Writen by ATI Instructor Robert Nelson for Via Satellite,

September, 1999
Antennas
The Interface with Space
by Robert A. Nelson
The antenna is the most visible part of the satellite communication
system. The antenna transmits and receives the modulated carrier signal
at the radio frequency (RF) portion of the electromagnetic spectrum. For
satellite communication, the frequencies range from about 0.3 GHz (VHF)
to 30 GHz (Ka-band) and beyond. These frequencies represent microwaves,
with wavelengths on the order of one meter down to below one centimeter.
High frequencies, and the corresponding small wavelengths, permit the
use of antennas having practical dimensions for commercial use. This
article summarizes the basic properties of antennas used in satellite
communication and derives several fundamental relations used in antenna
design and RF link analysis.
HISTORY OF ELECTROMAGNETIC WAVES
The quantitative study of electricity and magnetism began with the scientifc
research of the rench physicist !harles Augustin !oulomb. In "#$# !oulomb
proposed a law of force for charges that% li&e Sir Isaac Newton's law of gravitation%
varied inversely as the square of the distance. (sing a sensitive torsion balance% he
demonstrated its validity e)perimentally for forces of both repulsion and attraction.
*i&e the law of gravitation% !oulomb's law was based on the notion of +action at a
distance%+ wherein bodies can interact instantaneously and directly with one another
without the intervention of any intermediary.
At the beginning of the nineteenth century% the electrochemical cell was invented by
Alessandro ,olta% professor of natural philosophy at the (niversity of -avia in Italy.
The cell created an electromotive force% which made the production of continuous
currents possible. Then in "$./ at the (niversity of !openhagen% 0ans !hristian
1ersted made the momentous discovery that an electric current in a wire could
de2ect a magnetic needle. News of this discovery was communicated to the rench
Academy of Sciences two months later. The laws of force between current bearing
wires were at once investigated by Andre34arie Ampere and by 5ean36aptiste 6iot
and eli) Savart. 7ithin si) years the theory of steady currents was complete. These
laws were also +action at a distance+ laws% that is% e)pressed directly in terms of the
distances between the current elements.
Subsequently% in "$8"% the 6ritish scientist 4ichael araday demonstrated the
reciprocal e9ect% in which a moving magnet in the vicinity of a coil of wire produced
an electric current. This phenomenon% together with 1ersted's e)periment with the
magnetic needle% led araday to conceive the notion of a magnetic feld. A feld
produced by a current in a wire interacted with a magnet. Also% according to his law
of induction% a time varying magnetic feld incident on a wire would induce a voltage%
thereby creating a current. :lectric forces could similarly be e)pressed in terms of an
electric feld created by the presence of a charge.
araday's feld concept implied that charges and currents interacted directly and
locally with the electromagnetic feld% which although produced by charges and
currents% had an identity of its own. This view was in contrast to the concept of
+action at a distance%+ which assumed bodies interacted directly with one another.
araday% however% was a self3taught e)perimentalist and did not formulate his laws
mathematically.
It was left to the Scottish physicist 5ames !ler& 4a)well to establish the mathematical
theory of electromagnetism based on the physical concepts of araday. In a series of
papers published between "$;< and "$<;% 4a)well restated the laws of !oulomb%
Ampere% and araday in terms of araday's electric and magnetic felds. 4a)well thus
unifed the theories of electricity and magnetism% in the same sense that two
hundred years earlier Newton had unifed terrestrial and celestial mechanics through
his theory of universal gravitation.
As is typical of abstract mathematical reasoning% 4a)well saw in his equations a
certain symmetry that suggested the need for an additional term% involving the time
rate of change of the electric feld. 7ith this generali=ation% 4a)well's equations also
became consistent with the principle of conservation of charge.
urthermore% 4a)well made the profound observation that his set of equations% thus
modifed% predicted the e)istence of electromagnetic waves. Therefore% disturbances
in the electromagnetic feld could propagate through space. (sing the values of
&nown e)perimental constants obtained solely from measurements of charges and
currents% 4a)well deduced that the speed of propagation was equal to speed of light.
This quantity had been measured astronomically by 1laf Romer in "<#< from the
eclipses of 5upiter's satellites and determined e)perimentally from terrestrial
measurements by 0.*. i=eau in "$>?. 0e then asserted that light itself was an
electromagnetic wave% thereby unifying optics with electromagnetism as well.
4a)well was aided by his superior &nowledge of dimensional analysis and units of
measure. 0e was a member of the 6ritish Association committee formed in "$<" that
eventually established the centimeter3gram3second @!ASB system of absolute
electrical units.
