768CE.qxd jp.

1 1

Antenna Pattern
Measurement:
Concepts and Techniques
M I C H A E L D. F O E G E L L E

As high frequencies become more common, understanding antenna pattern

measurement and how to obtain useful measurements becomes critical.
he first article of this two-part series explores the basic to the relative information that makes up the antenna pattern

T concepts and techniques of antenna pattern measure-

ment and evaluates the benefits and drawbacks of var-
ious measurement methods. The concepts relating to near-
itself, and the various pieces of information that can be de-
termined from it, a variety of other results can be determined
from an active antenna system.
field and far-field pattern testing are discussed as well. The Although complex antenna-pattern measurement has been
second article (see page 34) presents the theory and equations a common requirement in the microwave antenna arena for
governing antenna properties and includes many years, it has only recently become more common to
a complete description of a site calibra- other areas such as electromagnetic compatibility
tion for pattern-measurement testing. (EMC) and wireless telecommunication. On the
Antenna pattern measurement EMC front, the interest in pattern measure-
refers to the determination of the ments appears to stem from a range of
radiation pattern of an antenna sources. The first is that, as EMC stan-
under test (AUT). It is the dards are forced to move higher in fre-
measurement of the relative quency, the effects of narrow-beam
magnitude and phase of an radiation from the equipment
electromagnetic signal re- under test (EUT) and the corre-
ceived from the AUT. Al- sponding interaction with the re-
though highly directional ceive antenna become increasingly sig-
antennas (i.e., horns) are nificant. It is important that the test
often measured by scanning a antenna is able to see all signals radiating
plane perpendicular to the from the EUT. In addition, broadband
bore-sight axis of the antenna antennas designed for EMC work are
(i.e., parallel to the face of the horn) finding their way into other applications
at some distance, this article focuses in which concern for antenna patterns
on total spherical pattern measurements. has always been an issue. Finally, many
A subset of this is the simple polar planar engineers with microwave back-
cut, in which the pattern is determined grounds now must deal with EMC is-
for a single azimuth rotation around the sues. These engineers want more in-
antenna. formation than has traditionally
Because a passive antenna is reciprocal, been provided on these antennas.
the pattern information could be ob- For the wireless industry, base sta-
tained by using it as either the transmit- tion antenna patterns have always
Illustration by TAISHA PAYTON

ter or receiver. This is in contrast to an been important in ensuring cover-

active antenna system, in which trans- age. Understanding the pattern of
mit and receive behavior may be con- each cell tower is critical to deter-
siderably different, and thus both rel- mining the required spacing be-
ative pattern and absolute power tween them. However, lately the
information is required. In addition industry has put considerable
Antenna Measurement

