax
y

= 0.

(21.27)

y = 0,W

This ensures vanishing H x and H y at z = 0 and z = L , as well as vanishing

H x and H z at y = 0 and y = W (magnetic walls), as follows from the
relation between the H-field and Ax :
14

1 Ax
1 Ax
.
(21.28)

, Hx =
z
y
It is obvious now that the solution must appear in terms of the functions
H x = 0, H y =

X ( x) = Anc cos(k xn x), k xn = n

n

Y ( y ) = Bnc cos(k yn y ), k yn = n
n

Z ( z ) = Cnc cos(k zn z ), k zn = n

(21.29)

L
The spectrum of the eigenmodes in the cavity is discrete. The frequencies of
those modes (the resonant frequencies) can be calculated from (21.23) as
2
2
2
m n p
( mnp ) 2
,
(21.30)

+
+
= r
h W L
n

f r( mnp ) =

1
2

m n p

+
+
.
h W L

(21.31)

The mode with the lowest resonant frequency is the dominant mode. Since
x
usually L > W, the lowest-frequency mode is the TM 001
mode, for which
1

c
f r(001) =
=
.
(21.32)
L
2
2L r
x
The dominant TM 001
mode is exactly the mode considered by the
transmission-line model (see previous sections). The field distribution of
some low-order modes is given in the following figure:

Fig. 14.13, pp. 741, C.B.

The general solution for the Ax( mnp ) (see (21.20) and (21.24)) is

Ax( mnp ) = Axc cos m x Bxc cos n y C xc cos p z ,
h
W
L

or



Ax( mnp ) = Amnp cos m x cos n y cos p z .
h
L
W
The respective solution for field vectors of the (m,n,p) mode is
k 2 k x2
Ex = j
Amnp cos(k x x) cos( k y y ) cos(k z z ) ,

kx k y

Ey = j
Ez = j

kx kz

Hy =
Hz =

kz

ky

(21.33)

(21.34)

(21.35)

Amnp sin(k x x) sin(k y y ) cos(k z z ) ,

(21.36)

Amnp sin(k x x) cos( k y y ) sin(k z z ) ,

(21.37)

Hx = 0,
Amnp cos(k x x) cos(k y y ) sin(k z z ) ,

Amnp cos(k x x) sin(k y y ) cos(k z z ) .

(21.38)
(21.39)
(21.40)

x
For the dominant TM 001
mode,

Ex = j (k 2 2 / h 2 ) /( ) A001 cos( z / L), E y = Ez = 0, (21.41)

H y = ( / L) A001 sin( z / L), H x = H z = 0.
(21.42)
15

16

4. Cavity model for the radiated field of a rectangular patch

The microstrip patch is represented by the cavity model reasonably well

assuming that the material of the substrate is truncated and does not extend
beyond the edges of the patch. The four side walls (the magnetic walls)
represent four narrow apertures (slots) through which radiation takes place.
To calculate the radiation fields, the equivalence principle is used. The
field inside the cavity is assumed equal to zero, and its influence on the field
in the infinite region outside is represented by the equivalent surface currents
on the surface of the cavity.

G G
Js , M s

G
Ms

G
Js = 0
G
G
M s = 2n E

G
Ms
The influence of the infinite ground plane
G is accounted for by the image
theory, according to which the currents M s in the presence of the infinite
plane radiate as if magnetic currents
G of double
G strength radiate in free space:
M s = 2n E .
(21.45)
G
Note that an Ex field at the slots corresponds to M s density vector, which is
tangential to the ground plane. Thus, its image is of the same direction. The
x
equivalent magnetic current densities for the dominant TM 001
mode are
sketched below.

G
G
J s = n H
G
G
M s = n E

G G
Js , M s
Because of the very small height (h) of the substrate, the field is concentrated
beneath the patch. There is some actual electrical current at the top metallic
plate, however, its contribution to radiation is negligible. That is because (1)
it is backed by a conductor, and (2) it is very weak compared to the
equivalent currents at the slots. The actual electrical current density of the top
patch is maximum at the edges of the patch, but still its values are negligible
in comparison with the radiation effect from the slots.
In the cavity model, the side walls
employ magnetic-wall boundary
G
condition, which sets the tangential H components at the slots equal to zero.
Therefore,
G
G
J s = n H = 0 .
(21.43)
Only the equivalent magnetic current
G densityG
M s = n E
(21.44)
has substantial contribution to the radiated field.