4a)well's theory was not accepted by scientists immediately% in part because it had
been derived from a bewildering collection of mechanical analogies and diCcult
mathematical concepts. The form of 4a)well's equations as they are &nown today is
due to the Aerman physicist 0einrich 0ert=. 0ert= simplifed them and eliminated
unnecessary assumptions.
0ert='s interest in 4a)well's theory was occasioned by a pri=e o9ered by the 6erlin
Academy of Sciences in "$#? for research on the relation between polari=ation in
insulators and electromagnetic induction. 6y means of his e)periments% 0ert=
discovered how to generate high frequency electrical oscillations. 0e was surprised to
fnd that these oscillations could be detected at large distances from the apparatus.
(p to that time% it had been generally assumed that electrical forces decreased
rapidly with distance according to the Newtonian law. 0e therefore sought to test
4a)well's prediction of the e)istence of electromagnetic waves.
In "$$$% 0ert= set up standing electromagnetic waves using an oscillator and spar&
detector of his own design and made independent measurements of their wavelength
and frequency. 0e found that their product was indeed the speed of light. 0e also
verifed that these waves behaved according to all the laws of re2ection% refraction%
and polari=ation that applied to visible light% thus demonstrating that they di9ered
from light only in wavelength and frequency. +!ertainly it is a fascinating idea%+ 0ert=
wrote% +that the processes in air that we have been investigating represent to us on a
million3fold larger scale the same processes which go on in the neighborhood of a
resnel mirror or between the glass plates used in e)hibiting Newton's rings.+
It was not long until the discovery of electromagnetic waves was transformed from
pure physics to engineering. After learning of 0ert='s e)periments through a
maga=ine article% the young Italian engineer Auglielmo 4arconi constructed the frst
transmitter for wireless telegraphy in "$?;. 7ithin two years he used this new
invention to communicate with ships at sea. 4arconi's transmission system was
improved by Darl . 6raun% who increased the power% and hence the range% by
coupling the transmitter to the antenna through a transformer instead of having the
antenna in the power circuit directly. Transmission over long distances was made
possible by the re2ection of radio waves by the ionosphere. or their contributions to
wireless telegraphy% 4arconi and 6raun were awarded the Nobel -ri=e in physics in
"?/?.
4arconi created the American 4arconi 7ireless Telegraphy !ompany in "$??% which
competed directly with the transatlantic undersea cable operators. 1n the early
morning of April ";% "?".% a ."3year old 4arconi telegrapher in New Eor& !ity by the
name of Favid Sarno9 received a wireless message from the 4arconi station in
Newfoundland% which had pic&ed up faint S1S distress signals from the steamship
Titanic. Sarno9 relayed the report of the ship's sin&ing to the world. This singular
event dramati=ed the importance of the new means of communication.
Initially% wireless communication was synonymous with telegraphy. or
communication over long distances the wavelengths were greater than .// meters.
The antennas were typically dipoles formed by long wires cut to a submultiple of the
wavelength.
!ommercial radio emerged during the "?./s and "?8/s. The American 4arconi
!ompany evolved into the Radio !orporation of America @R!AB with Favid Sarno9 as
its director. Technical developments included the invention of the triode for
amplifcation by *ee de orest and the perfection of A4 and 4 receivers through the
wor& of :dwin 0oward Armstrong and others. In his boo& Empire of the Air: The Men
Who Made Radio% Tom *ewis credits
de orest% Armstrong% and Sarno9 as the three visionary pioneers most responsible
for the birth of the modern communications age.
Stimulated by the invention of radar during 7orld 7ar II% considerable research and
development in radio communication at microwave frequencies and centimeter
wavelengths was conducted in the decade of the "?>/s. The 4IT Radiation
*aboratory was a leading center for research on microwave antenna theory and
design. The basic formulation of the radio transmission formula was developed by
0arald T. riis at the 6ell Telephone *aboratories and published in "?><. This equation
e)pressed the radiation from an antenna in terms of the power 2ow per unit area%
instead of giving the feld strength in volts per meter% and is the foundation of the R
lin& equation used by satellite communication engineers today.
TYPES OF ANTENNAS
A variety of antenna types are used in satellite communications. The most widely
used narrow beam antennas are re2ector antennas. The shape is generally a
paraboloid of revolution. or full earth coverage from a geostationary satellite% a horn
antenna is used. 0orns are also used as feeds for re2ector antennas.