pendicular) field components. This measurement is usually

accomplished by using a dual-polarized horn, log-periodic
dipole array, or dipole antenna as the measurement antenna
(MA). Although it provides the best result, this technique re-
quires two receivers or the ability to automatically switch the
polarization of a single receiver, which can increase the cost
of the test. A slower, and possibly less accurate, option is to re-
peat an identical pattern test for each MA polarization. This
option could result in time variations and alignment issues
that could have significant effects.
Figure 1 shows a typical polar-pattern test setup. The AUT
(a cell phone in this case) is placed on a rotating turntable, and
a dual-polarized antenna is placed level with the AUT a fixed
distance away. The turntable is rotated 360°, and the response
Figure 1. Test setup for single-axis polar pattern measurement. between the antennas is measured as a function of angle. Nor-
mally, these measurements are performed in a fully anechoic
emphasis on handset pattern measurement as well. (simulated free-space) environment, but sometimes it may be
The Cellular Telecommunications and Internet Association desirable to measure the pattern over conducting ground, or
(CTIA) has drafted a set of test plans aimed at verifying the in some other as-used geometry to get real-world pattern in-
performance of cellular telephone handsets. One of the CTIA formation. Figure 2 shows some polar patterns for typical an-
plans provides tests for verifying radiated signal performance.1 tenna types and polarizations.
Previously, cell phones were required to meet a peak-signal To generate a full spherical-pattern measurement, it is nec-
requirement, but now they are required to meet a total radiated essary to change the relationship between the AUT and the
power requirement. This requirement ensures that a cell phone MA and repeat the previous polar test for each new orienta-
is transmitting energy in a broad pattern rather than in a nar- tion. The changes in orientation must be perpendicular to the
row beam and, therefore, is less likely to lose contact with the plane of measurement to completely cover a spherical surface.
cellular network. In simpler terms, the second axis of rotation must be perpen-
The tests are also designed to characterize both transmitted dicular to and intersect the first axis of rotation.
and received power and pattern, as well as the minimum sig- The two axes correspond to the θ and φ angles of the spher-
nal that the phone can properly detect. There are also calcula- ical coordinate system and are typically referred to as elevation
tions designed to determine the effectiveness of the phone and azimuth, respectively. Just as in the spherical coordinate
when the base station antennas are located along the horizon system, only one axis needs to be rotated through 360°, where-
(the typical configuration). The tests help to ensure that not all as the other is rotated only through 180°. With the proper pro-
of the radiated energy is directed up into space or down into cessing of the resulting data, it really does not matter which
the ground. axis is which. Either antenna can be rotated around this sec-
Whereas cell phone manufacturers are often interested in ond axis to generate the same pattern, but each technique has
the performance of the phone by itself, CTIA also requires both advantages and disadvantages.
testing with a liquid-filled phantom head or torso to simulate
the effect of human interaction with the phone. Conical-Section Method
In addition to cell phones, other products with growing The conical-section method uses an elevated turntable to
wireless testing requirements include wireless personal digital support the AUT and rotates the MA around the AUT on an
assistants, which are typically covered under the cellular re- axis perpendicular to the vertical rotational axis of the
quirements, and home- and office-based wireless networks turntable (see Figure 3). This method fits the geometric pic-
such as wireless local-area networks and Bluetooth devices. ture that most people have for spherical coordinate systems,
and, therefore, it is often the method used for pattern mea-
Measurement Techniques surements. The turntable continues to provide the azimuth
The basic pattern-measurement technique that most people (φ) rotation, whereas the MA is raised (elevated) or lowered in
are familiar with uses a single-axis rotational pattern. This an arc around the AUT, and, thus, the term elevation axis.
technique involves an AUT placed on a rotational positioner A common misconception when visualizing this technique
and rotated about the azimuth to generate a two-dimensional is to consider moving the MA in a 180° arc across the top of the
polar pattern. This measurement is commonly done for the AUT. However, a quick look at Figure 3 shows that this would
two principal axes of the antenna to determine parameters just duplicate the measurement across the top half of the AUT
such as antenna beam width in both the E and H planes. Such and never measure the bottom half of the pattern. The data
data are typically only measured for the copolar field compo- points at (φ = 0°, θ = +x°) and (φ = 180°, θ = –x°), where θ =
nent for simple horns or dipoles for which the general polar- 0° directly above the antenna, are the same.
ization of the pattern is well known. This method results in the MA describing circles of varying
For more-complicated radiators, for which the polarization diameter, and thus the reference to conical sections. The cir-
may not be known, or may vary as a function of angle, it is cles may be thought of as latitude lines on a globe, from the
important to be able to measure two orthonormal (i.e., per- north (+z) to south (–z) poles, with the largest circle located
 768CE.qxd jp.1 3

at the equator. Only the one

circle where the MA is at the
same height as the AUT (i.e.,
the equator) results in a true
polar pattern measurement.
Although the conical-sec-
tion method is conceptually
simple, it has a number of
drawbacks. A large pivot arm
or arch support is required to
manipulate the MA. For long
range lengths, this require-
ment can be a difficult
proposition. Similarly, if this
test is to be performed in a
fully anechoic chamber, the
chamber must be much larger
than would normally be nec-
essary to support the required
range length because the floor
and ceiling must be the same
distance away as the rear wall
behind the MA. This can dra-
matically increase the cost of
antenna measurement.
To perform a full surface
measurement, the turntable
must also be cantilevered out
from a wall or other support
to allow the MA to be moved
under the turntable. Other- Figure 2. Copolarized polar patterns for a vertically polarized dipole, horizontally polarized
wise, there will be a dead zone dipole, and standard-gain horn.
where the antenna is blocked
by the supporting structure. In any case, the turntable itself cover the entire sphere because the great circles cover the
can significantly affect the pattern measured if it is too mas- front and back of the sphere simultaneously.
sive or made of the wrong materials. With the shift in coordinate systems, the turntable is now an
elevation positioner rather than an azimuth positioner be-
Great-Circle Method cause it changes the MA position from pole to pole rather
For the great-circle method, the MA is fixed and the AUT is than along latitudinal lines parallel to the equator. The
repositioned on the turntable to generate each polar cut. Be- horizontal rotation axis of the AUT now provides the az-
cause the MA is fixed, pointing perpendicular to the rotation imuth positioning.
axis in this case, every cut is a true polar pattern. Therefore, The great-circle method has the advantage of being rela-
each rotation of the turntable provides the greatest diameter tively easy to perform with a low-cost system by rotating the
circle possible. AUT manually about the horizontal axis, but, as with most
To compare the two methods, the AUT must be laid on its such endeavors, it can be extremely tedious without additional
side with respect to the setup for the conical-section meth- automation. The method has an added benefit. The path be-
od to represent the associated shift in coordinate systems tween the AUT and MA is never obscured by the support struc-
(see Figure 4). ture, although care must be taken to ensure that the existing
By rotating the AUT about the horizontal axis between each support structure does not have reflective properties that could
great-circle cut, the entire spherical surface can be covered alter the antenna pattern, especially if additional material is
(see Figure 5). Each polar cut passes through the others at the required to support the AUT in different orientations.
horizontal axis of rotation, and the intersection points at the Finally, because the MA is fixed, the chamber only needs to
horizontal axis are equivalent to the top and bottom MA po- support the required range length in one dimension. This
sitions in the conical-section method. This is why the AUT opens the possibility of using tapered chambers and the like to
was laid on its side, to support the change in coordinates. obtain high performance and long range lengths affordably.
For the great-circle method, the circles can be thought of as
longitude lines, running from the north (+z) to the south (–z) Comparison of Methods
pole and back around the other side. As before, it is only nec- Although each method has advantages and disadvantages,
essary to rotate the AUT (instead of the MA) through 180° to it is important to verify that they are both capable of produc-
Antenna Measurement

