11/8/2007 Antenna Pattern notes 1/1

C. Antenna Pattern

Radiation Intensity is dependent on both the antenna and the

radiated power. We can normalize the Radiation Intensity
function to construct a result that describes the antenna
only. We call this normalized function the Antenna
Directivity Pattern.

HO: Antenna Directivity

The antenna directivity function essentially describes the

antenna pattern, from which we can ascertain fundamental
antenna parameters such as (maximum) directivity,
beamwidth, and sidelobe level.

HO: The Antenna Pattern

We find that conservation of energy requires a tradeoff

between antenna (maximum) directivity and beamwidth—we
increase one, we decrease the other.

HO: Beamwidth and Directivity

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Antenna Directivity 1/7

Antenna Directivity
Recall the intensity of the E.M. wave produced by the
mythical isotropic radiator (i.e., an antenna that radiates
equally in all directions) is:
Prad
U0 =
4π

Prad U (θ , φ ) = U0

Tx

But remember, and isotropic radiator is actually a physical

impossibility!

If the electromagnetic energy is monochromatic—that is, it is

a sinusoidal function of time, oscillating at a one specific
frequency ω —then an antenna cannot distribute energy
uniformly in all directions.

The intensity function U (θ , φ ) thus describes this uneven

distribution of radiated power as a function of direction, a
function that is dependent on the design and construction of
the antenna itself.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Antenna Directivity 2/7

U (θ , φ )
Prad

Tx

Q: But doesn’t the radiation intensity also depend on the

power delivered to the antenna by transmitter?

A: That’s right! If the transmitter delivers no power to the

antenna, then the resulting radiation intensity will likewise be
zero (i.e., U (θ , φ ) = 0 ).

Q: So is there some way to remove this dependence on the

transmitter power? Is there some function that is dependent
on the antenna only, and thus describes antenna behavior
only?

A: There sure is, and a very important function at that!

Will call this function D (θ , φ ) —the directivity pattern of the

antenna.

The directivity pattern is simply a normalized intensity

function. It is the intensity function produce by an antenna
and transmitter, normalized to the intensity pattern produced
when the same transmitter is connected to an isotropic
radiator.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Antenna Directivity 3/7

U (θ , φ ) intensity of antenna
D (θ , φ ) = =
U0 intensity of isotropic radiator

Using U0 = Prad 4π , we can likewise express the directivity

pattern as:
4π U (θ , φ )
D (θ , φ ) =
Prad

Q: Hey wait! I thought that this function was supposed to

remove the dependence on transmitter power, but there is Prad
sitting smack dab in the middle of the denominator.

A: The value Prad in the denominator is necessary to normalize

the function. The reason of course is that U (θ , φ ) (in the
numerator) is likewise proportional to the radiated power.

In other words, if Prad doubles then both numerator and

denominator increases by a factor of two—thus, the ratio
remains unchanged, independent of the value Prad .

Another indication that directivity

U (θ , φ ) pattern D (θ , φ ) is independent of the
D (θ , φ ) =
U0 transmitter power are it units. Note
that the directivity pattern is a
coefficient—it is unitless!

Jim Stiles The Univ. of Kansas Dept. of EECS

 11/8/2007 Antenna Directivity 4/7

Perhaps we can rearrange the above expression to make this

all more clear:
Dependent on
Prad antenna only.
U (θ , φ ) = D (θ , φ )
4π
Dependent on
Tx power and Dependent on
the antenna. Tx power

Hopefully it is apparent that the value of this function D (θ , φ )

in some direction θ and φ describes the intensity in that
direction relative to that of an isotropic radiator (when
radiating the same power Prad ).

For example, if D (θ , φ ) = 10 in some direction, then the

intensity in that direction is 10 times that produced by an
isotropic radiator in that direction.

If in another direction we find D (θ , φ ) = 0.5 , we conclude that

the intensity in that direction is half the value we would find
if an isotropic radiator is used.

Q: So, can the directivity function take any form? Are

there any restrictions on the function D (θ , φ ) ?

A: Absolutely! For example, let’s integrate the directivity

function over all directions (i.e., over 4π steradians).