In a direct feed re2ector% such as on a satellite or a small earth terminal% the feed
horn is located at the focus or may be o9set to one side of the focus. *arge earth
station antennas have a subre2ector at the focus. In the !assegrain design% the
subre2ector is conve) with an hyperboloidal surface% while in the Aregorian design it
is concave with an ellipsoidal surface.
The subre2ector permits the antenna optics to be located near the base of the
antenna. This confguration reduces losses because the length of the waveguide
between the transmitter or receiver and the antenna feed is reduced. The system
noise temperature is also reduced because the receiver loo&s at the cold s&y instead
of the warm earth. In addition% the mechanical stability is improved% resulting in
higher pointing accuracy.
-hased array antennas may be used to produce multiple beams or for electronic
steering. -hased arrays are found on many nongeostationary satellites% such as the
Iridium% Alobalstar% and I!1 satellites for mobile telephony.
GAIN AND HALF POWER BEAMWIDTH
The fundamental characteristics of an antenna are its gain and half power
beamwidth. According to the reciprocity theorem% the transmitting and receiving
patterns of an antenna are identical at a given wavelength
The gain is a measure of how much of the input power is concentrated in a particular
direction. It is e)pressed with respect to a hypothetical isotropic antenna% which
radiates equally in all directions. Thus in the direction @ % B% the gain is
G( , ) = (dP/d)/(P
in
/4 )
where Pin is the total input power and dP is the increment of radiated output power in
solid angle d. The gain is ma)imum along the boresight direction.
The input power is Pin G Ea
.
A H Z/ where Ea is the average electric feld over the
area A of the aperture% Z/ is the impedance of free space% and is the net antenna
eCciency. The output power over solid angle dis dP G E
.
r
.
dH Z/% where E is the
electric feld at distance r. 6ut by the raunhofer theory of di9raction% E G Ea A H r
along the boresight direction% where is the wavelength. Thus the boresight gain is
given in terms of the si=e of the antenna by the important relation
G G @> H
.
B A
This equation determines the required antenna area for the specifed gain at a given
wavelength.
The net eCciency is the product of the aperture taper eCciency a % which depends
on the electric feld distribution over the antenna aperture @it is the square of the
average divided by the average of the squareB% and the total radiation eCciency I
G PHPin associated with various losses. These losses include spillover% ohmic heating%
phase nonuniformity% bloc&age% surface roughness% and cross polari=ation. Thus G
a I. or a typical antenna% G /.;;.
or a re2ector antenna% the area is simply the proJected area. Thus for a circular
re2ector of diameter % the area is A G
.
H> and the gain is
G = ( / )
!
which can also be written


G = ( f / c)
!
since c G f% where c is the speed of light @8 "/
$
mHsB% is the wavelength% and f is
the frequency. !onsequently% the gain increases as the wavelength decreases or the
frequency increases.
or e)ample% an antenna with a diameter of . m and an eCciency of /.;; would have
a gain of $<$; at the !3band uplin& frequency of < A0= and wavelength of /./;/ m.
The gain e)pressed in decibels @d6B is
"/ log@$<$;B G 8?.> d6. Thus the power radiated by the antenna is $<$; times more
concentrated along the boresight direction than for an isotropic antenna% which by
defnition has a gain of " @/ d6B. At Du3band% with an uplin& frequency of "> A0= and
wavelength /./." m% the gain is >?%.8< or ><.? d6. Thus at the higher frequency% the
gain is higher for the same si=e antenna.
The boresight gain G can be e)pressed in terms of the antenna beam solid angle A
that contains the total radiated power as
G = " (4 /
A
)
which ta&es into account the antenna losses through the radiation eCciency I. The
antenna beam solid angle is the solid angle through which all the power would be
concentrated if the gain were constant and equal to its ma)imum value. The
directivity does not include radiation losses and is equal to G H I.
The half power beamwidth is the angular separation between the half power points
on the antenna radiation pattern% where the gain is one half the ma)imum value. or
a re2ector antenna it may be e)pressed
0-67 G G # H
where # is a factor that depends on the shape of the re2ector and the method of
illumination. or a typical antenna% # G #/ @"... if is in radiansB. Thus the half
power beamwidth decreases with decreasing wavelength and increasing diameter.
or e)ample% in the case of the . meter antenna% the half power beamwidth at < A0=
is appro)imately ".#; . At "> A0=% the half power beamwidth is appro)imately /.#; .
As an e)treme e)ample% the half power beamwidth of the Feep Space Networ& <>
meter antenna in Aoldstone% !alifornia is only /./> at K3band @$.> A0=B.