Two-Axis Positioners
By adopting the great-circle method and manipulating the
AUT in two axes, it is possible to automate the test such that
data can be acquired according to the measurement sequence
of either method. Figure 7 shows a simple two-axis position-
er that can automate the rotation of the AUT on both axes. By
rotating the turntable (elevation) 360° and stepping the hor-
izontal axis (azimuth) of the AUT between each turntable ro-
tation, the great-circle method (see Figure 8a) can be dupli-
cated. Alternatively, by rotating the horizontal axis (azimuth)
of the AUT 360° and stepping the turntable (elevation), the
conical-section method (see Figure 8b) can be duplicated.
The two-axis positioner does suffer from one of the limita-
tions mentioned for the conical-section method. That is, for
Figure 3. Illustration of the conical-section method for spheri- some portion of the pattern (the south pole in Figures 7 and
cal antenna-pattern measurement. 8), the support structure is between the AUT and the MA.
This effect can be minimized by matching the support struc-
ture to the load being rotated, thereby reducing the amount
of interposing material to a minimum. Controlling the orien-
tation of the AUT with respect to the support can also im-
prove results. By making sure that the support is in a null or
back-lobe, its effects on pattern-related measurements can be
minimized.

Three-Dimensional Patterns
No matter which method is used to acquire the data, the
analysis of the result is made easier by the use of a three-
dimensional spherical plot to graph the output. Figure 9 gives
an example of a dipole pattern (a) and a standard-gain horn
pattern (b) plotted in three dimensions. This type of graphing
capability allows the pattern to be rotated around for different
Figure 4. Great-circle configuration of antenna under test.
views to help get an idea of the relative magnitude of the sig-
nal in various directions.

Near-Field versus Far-Field Measurements

Regardless of how the data are acquired, one of the available
system variables is the range length. Usually, when one refers
to the properties of an antenna, be it antenna pattern, gain, or
another property, the reference is to the far-field, free-space
properties of the antenna. In the far-field, free-space condition,
the measured properties of the antenna do not appear to vary
as a function of separation distance or antenna location. That
is not to say that the measured field levels themselves do not
vary, but that the measured gain or pattern does not vary.
To state it simply, the far-field, free-space condition is the
condition in which all of the theoretical equations typically
used for calculating antenna properties are valid.
Figure 5. Illustration of the great-circle method for spherical
antenna-pattern measurement. The back sides of the polar cuts
In a near-field or non-free-space environment, the antenna
have been removed for clarity. properties that are measured appear to vary as a function of
their environment. Effects such as mutual coupling between
the AUT and the measurement antenna or the antennas and
ing the same results. Figure 6 shows both conical section (a) other objects around them, as well as other near-field pertur-
and great circle (b) results with the same step size between bations, prevent the direct determination of the desired an-
measurement points and in which the coordinate systems have tenna properties. Even assuming a good free-space environ-
been aligned. Overlaying the two plots (see Figure 6c) shows ment (i.e., a fully anechoic chamber), there are still limitations
that the actual measured data points are identical, regardless to near-field testing.
of the method used. Therefore, given just the resulting data Most readers will be familiar with at least one rule of thumb
points (see Figure 6d), it is not possible to determine which for near- versus far-field determinations. In reality, there are
method was used to generate them. two very different definitions. The first, which is usually more
 768CE.qxd jp.1 5

The second far-field require-

ment, which is more famil- iar
to microwave engineers, is usu-
ally the dominant factor at
higher frequencies. In this case,
the objects involved (either the
actual antennas or larger de-
vices containing small anten-
nas) are large compared with
the wavelength.
The effects of scattering from
different points on the object,
or from different emissions
points in the case of an antenna
array or a leaky shielded enclo-
sure with multiple openings, re-
sult in wave fronts propagating
in multiple directions. The far-
field condition is met when all
of these different wave fronts
merge to form one wave front;
that is, when the multiple
sources are indistinguishable
from a single source (when sep-
aration distance r > 2D2/λ).
Therefore, the bigger the
object or the shorter the wave-
length, the farther away the re-
ceive antenna has to be for that
object to appear as a single
source. The region inside the
2D2/λ distance, but outside the
Figure 6. Comparison of measurement points between (6a) conical-section method and (6b) reactive near-field region, is
great-circle method. (6c) shows the two results overlaid, and (6d) indicates that it is impossi- referred to as the radiating
ble to tell which method was used given only the resulting data points. near-field or Fresnel region,
whereas the region outside
this distance is the far-field or
important at low frequencies, is represented by the near-field Fraunhofer region.2
term(s) of the electric and/or magnetic field equations. These In terms of antenna-pattern measurements, normally there
are the terms that behave as 1/rn, where n > 1. These terms rep- is little useful information to be gained within the reactive
resent the nonpropagating or evanescent electric and mag- region of an antenna. The one possible exception would be
netic fields—those caused by capacitively or inductively stored when the antenna is to be used in the reactive region as well.
energy in the antenna. Therefore, this region is referred to as
the reactive region of the antenna.2
The reactive fields decay rapidly with distance from the
→ →
antenna, leaving only the E × H term, which has a 1/r be-
havior. In this case, the far-field condition is satisfied by λ/r
<< 1, that is, where the measurement distance r is much
greater than wavelength λ. The reactive region is commonly
defined as