17

L
x

#1
t
G
lo
s
M
s
z

G
E (001)

#2
t
slo G
Ms

G
At slots #1 and #2, the equivalent M s currents are co-directed and with
equal amplitudes. They are constant along x and y.
18

Radiation from a slot with constant current density

P ( x, y , z )

G
rPQ

G
r

Q ( x, z )
G

G r
Ms

G
The radiation from an (x-y) slot of constant M s currents is found using
G
G
the electric vector potential F . Since M s has only a y component, so does
G G
F : F = Fy y .

Fy (r , , ) =

h/2 W /2

My

r
h / 2 W / 2 PQ

jkrPQ

dxdz .

(21.46)

Here, M y = 2 E0 , E0 being the phasor of the E-field at the radiating slot,

G
and rPQ = r r r = r x sin cos y sin sin .
Fy =2 E0

e jkr
4 r

h/2

W /2

exp( jkx sin cos )dx

h / 2

exp( jky sin sin )dy (21.47)

W / 2

E Wh jkr sin X sin Y

Fy = 0
e

2 r
X
Y

(21.48)

where
kh
sin cos ,
2
kW
Y=
sin sin .
2
X=

19

According to the relation between the far-zone E-field and the vector
potential,
Er  0, E = j F , E = j F ,
(21.49)
where = / , F = Fy cos sin , and F = Fy cos .
WhE0 jkr
sin X sin Y
E = j
e
cos sin
(21.50)
2 r
X
Y
WhE0 jkr
sin X sin Y
E = j
e
cos
(21.51)
2 r
X
Y
Since = k ,
V
sin X sin Y

E = jkW 0 e jkr cos sin

(21.52)
,
X
Y
2 r

V
sin X sin Y

E = jkW 0 e jkr cos

(21.53)
.
2 r
X
Y

Here V0 = hE0 is the voltage between the patch edge and the ground plane.
Slots #1 and #2 form an array of two elements with excitation of equal
phase and magnitude, separated by a distance L. Their AF is
kLeff

AF12 = 2cos
cos .
(21.54)
2

Here Leff = L + 2L is the effective patch length. Thus, the total radiation
field is

kWV0 jkr
sin X sin Y kLeff
cos , (21.55)
Et = j
e
cos sin
cos
X
Y 2
r

kL

eff

kWV0 jkr
sin X sin Y
cos . (21.56)
Et = j
e
cos
cos
X
Y 2
r

kLeff
cos , the pattern of the patch is obtained as
Introducing Z =
2
sin X sin Y
f ( , ) = E2 + E2 = 1 sin 2 sin 2
cos Z . (21.57)
X
Y

20

E-plane pattern (x-z plane, = 0, 0

180)

kh

sin sin
2
cos kLeff cos .
f E ( ) =

kh
2

sin
2

(21.58)

H-plane pattern (x-y plane, = 90, 0 90 and 270

kh

kW

sin cos sin

sin
2
2

f H ( ) = cos
kh
kW
cos
sin
2
2

360)

(21.59)

90

30

60

330

1
120

0.8

60

30

300

0.6
150
0.4

r = cos(t) (sin(0.01 cos(t))/(0.01 cos(t))) (sin(0.6 sin(t))/(0.6 sin(t)))

0

90

270

0.2

r = (cos(.674 cos(t)/2)) (sin(0.05 sin(t))/(0.05 sin(t)))

0.4

0.6

r = 2.2

180

90

0.8

0.2

30

60

330

1
120
0.8

60

30

300

0.6
150
0.4

21

0.4

0.2

r = cos(t) (sin(0.01 cos(t))/(0.01 cos(t))) (sin(1.1 sin(t))/(1.1 sin(t)))

270

r = (cos(.5 cos(t)/2)) (sin(0.05 sin(t))/(0.05 sin(t)))

0.6

r = 4

90

180

0.8

0.2

22

Fig.14.16, p. 744, Balanis

Fig. 14.18, p. 747, Balanis

Fig. 14.17, p. 746, Balanis

Non-radiating slots: It can be shown that the slots at y = W / 2 and

y = W / 2 do no radiate in the principle E- and H-planes. In general, these
two slots do radiate away from the principle planes, but their field intensity is
everywhere small compared to that radiated by slots #1 and #2.