The gain may be e)pressed directly in terms of the half power beamwidth by
eliminating the factor H . Thus%
G G @ # H B
.
Inserting the typical values G /.;; and # G #/ % one obtains
G = !$,%%%/ ( )
!
where is e)pressed in degrees. This is a well &nown engineering appro)imation for
the gain @e)pressed as a numericB. It shows directly how the si=e of the beam
automatically determines the gain. Although this relation was derived specifcally for
a re2ector antenna with a circular beam% similar relations can be obtained for other
antenna types and beam shapes. The value of the numerator will be somewhat
di9erent in each case.
or e)ample% for a satellite antenna with a circular spot beam of diameter " % the
gain is .#%/// or >>.8 d6. or a Du3band downlin& at ". A0=% the required antenna
diameter determined from either the gain or the half power beamwidth is ".#; m.
A horn antenna would be used to provide full earth coverage from geostationary
orbit% where the angular diameter of the earth is "#.> . Thus% the required gain is
$?.. or "?.; d6. Assuming an eCciency of /.#/% the horn diameter for a !3band
downlin& frequency of > A0= would be .# cm.
EIRP AND G/T
or the R lin& budget% the two required antenna properties are the equivalent
isotropic radiated power @:IR-B and the +fgure of merit+ GHT. These quantities are the
properties of the transmit antenna and receive antenna that appear in the R lin&
equation and are calculated at the transmit and receive frequencies% respectively.
The equivalent isotropic radiated power @:IR-B is the power radiated equally in all
directions that would produce a power 2u) density equivalent to the power 2u)
density of the actual antenna. The power 2u) density is defned as the radiated
power P per unit area &% or G PH&. 6ut P G I Pin % where Pin is the input power and
I is the radiation eCciency% and
& G d
.
A %where d is the slant range to the center of coverage and A is the solid
angle containing the total power. Thus with some algebraic manipulation%
G I @> H A B@ Pin H > d
.
B G G Pin H > d
.
Since the surface area of a sphere of radius d is > d
.
% the 2u) density in terms of the
:IR- is
G :IR- H > d
.
:quating these two e)pressions% one obtains
:IR- G G Pin
Therefore% the equivalent isotropic radiated power is the product of the antenna gain
of the transmitter and the power applied to the input terminals of the antenna. The
antenna eCciency is absorbed in the defnition of gain.
The +fgure of merit+ is the ratio of the antenna gain of the receiver G and the system
temperature T. The system temperature is a measure of the total noise power and
includes contributions from the antenna and the receiver. 6oth the gain and the
system temperature must be referenced to the same point in the chain of
components in the receiver system. The ratio GHT is important because it is an
invariant that is independent of the reference point where it is calculated% even
though the gain and the system temperature individually are di9erent at di9erent
points.
ANTENNA PATTERN
Since electromagnetic energy propagates in the form of waves% it spreads out
through space due to the phenomenon of di9raction. Individual waves combine both
constructively and destructively to form a di9raction pattern that manifests itself in
the main lobe and side lobes of the antenna.
The antenna pattern is analogous to the +Airy rings+ produced by visible light when
passing through a circular aperture. These di9raction patterns were studied by Sir
Aeorge 6iddell Airy% Astronomer Royal of :ngland during the nineteenth century% to
investigate the resolving power of a telescope. The di9raction pattern consists of a
central bright spot surrounded by concentric bright rings with decreasing intensity.
The central spot is produced by waves that combine constructively and is analogous
to the main lobe of the antenna. The spot is bordered by a dar& ring% where waves
combine destructively% that is analogous to the frst null. The surrounding bright rings
are analogous to the side lobes of the antenna pattern. As noted by 0ert=% the only
di9erence in this behavior is the si=e of the pattern and the di9erence in wavelength.
7ithin the main lobe of an a)isymmetric antenna% the gain G@ B in a direction with
respect to the boresight direction may be appro)imated by the e)pression
G( ) = G '! ( / )
!
where G is the boresight gain. 0ere the gains are e)pressed in d6. Thus at the half
power points to either side of the boresight direction% where G H.% the gain is
reduced by a factor of .% or 8 d6. The details of the antenna% including its shape and
illumination% are contained in the value of the half power beamwidth . This equation
would typically be used to estimate the antenna loss due to a small pointing error.
The gain of the side lobes can be appro)imated by an envelope. or new earth
station antennas with
H L "//% the side lobes must fall within the envelope .? .; log by international
regulation. This envelope is determined by the requirement of minimi=ing
interference between neighboring satellites in the geostationary arc with a nominal
. spacing.