r < 0.62 √ D3 /λ ,

where D is the largest dimension of the radiating object. For

practical applications, a simple rule of thumb suitable for most
antennas is given by r < 2λ. Within this region, any measure-
ment antenna or probe would have a significant effect on the Figure 7. Example of a two-axis positioner setup for pattern-
transmit antenna. measurement testing.
 768CE.qxd jp.1 6
Antenna Measurement

However, it would not be testing. EMC emissions

possible to eliminate the are far from being con-
effect of the measure- tinuous wave, often con-
ment antenna on the sisting of harmonics,
AUT, and therefore the broadband noise, and
usefulness of such data spurious signals. Obtain-
would be limited. The ing the same radiation
Fresnel region contains signature at each point of
propagating electromag- a near-field scan is very
netic energy, but not in a unlikely.
cohesive form. There- To further complicate
fore, pattern measure- (a) (b) matters, low-frequency
ments done in this EMC measurements are
region can readily deter- Figure 8. (a) Great-circle method and (b) conical-section method per- often performed in the
mine quantities such as formed using the same two-axis positioner. reactive region of both
total radiated power but the EUT and the receive
may only provide an ap- antenna. Although near-
proximation of the far- field reactive terms can be
field pattern, gain, and easily determined for sim-
other properties. ple dipole elements, such
predictions for more-
Converting from complicated antennas or
Near Field to Far emitters are extremely dif-
Field ficult. The amount of data
A common practice in and processing required
microwave antenna mea- to correctly separate the
surements, and something effects of the EUT from
of a Holy Grail for EMC (a) (b) the receive antenna and
measurements, is the use the rest of the environ-
of near-field measure- Figure 9. Three-dimensional spherical plot of (a) simple dipole and (b) ment to truly predict a
ments to predict far-field standard-gain horn. Note the expected toroidal (donut) shape of the far-field result is far be-
results. In the Fresnel re- dipole pattern and the strong directionality and sidelobes of the stan- yond the current state of
gion, it is possible to scan dard-gain horn. the art.
the magnitude and phase
of the field along a closed Conclusion
surface (or, in the case of planar near-field scanning, an open The need for antenna-pattern information is increasing as
surface intersecting the vast majority of the propagating en- the EMC community moves to higher frequencies and more-
ergy) and predict the far-field levels. Acquiring the relative advanced techniques, and as wireless devices continue to per-
phase and magnitude at each point on the surface requires vade our everyday radio-frequency (RF) environment. The
the use of a reference signal in addition to the measurement techniques for complex-pattern measurement are rather
antenna signal. The fixed reference is needed to track the rel- straightforward, but there are some pitfalls. Useful pattern in-
ative phase of the signal in time because each point in space formation can be obtained using either the radiating near-
is not sampled at the same instant in time. field or far-field, but not the reactive, region of the AUT. The
For passive antennas, a vector network analyzer is normal- conversion of near-field pattern information to far-field results
ly used, which acquires both magnitude and phase information is possible, but it requires specialized software and measure-
against its own reference signal. Active devices are more com- ment capabilities.
plicated, requiring the use of a fixed reference antenna or sen-
sor in addition to the measurement antenna to obtain both References
phase and magnitude references (because an active device may 1. “Method of Measurement for Radiated RF Power and Receiver Per-
not maintain a constant magnitude or phase relationship). In formance, Draft Revision 1.2” (Washington, DC: CTIA, 2001).
either case, the calculations required to do the conversion are 2. CA Balanis, Antenna Theory, Analysis and Design (New York:
beyond the scope of this article. Harper & Row, 1982), 22–23.
For EMC testing, the conversion of radiated-emissions
measurements from near field to far field is made much more Michael D. Foegelle, PhD, is senior principal design engineer at
difficult by the nature of the electromagnetic signature of the ETS-Lindgren (Cedar Park, TX). He can be reached at 512-531-
device under test and the frequency range required for EMC 6444 or [email protected]. ■
 768CE.qxd jp.1 7

Antenna Pattern
Measurement:
Theory and Equations
M I C H A E L D. F O E G E L L E

The second installment on antenna pattern measurement describes the

calculations involved in determining properties such as TRP, EIRP,
directivity, and efficiency.

his is the second article in a two-part series on anten-

T na pattern measurement. This installment presents the

theory and equations governing a variety of antenna
properties and includes a complete description of a site cali-
bration for pattern-measurement testing.