23

24

MICROSTRIP ANTENNAS PART III

(Circular patch antennas: the cavity model. Radiation field of the circular
patch. Circularly polarized radiation from patches. Arrays and feed
networks.)
1. The circular patch: the cavity model

The circular patch cannot be analyzed using the TL method, but can be
accurately described by the cavity method. It is again assumed that only TMz
modes
are supported in the cavity. They are fully described by the VP
G

A = Az z . The Az VP function satisfies the Helmholtz equation:

2 Az + k 2 Az = 0
which now is solved in cylindrical coordinates:
1 Az 1 2 Az 2 Az
+ 2
+
+ k 2 Az = 0

2
2

z
2
2
Az 1 Az 2 Az
1 Az
+
+
+ k 2 Az = 0

+

2 2 2
z 2
1 Az 2 Az
1 2 Az 2 Az
+
+
+ 2 + k 2 Az = 0
2 2 2
z
Az = R ( ) F ( ) Z ( z )

R
R RZ F
Z
+ FZ 2 + 2
+ RF 2 + k 2 RFZ = 0
2

z

2

FZ

(22.1)
(22.2)
(22.3)

(22.4)
(22.5)

(22.6)

1 R 1 2 R
1 2 F 1 2Z
+
+
+
= k 2
(22.7)
R R 2 2 F 2 Z z 2
The 4th term is independent on and , and is being separated:
1 2Z
= k z 2 .
(22.8)
Z z 2
Then,
1 R 1 2 R
1 2F
+
+
= k 2 k z 2 = const
(22.9)
R R 2 2 F 2
R 2 2 R 1 2 F

+
+
+ k 2 kz2 2 = 0 .
(22.10)
R R 2 F 2
Now, the 3rd term is independent of , and the other terms are independent
of . Thus, (22.10) is separated into two equations:
1 2F
= k2
(22.11)
F 2
and
R 2 2 R
+
+ k 2 k z 2 2 k 2 = 0 .
(22.12)
R R 2
We define
k 2 = k 2 kz 2 .
(22.13)

Then (22.12) can be written as (note that (22.12) depends only on ):

2
R
+ ( k ) k 2 R = 0 .
(22.14)

Thus, equation (22.1) has been separated into three ordinary differential
equations - (22.8), (22.11) and (22.14).
(a) Equation (22.8) is complemented by Neumann BC at the top patch and
the grounded plane (electric walls):
Az
Z
=0
= 0.
(22.15)
z
z
Its solution, therefore, is in the form


Z ( z ) = cn cos n z
(22.16)
h
n
with the eigenvalues k z = n / h .
(b) The solution of (22.11) is also a harmonic function. Since we are
interested in real-valued harmonic functions,
F ( ) = bmc cos km + bms sin km .
(22.17)
m

Since there are no specific BCs to be imposed at certain angular positions,

the only requirement for the eigenvalues km comes from the condition that
the F ( ) must be periodic in :
F ( ) = F ( + 2 ) .

(22.18)

Equation (22.18) is true only if k are integers. That is why the usual
construction of a general solution for F ( ) for a complete cylindrical region
( = 0 to 2 ) is in the form
F ( ) = bmc cos ( m ) + bms sin ( m ) .

Fig. D-1, Harrington, p. 461

(22.19)

This is the well-known Fourier-series expansion.

(c) Equation (22.14) is a Bessel equation in which k is an integer (let k
= n). Solutions are of the form of the following special functions:
J n ( k ) Bessel function of the first kind,
N n ( k ) Bessel function of the second kind (Neumann function),

H n(1) ( k ) Hankel function of the first kind,

H n(2) ( k ) Hankel function of the second kind.

Note: H n(1) = J n + jN n ; H n(2) = J n jN n . The eigenvalues are determined

according to the boundary conditions.