TAPER
The gain pattern of a re2ector antenna depends on how the antenna is illuminated by
the feed. The variation in electric feld across the antenna diameter is called the
antenna taper.
The total antenna solid angle containing all of the radiated power% including side
lobes% is
A G I @> H GB G @"HaB @
.
H AB
where a is the aperture taper eCciency and I is the radiation eCciency associated
with losses. The beam eCciency is defned as
G M H A
where M is the solid angle for the main lobe. The values of a and are calculated
from the electric feld distribution in the aperture plane and the antenna radiation
pattern% respectively.
or a theoretically uniform illumination% the electric feld is constant and the aperture
taper eCciency is ". If the feed is designed to cause the electric feld to decrease
with distance from the center% then the aperture taper eCciency decreases but the
proportion of power in the main lobe increases. In general% ma)imum aperture taper
eCciency occurs for a uniform distribution% but ma)imum beam eCciency occurs for
a highly tapered distribution.
or uniform illumination% the half power beamwidth is ;$.> H and the frst side
lobe is "#.< d6 below the pea& intensity in the boresight direction. In this case% the
main lobe contains about $> percent of the total radiated power and the frst side
lobe contains about # percent.
If the electric feld amplitude has a simple parabolic distribution% falling to =ero at the
re2ector edge% then the aperture taper eCciency becomes /.#; but the fraction of
power in the main lobe increases to ?$ percent. The half power beamwidth is now
#..$ H and the frst side lobe is .>.< d6 below pea& intensity. Thus% although the
aperture taper eCciency is less% more power is contained in the main lobe% as
indicated by the larger half power beamwidth and lower side lobe intensity.
If the electric feld decreases to a fraction ( of its ma)imum value% called the edge
taper% the re2ector will not intercept all the radiation from the feed. There will be
energy spillover with a corresponding eCciency of appro)imately " (
.
. 0owever% as
the spillover eCciency decreases% the aperture taper eCciency increases. The taper
is chosen to ma)imi=e the illumination eCciency% defned as the product of aperture
taper eCciency and spillover eCciency.
The illumination eCciency reaches a ma)imum value for an optimum combination of
taper and spillover. or a typical antenna% the optimum edge taper ( is about /.8"<%
or "/ d6 @./ log (B. 7ith this edge taper and a parabolic illumination% the aperture
taper eCciency is /.?.% the spillover eCciency is /.?/% the half power beamwidth is
<;.8 H% and the frst side lobe is ...8 d6 below pea&. Thus the overall illumination
eCciency is /.$8 instead of /.#;. The beam eCciency is about ?; percent.
COVERAGE AREA
The gain of a satellite antenna is designed to provide a specifed area of coverage on
the earth. The area of coverage within the half power beamwidth is
& = d
!

where d is the slant range to the center of the footprint and is the solid angle of a
cone that intercepts the half power points% which may be e)pressed in terms of the
angular dimensions of the antenna beam. Thus
= K
where and are the principal plane half power beamwidths in radians and K is a factor that depends on
the shape of the coverage area. For a square or rectangular area of coverage, K = 1, while for a circular or
elliptical area of coverage, K = /4.
The boresight gain may be appro)imated in terms of this solid angle by the relation
G = (4 / ) = ( / ))(4',!*+ / )
where and are in degrees and is an eCciency factor that depends on the the
half power beamwidth. Although is conceptually distinct from the net eCciency %
in practice these two eCciencies are roughly equal for a typical antenna taper. In
particular% for a circular beam this equation is equivalent to the earlier e)pression in
terms of if G @ # H >B
.
.
If the area of the footprint & is specifed% then the si=e of a satellite antenna increases
in proportion to the altitude. or e)ample% the altitude of *ow :arth 1rbit is about
"/// &m and the altitude of 4edium :arth 1rbit is about "/%/// &m. Thus to cover
the same area on the earth% the antenna diameter of a 4:1 satellite must be about
"/ times that of a *:1 satellite and the gain must be "// times% or ./ d6% as great.
1n the Iridium satellite there are three main mission *3band phased array antennas.