Range Calibration
With a two-axis positioner setup, it is quite straightforward
to perform general pattern measurements and determine a
variety of relative data such as 3-dB beam width, front-to-
back ratio, and directivity. However, before accurate mea-
surements of values such as total radiated power (TRP), ef-
fective isotropic radiated power (EIRP), or antenna gain can
be made, it is necessary to perform a reference calibration to
correct for the various factors affecting these tests. The factors
include components such as range-length loss, gain of the re-
ceive antenna, cable losses, and so forth.
Normally, this calibration is done using a reference anten-
na (typically either a dipole or standard-gain horn) with
known gain characteristics. The reference antenna is mount-
ed at the center of the positioner as the antenna under test
(AUT) and adjusted to be at bore-sight level with the receive
antenna. The reference calibration is repeated for each polar-
ization of the receive antenna, with the reference antenna
polarized parallel to the corresponding receive element. Figure
1 shows a typical range-calibration setup and calls out vari-
ous components that are included in the measurement.
Typically, a signal generator or the output of a network an-
alyzer is connected to the reference antenna by one or more ca-
Illustration by TAISHA PAYTON

bles, possibly through a power amplifier. The receive antenna

is connected to a receiver or the input of a network analyzer
through one or more additional cables, possibly through a
preamplifier.
The power at the transmit antenna input port, Pt, is given in
terms of the signal generator output, PSG, by
 768CE.qxd jp.1 8
Antenna Measurement

mitted is the difference between

Pt Pr the applied forward incident
power and the power reflected
cl2 r back to the source:
Gt cl3
Gr
ga gpa Pnet = Pinc – Prefl.
(6)
PSG cl4 PRX
cl1
If a theoretical gain value is
Signal used in the Friis equation, then
Receiver
generator Pnet should be used for Pt be-
cause the theoretical formula
typically would not be able to
account for the voltage stand-
Figure 1. Some typical components of a range-calibration setup.
ing wave ratio (VSWR) caused
by the impedance mismatch. This requires either using a bidi-
PSGga rectional coupler and power meter configuration at the trans-
Pt = , (1) mit antenna to determine P directly, or measuring the
cl1cl2 net
VSWR of the antenna and performing additional calcula-
where ga is the gain of the amplifier, and cl1 and cl2 are the tions to predict the net power from the forward power.
cable losses of the corresponding transmit cables. If measured gain values are used, it is important to know
The power at the receiver, PRX, is given in terms of the power how those gain values were determined and whether they al-
at the receive antenna output port, Pr, by ready contain a contribution from the VSWR. Because any

Prgpa
PRX = , (2)
cl3cl4 The difference between
where gpa is the gain of the preamplifier, and cl3 and cl4 are the rms and peak field values
cable losses of the corresponding receive cables. can result in an immediate
If any of the components are missing, the corresponding
gain or loss for that variable in the equation should be 1. In 3-dB error in reported
terms of decibels, these formulas become measurement results.
Pt = PSG + ga – cl1 – cl2 (3)
calibration technique is inherently governed by this same for-
and mula, the resulting gain will be different depending on whether
VSWR effects have been accounted for separately. If not,
PRX = Pr + gpa – cl3 – cl4, (4) the gain will be changed simply by the ratio of net power to
forward power:
and the gain or loss of missing components would be 0 dB.
The Friis transmission equation governs the interaction be- PnetGt netGrλ2 Pinc (Gt net • Pnet/Pinc) Grλ2 PincGt incGrλ2
tween two antennas in the far field: Pr = = = (7)
(4πr)2 (4πr)2 (4πr)2
PtGtGrλ2 An impedance mismatch is just as likely to happen with the
Pr = , (5)
(4πr)2 receive antenna, leading to similar measurement issues, but it
would not be as easy to observe directly because, in this case,
where Pr is the power measured at the receive antenna output the reflected energy would be reradiated. There is no good
port; Pt is the power measured at the transmit antenna input way to measure the forward and reflected received energy.
port; Gt is the gain of the transmit antenna; Gr is the gain of the However, the VSWR of the receive antenna can be used to
receive antenna; λ is the wavelength; and r is the separation be- determine this effect. Fortunately, the gain of the receive
tween the two antennas (the range length).1, 2 antenna does not need to be known exactly (other than to
The exact definition of Pt is often a source of some confu- double-check the calibration result against theoretical pre-
sion and is somewhat dependent on what terms are included dictions) because it will be measured as part of the range
in the definition of gain. If the antenna is perfectly matched to calibration process.
the source cable, then all power applied to the antenna is As indicated in Figure 1, other factors are typically involved
radiated (or absorbed by losses in the antenna). However, in in the measurement, unless power meters and directional cou-
the more common case of a mismatch between the source plers are used right at the antennas to measure the net trans-
impedance and the antenna impedance, a portion of the en- mitted and received power. These factors include cable losses
ergy is reflected back to the source so that the net power trans- and the gain of any power amplifiers or preamplifiers.
 768CE.qxd jp.1 9