Fig. D-2, Harrington, p. 462

In the cavity model, it is required that (magnetic wall):

Az
R
=0
,
=0

=a

(22.20)

and that the field is finite for = a . The Bessel functions of the first kind
J n (k ) are the suitable choice. The eigenvalues k are determined from
(22.20):

J n (k )

nm

, k

=0
=a

nm
a

(22.21)

is the m null of the derivative of the Bessel function of the nth

where nm
order J n . Thus, the solution of the Helmholtz equation for Az can be given in
a modal form as, see (22.5),
c

Az( mnp ) = M mnp J m nm

bn cos ( n ) + bn sin ( n ) cos p z . (22.22)
a

h
The characteristic equation (22.13) is finally obtained as
k 2 = 2 = k 2 + k z 2 .
(22.23)
From (22.23), the resonant frequencies of the patch can be obtained:
2
2

2
(22.24)
= nm + p ,
mnp
a h
th

nm
(22.25)
a + p h .
2

Equation (22.25) does not take into account the fringing effect of the circular
patch. The actual radius a should be replaced by
f r ( mnp ) =

1/ 2

2h a
(22.26)
ae = a 1 +
ln
+ 1.7726 .

a r 2h
The first four modes in ascending order are TMz110, TMz210, TMz010,
are
TMz310 where the respective nulls nm
= 1.8412
11
= 3.0542
21

The resonant frequency of the dominant TMz110 mode can be determined

from (22.26) as
1.8412c
f r (110) =
,
(22.27)
2 ae r
where c is the speed of light in vacuum.
The VP of the dominant TMz110 mode is
c
s
Az(110) = M110 J1 11
(22.28)
bn cos + bn sin .
a

Assuming excitation at = 0 ( Az has vanishing angular first derivative), we

set bns = 0 . The field components are computed from Az according to the
field-potential relations
j 2 Az
z
j 1 2 Az
E =
z

j 2 Az
2
Ez =
2 + k Az
z

E =

1 1 Az

1 Az
H =

Hz = 0
H =

(22.29)

For the dominant TMz110 mode,

E = E = 0
/ a ) cos
Ez = E0 J1 ( 11

H = j
H = j

E0 1

0
E0

/ a )sin
J1 ( 11

(22.30)

/ a )cos
J1( 11

= 3.8318
01
= 4.2012
31
5

E-plane amplitude pattern:

, E = 0
E (0D 90D , = 0D ,180D ) = J 02

2. Radiated fields and equivalent surface currents of the circular patch

H-plane amplitude pattern:

E (0D 90D , = 90D , 270D ) = cos J 02 , E = 0

As with the rectangular patch, the field radiated by the circular slot is
determined using the equivalence principle. The circumferential wall of the
cavity is replaced by an equivalent circular sheet of magnetic current density
M = 2 Ez
(22.31)
=a

radiating in free space. The factor of 2 accounts for the ground plane. Since
the height of the slot h is very small and the slot field is independent of z, we
can substitute the surface magnetic current density over the slot with a
filamentary magnetic current I m = M h :
)cos .
I m = 2hE0 J1 ( 11
(22.32)
 

2V0

) is the voltage between ground and the top plate of the

Here, V0 = hE0 J1 ( 11
patch at the feed ( = 0 ).
Using the theory for the radiation field of a circular slot, the following
expressions are obtained for the far field of the circular patch:
, E = C (r ) cos sin J 02 ,
Er = 0, E = C (r ) cos J 02
(22.33)
where
k a V e jk0 r
C (r ) = j 0 e 0
,
2r
J 02 = J 0 (k0 ae sin ) + J 2 (k0 ae sin ) ,
= J 0 (k0 ae sin ) J 2 (k0 ae sin ) .
J 02

Fig. 14.23, p. 758, Balanis

W =L

3. Circular polarization with patch antennas

Circular polarization can be obtained if two orthogonal modes are excited

with a 90 time-phase difference between them. This can be accomplished by
adjusting the physical dimensions of the patch and using either single or two
feed points.

W =L

d
L

A. Square patch with circularly polarized field

(e) CP for square patches with thin slots: c = L / 2.72 = W / 2.72 , d = c /10
B. Circular patch with circularly polarized field

1
L = W 1 +
Qt
(c) Nearly square patch with microstrip-line feed for CP accounting for losses

(d) Coax-feeds for CP

9

10

FEED-PROBE ANGULAR SPACING OF DIFFERENT MODES FOR CIRCULAR

POLARIZATION

TM110

90

TM210
45
or
135

TM310
30
or
90

TM410
22.5
or
67.5

TM510
TM610
18, 54 15, 45
or
or
90
75

4. Array and feed networks

11

12