:ach antenna has "/< elements% distributed into $ rows with element separations of
"".; cm and row separations of ?.> cm over an antenna area of "$$ cm $< cm. The
pattern produced by each antenna is divided into "< cells by a two3dimensional
6utler matri) power divider% resulting in a total of >$ cells over the satellite coverage
area. The ma)imum gain for a cell at the perimeter of the coverage area is .>.8 d6.
rom geostationary orbit the antenna si=e for a small spot beam can be considerable.
or e)ample% the spacecraft for the Asia !ellular Satellite System @A!eSB% being built
by *oc&heed 4artin for mobile telephony in Southeast Asia% has two unfurlable mesh
antenna re2ectors at *3band that are ". meters across and have an o9set feed.
0aving di9erent transmit and receive antennas minimi=es passive intermodulation
@-I4B interference that in the past has been a serious problem for high power *3band
satellites using a single re2ector. The antenna separation attenuates the -I4
products by from ;/ to #/ d6.
SHAPED BEAMS
1ften the area of coverage has an irregular shape% such as one defned by a country
or continent. (ntil recently% the usual practice has been to create the desired
coverage pattern by means of a beam forming networ&. :ach beam has its own feed
and illuminates the full re2ector area. The superposition of all the individual circular
beams produces the specifed shaped beam.
or e)ample% the !3band transmit hemiH=one antenna on the Intelsat < satellite is 8..
meters in diameter. This is the largest diameter solid circular aperture that fts within
an Ariane > launch vehicle fairing envelope. The antenna is illuminated by an array of
">< -otter horns. The beam diameter for each feed is ".< at 8.# A0=. 6y
appropriately e)citing the beam forming networ&% the specifed areas of coverage are
illuminated. or .# d6 spatial isolation between =ones reusing the same spectrum%
the minimum spacing is given by the rule of thumb ".> % so that ... . This
meets the specifcation of G ..; for Intelsat <.
Another e)ample is provided by the 0S38#<
dual3spin stabili=ed Aala)y ; satellite% operated by -anAmSat. The re2ector diameter
is ".$/ m. There are two linear polari=ations% hori=ontal and vertical. In a given
polari=ation% the contiguous (nited States @!1N(SB might be covered by four beams%
each with a half power beamwidth of 8 at the !3band downlin& frequency of > A0=.
rom geostationary orbit% the angular dimensions of !1N(S are appro)imately
< 8 . or this rectangular beam pattern% the ma)imum gain is about 8" d6. At
edge of coverage% the gain is 8 d6 less. 7ith a T7TA ouput power of "< 7 @". d67B%
a waveguide loss of ".; d6% and an assumed beam3forming networ& loss of " d6% the
ma)imum :IR- is >/.; d67.
The shaped re2ector represents a new technology. Instead of illuminating a
conventional parabolic re2ector with multiple feeds in a beam3forming networ&% there
is a single feed that illuminates a re2ector with an undulating shape that provides the
required region of coverage. The advantages are lower spillover loss% a signifcant
reduction in mass% lower signal losses% and lower cost. 6y using large antenna
diameters% the rollo9 along the perimeter of the coverage area can be made sharp.
The practical application of shaped re2ector technology has been made possible by
the development of composite materials with e)tremely low coeCcients of thermal
distortion and by the availability of sophisticated computer software programs
necessary to analy=e the antenna. 1ne widely used antenna software pac&age is
called ARAS-% produced by TI!RA of !openhagen% Fenmar&. This program calculates
the gain from frst principles using the theory of physical optics.
SUMMARY
The gain of an antenna is determined by the intended area of coverage. The gain at a
given wavelength is achieved by appropriately choosing the si=e of the antenna. The
gain may also be e)pressed in terms of the half power beamwidth.
Re2ector antennas are generally used to produce narrow beams for geostationary
satellites and earth stations. The eCciency of the antenna is optimi=ed by the
method of illumination and choice of edge taper. -hased array antennas are used on
many *:1 and 4:1 satellites. New technologies include large% unfurlable antennas
for producing small spot beams from geostationary orbit and shaped re2ectors for
creating a shaped beam with only a single feed.
MMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMM
Dr. Robert A. Nelso! -.:. is president of Satellite :ngineering Research
!orporation% a satellite engineering consulting frm in 6ethesda% 4aryland. Fr. Nelson
is instructor for the courses &ate,,ite (omm-nication &./tem/ En0ineerin0: 1E2, ME2,
GE2 and &ate,,ite 3nd-/tr. 4pdate: Pre/ent &tat-/ and 5-t-re Trend/ o9ered by
Applied Technology Institute @mailto:[email protected])B. 0e is a *ecturer in the
Fepartment of Aerospace :ngineering at the (niversity of 4aryland and is Technical
:ditor of 6ia &ate,,ite maga=ine.