To minimize the uncertainty of resulting measurements, it where E is the electric field generated at the distance r from the
is usually desirable to perform the range calibration with all ca- transmit antenna, Pt is the power measured at the transmit
bles in place and use the same configuration for both calibra- antenna input port, Gt (θ, φ) is the angle-dependent gain of the
tion and pattern measurements. Should any component be transmit antenna, and r is the distance from the transmit an-
changed or damaged, the entire calibration must be redone. It tenna to the test point (the range length).1
is possible to perform individual calibrations on various system Combining the equation for the power density with that of
components, but each additional measurement increases the the electric field gives
total measurement uncertainty involved. Therefore, it is prefer-
able to calibrate the system as a whole whenever possible. PtGt(θ,φ)
ρ= . (13)
4πr2
TRP
To determine exactly how to apply the range calibration, it Combining this result with the equation for TRP gives
is important to make a comparison between the desired mea-
P π 2π
surement quantities and what will actually be measured by TRP = 4πt ∫θ=0 ∫φ=0 Gt(θ,φ) sin(θ)dθdφ. (14)
the test system. The primary quantity of interest is the TRP,
which can be obtained by integrating the time-averaged power
density of the radiated signal across the entire spherical surface Received Power
enclosing the AUT. Unfortunately, the receiver used to perform the test cannot
The time-averaged power density of a radiating signal is measure power density directly; instead, it measures received
given by the real part of the Poynting vector: power (again, neglecting cable losses, etc.). A related quantity
to the TRP would then be the total received power, given by in-
1 1 |E|2 |Erms|2 |Erms|2 tegrating the received power across all of the measurement
ρ= Re(E × H) = = = , (8)
2 2 η η 120π points of the AUT. The total power received is

where ρ is the time-averaged power density, E is the peak elec- π 2π

TPr = ∫θ=0 ∫φ=0 Prsin(θ)dθdφ, (15)
tric field strength, H is the peak magnetic field strength,
Erms is the root-mean-square (rms) electric field strength, and
η is the impedance of free space (120π).3, 4 where TPr is the total power received and Pr is the power mea-
The factor of 1⁄2 in the definition of the power density orig- sured at the receive antenna output port.
inates from the time averaging of the power across a complete The received power is given by the Friis transmission equa-
period. Although most reference materials and numerical anal- tion described earlier, so in terms of the transmit power and
ysis tools refer to wave magnitudes by their peak values, the angle-dependent gain, the equation becomes

E = Ee–jωt , (9) PtGrλ2 π 2π

TPr = ∫θ=0 ∫φ=0 Gt(θ,φ) sin(θ)dθdφ. (16)
(4πr)2
most measurement instrumentation reports rms values,
Because the desired value is TRP, the required correction fac-
1 tor is simply the ratio of TRP to the total power received:
Erms = E. (10)
√2
Pt π 2π
Therefore, when determining the power density from the rms ∫ ∫ G (θ,φ) sin(θ)dθdφ
TRP 4π θ=0 φ=0 t (17)
electric field, the factor of 1⁄2 has already been accounted for. = ,
TPr PtGrλ2 π 2π
The difference between rms and peak field values can result in ∫ ∫ G (θ,φ) sin(θ)dθdφ
an immediate 3-dB error in reported measurement results if (4πr)2 θ=0 φ=0 t
it is not treated correctly.
The TRP is given by integrating the power density across the which, when simplified, becomes
surface of the reference sphere:
TRP 4πr2 .
= (18)
π 2π TPr Grλ2
TRP = ∫ θ=0 ∫ φ=0 ρr2 sin(θ)dθdφ, (11)
This constant makes sense because the factor is related to the
where TRP is the total radiated power, ρ is the time-averaged range length and the gain of the receive antenna, both of which
power density, r is the radius of the sphere (the range length), are exactly what needs to be calibrated out of the system. Going
θ is the elevation angle, and φ is the azimuth angle. back to the Friis equation, the reference measurement
The electric field generated at a point in the far field as a performed with the reference antenna results in a site reference
function of the transmitted power is given by constant given by

√30PtGt(θ,φ) Pr GtGrλ2
E= , (12) C= = , (19)
r Pt (4πr)2
 768CE.qxd jp.1 10
Antenna Measurement

where C is the ratio of received power to transmitted power. nitude and phase. The magnitude of the reflection coefficient
Substituting this into the previous equation gives a correction is then
factor of
Vrefl VSWR –1
|ρ| = = . (24)
TRP G Vinc VSWR +1
= t . (20)
TPr 4πC
The transmission coefficient τ is defined as the ratio of trans-
The required site-calibration constant is now represented mitted to incident waves and is given by
in terms of the gain of the reference antenna and a single-
path loss measurement for each polarization. The ratio C VL
τ= , (25)
could contain contributions from other terms, such as V+
cable loss and so forth, as long as those contributions are
present in both the reference calibration and the pattern or, in terms of impedance,
measurements.
2Zo
τ= , (26)
Accounting for VSWR ZL + Zo
The treatment of the transmit antenna VSWR is an impor-
tant part of both the range calibration and the measurement where VL is the wave transmitted through the mismatch to
of various antenna properties. In general, VSWR is a mea- the load side (magnitude and phase).
surement of the mismatch between two transmission lines. By definition, τ – ρ = 1. However, the transmission coeffi-
It provides a measurement of the amount of signal being cient is not very useful for determining the net transmitted
reflected back from the mismatch, which is directly related to power from the VSWR because it also requires some knowl-
the amount of energy that is transmitted. edge of the impedance of the load. Although the necessary
For many antennas, the VSWR represents the largest com- information could be determined from the reflection co-
ponent of the antenna efficiency (the rest results from ohmic efficient, it is considerably easier to determine the ratio of
losses in the antenna itself). To determine the contribution the reflected power to the incident power, and then use that
from VSWR, it is necessary to calculate the ratio of the net to determine the net transmitted power:
power to the forward power.
VSWR is defined as the ratio of maximum to minimum Prefl Vrefl2 2
= 2 = |ρ| , (27)
voltage on the transmission line and is given by Pinc Vinc

Vmax Vinc + Vrefl so that

VSWR = = , (21)
Vmin Vinc – Vrefl
Pnet = Pinc – Prefl
where Vmax is the maximum voltage on the transmission line
(feed cable), Vmin is the minimum voltage on the transmission = Pinc – Pinc • |ρ|2 (28)
line, Vinc is the magnitude of the incident wave, and Vrefl is
the magnitude of the reflected wave.5 = Pinc (1 – |ρ|2).
The reflection coefficient ρ (not to be confused with the
power density described previously) is the ratio of reflected to This results in a VSWR correction factor given in dB by
incident waves and is given by
CVSWR = 10log10 (1 – |ρ|2)
V–
ρ= , (22) 4 • VSWR . (29)
V+ =10log10
(VSWR + 1)2
or, in terms of impedance,
The VSWR component covered here is not the only anten-
Z –Z na VSWR term related to antenna measurements. If an an-
ρ= L o, (23)
ZL + Zo tenna is not in a free-space environment, energy reflected back
from other objects will affect the VSWR measurement. How-
where V+ is the incident wave (magnitude and phase), V– is the ever, this term is a measure of the antenna’s interaction with
reflected wave (magnitude and phase), Zo is the characteristic its environment rather than a measurement of an inherent
impedance of the transmission line (magnitude and phase), property of the antenna.
and ZL is the impedance of the load line (magnitude and Care should be taken when measuring VSWR to be used
phase). for range calibrations to ensure that the measurement repre-
If the load impedance is equal to the characteristic sents a true free-space VSWR. A simple way to do this is to alter
impedance of the transmission line, the reflection coefficient the orientation and location of the reference antenna when
would be zero because there is no mismatch in this case. In ad- measuring VSWR. If no variation is seen in the resulting
dition, unlike VSWR, the reflection coefficient has both mag- VSWR measurements, then the environment probably does
 768CE.qxd jp.1 11

not have a significant effect. where ∆t(θ, φ) is the relative magnitude of the AUT pattern
at any angle with respect to the maximum.2 For an isotropic
Gain, Directivity, Efficiency, and EIRP radiator, ∆t(θ, φ) = 1, so that D = 1. For any real antenna,
Once the range has been calibrated, a number of anten- ∆t(θ, φ) < 1 for much of the surface, resulting in D > 1. Direc-
na properties can be determined from the pattern mea- tivity is the only term related to the antenna gain, which is
surement. The first property of interest is EIRP. EIRP is solely a relative term. Range calibration does not show up in
the power required for a theoretical isotropic radiator (one this equation.
that radiates the same power in all directions) to generate As with the TRP measurement, the measurement system is
the same field level in all directions as the maximum field only capable of measuring received power, so instead of EIRP,
seen from the AUT. Starting from the definition of TRP, the corresponding value calculated would be the effective
EIRP is given by isotropic received power:
π 2π
π 2π
EIPr = ∫θ=0 ∫φ=0 Pr max sin(θ)dθdφ
EIRP = ∫θ=0 ∫φ=0 ρmaxr2 sin(θ)dθdφ, (30) (35)
= 4πPr max ,
where ρmax is the maximum time-averaged power density
found over the surface of the measurement sphere. where Pr max is the maximum received power from the pattern
Assuming that the maximum power density can be defined measurement.
using the bore-sight gain of the AUT, Assuming again that the maximum received power is the
bore-sight transmission response, the same site-reference con-
PtGt stant, C, can be used:
ρmax = . (31)
4πr2
EIRP PtGt Gt
= = . (36)
Combine this with the equation for EIRP to get EIPr 4πPr max 4πC

π 2π PtGt It is apparent that the same range calibration holds in this

EIRP = ∫θ=0 ∫φ=0 r2 sin(θ)dθdφ, (32)
4πr2 case as well. Therefore, the directivity can also be represented
=PtGt . directly in terms of measured quantities as

EIRP is simply the transmitted power increased by the AUT EIPr

D= . (37)
gain, which brings some clarity to the definition of gain. Gain TPr
(over isotropic) is defined as the increase in received signal
from the AUT over that which would be received from an The efficiency of the AUT is defined as the ratio of TRP to
isotropic radiator with the same source power. Therefore, to APIP. The choice of defining APIP as incident power or net
create an isotropic radiator that generates the same field level power determines whether VSWR is part of the efficiency
as the maximum seen from the AUT, the source power must be term. If the net power definition is used, the efficiency only
increased by the gain. represents the ohmic losses of the antenna and not the mis-
Rearranging the equation for EIRP gives the definition of match effects:
gain:
TRP
ε= . (38)
EIRP Pt
Gt = , (33)
Pt
Comparing this to the definition of gain and directivity makes
where Pt is often referred to as the antenna-port input power it clear that gain is given by the product of directivity and
(APIP). Again, there is the question of whether this term efficiency:
should refer to the incident power or whether it should
refer to the net power. This decision affects the calculation of EIRP EIRP TRP
Gt = = = Dε. (39)
the efficiency of the antenna. Pt TRP Pt
The ratio of EIRP to TRP is defined as the directivity of the
antenna: If the AUT has no losses or mismatch, the directivity and gain
should be equivalent.
EIRP PtGt Other Antenna Properties
D= = π 2π
TRP Pt There are plenty of other properties that can be determined
4π ∫θ=0 ∫φ=0 Gt(θ,φ) sin(θ)dθdφ
from an antenna pattern, such as front-to-back ratio, average
(34) radiated power, average gain, and beam widths. The calculation
4π ,
= π 2π of most of these properties is straightforward, usually using
∫θ=0 ∫φ=0 ∆t(θ,φ) sin(θ)dθdφ simple formulas. The most important part of many of the cal-
culations is the data search algorithms used to find values like
the maximum point, minimum point, and –3-dB points.
 768CE.qxd jp.1 12
Antenna Measurement

Some of these antenna properties have little or no meaning draft form and subject to change, the details of these calcula-
for some antennas. In addition, the orientation of the AUT tions are not covered in this article.
can affect the result of an automated calculation without ad-
ditional input from the user to indicate the desired alignment Conclusion
information. For example, the meaning of E- and H-plane The techniques for complex pattern measurement are rather
beam widths is commonly understood. However, if an AUT is straightforward, but the calculations involved in determin-
randomly oriented for the pattern test, or has an unusual pat- ing certain antenna properties can be much more compli-
tern, there is no simple way to determine automatically what cated. Nonetheless, with appropriate care and understanding
constitutes each plane. of the associated quantities, it is not difficult to obtain excel-
lent results.
CTIA Requirements The information provided in this article can help even the
The Cellular Telecommunications and Internet Associa- novice RF or EMC engineer to determine a variety of an-
tion (CTIA) has developed some very specific antenna-prop- tenna properties.
erty requirements in addition to the EIRP and TRP measure-
ments.4 One of these is the near-horizon partial radiated References
power, which is used to determine the power radiated in a 1. “Antenna Calculations,” ETS-Lindgren Antenna Catalog (Cedar
small band (typically ±22.5° or ±45°) along the azimuth axis. Park, TX: ETS-Lindgren, 2002), 71.
This requirement is intended to determine how a cellular 2. CA Balanis, Antenna Theory, Analysis and Design (New York:
phone will interact with the network of cellular base stations Harper & Row, 1982), 29, 65.
arranged around it along the horizon during normal opera- 3. JD Jackson, Classical Electrodynamics, 2nd ed. (New York: Wiley,
tion. The orientation of the AUT will have a great effect on this 1975), 347.
result, so the standard calls out precise positioning require- 4. “Method of Measurement for Radiated RF Power and Receiver Per-
ments for the phone. formance, Draft Revision 1.2” (Washington, DC: Cellular Telecom-
Because a cellular phone has both transmit and receive munications and Internet Association, 2001).
modes, the CTIA standard also contains receive property re- 5. BC Wadell, Transmission Line Design Handbook (Boston: Artech
quirements, including total isotropic sensitivity (TIS) and House, 1991), 497.
near-horizon partial isotropic sensitivity (NHPIS), in addition
to the radiated pattern requirements. These values are calcu- Michael D. Foegelle, PhD, is senior principal design engineer at
lated from the received power pattern instead of the trans- ETS-Lindgren (Cedar Park, TX). He can be reached at 512-531-
mitted power pattern. Because the CTIA standard is still in 6444 or [email protected]. ■

Reprinted from Compliance Engineering, Annual Reference Guide 2002 • Copyright © 2002 Canon Communications LLC