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Antenna Directivity 5/7

2π π 2π π
U (θ , φ )
∫0 0
∫ D (θ , φ ) sin θ d θ d φ = ∫
0
∫0 U0 sin θ d θ d φ
2π π
1
=
U0 ∫ ∫U (θ , φ ) sin θ d θ d φ
0 0
2π π
4π
U (θ , φ ) sin θ d θ d φ
Prad ∫0 ∫0
=

4π
= (Prad )
Prad
= 4π

Thus, we find that the directivity pattern D (θ , φ ) of any and

all antenna must satisfy the equation:

2π π

∫ ∫ D (θ , φ ) sin θ d θ d φ = 4π
0 0

We can slightly rearrange this integral to find:

2π π
1
4π ∫ ∫ D (θ , φ ) sin θ d θ d φ = 1.0
0 0

The left side of the equation is simply the average value of

the directivity pattern (Dave ), when averaged over all
directions—over 4π steradians!

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Antenna Directivity 6/7

The equation thus says that the average directivity of any

and all antenna must be equal to one.

Dave = 1.0

This means that—on average—the intensity created by an

antenna will equal the intensity created by an isotropic
radiator.

Æ In some directions the intensity created by any and all

antenna will be greater than that of an isotropic
radiator (i.e., D > 1 ), while in other directions the
intensity will be less than that of an isotropic
resonator(i.e., D < 1 ).

Tx D (θ , φ )

D (θ , φ ) = 1

Q: Can the directivity pattern D (θ , φ ) equal one for all

directions θ and φ ? Can the directivity pattern be the
constant function D (θ , φ ) = 1.0 ?

A: Nope! The directivity function cannot be isotropic.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Antenna Directivity 7/7

In other words, since:

U (θ , φ ) ≠ U0
then:

U (θ , φ ) U 0
U (θ , φ ) ≠ U 0 ⇒ ≠ ⇒ D (θ , φ ) ≠ 1 . 0
U0 U0

Q: Does this mean that there is no value of θ and φ for which

D (θ , φ ) will equal 1.0?

A: NO! There will be many values of θ and φ (i.e., directions)

where the value of the directivity function will be equal to
one!

Instead, when we say that:

D (θ , φ ) ≠ 1.0

we mean that the directivity function cannot be a constant

(with value 1.0) with respect to θ and φ .

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 The Antenna Pattern 1/6

The Antenna Pattern

Another term for the directivity pattern D (θ , φ ) is the
antenna pattern. Again, this function describes how a
specific antenna distributes energy as a function of direction.

An example of this function is:

2
D (θ , φ ) = c (1 + cos φ ) sin 2θ

where c is a constant that must be equal to:

4π
c= 2π π
2
∫∫ ( 1 + cos φ ) sin 2θ sin θ d θ d φ
0 0

Do you see why c must be equal to this value?

Q: How can we determine the antenna pattern of given

antenna? How do we find the explicit form of the function
D (θ , φ ) ?

A: There are two ways of determining the pattern of a given

antenna

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 The Antenna Pattern 2/6

1. By electromagnetic analysis - Given the size, shape,

structure, and material parameters of an antenna, we can use
Maxwell’s equations to determine the function D (θ , φ ) .
However, this
analysis often must
resort to
approximations or
assumption of ideal
conditions that can
lead to some error.

2. By direct measurement - We can directly measure the

antenna pattern in the laboratory. This has the advantage
that it requires no assumptions or approximations, so it may
be more accurate.

However, accuracy
ultimately depends
on the precision of
your measurements,
and the result
D (θ , φ ) is provided as a table of measured data, as opposed to
an explicitly mathematical function.

Q: Functions and tables!? Isn’t there some way to simply plot

the antenna pattern D (θ , φ ) ?

A: Yes, but it is a bit tricky.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 The Antenna Pattern 3/6

Remember, the function D (θ , φ ) describes how an antenna

distributes energy in three dimensions. As a result, it is
difficult to plot this function on a two-dimensional sheet (e.g.,
a page of your notes!).

Antenna patterns are thus typically plotted as “cuts” in the

antenna pattern—the value of D (θ , φ ) on a (two-dimensional)
plane.

* For example, we might plot D (θ = 90 D , φ ) as a function of φ.

This would be a plot of D (θ , φ ) on the x-y plane.

* Or, we might plot D (θ , φ = 0 ) as a function of θ . This would

be a plot of D (θ , φ ) along the x-z plane.

Sometimes these cuts are plotted in polar format, and other

times in Cartesian.

Jim Stiles The Univ. of Kansas Dept. of EECS

 11/8/2007 The Antenna Pattern 4/6

Polar plot of
antenna cut
(
D θ = π 2,φ )
as a function
of φ .

Cartesian plot
of antenna cut
(
D θ = π 2,φ )
as a function
of φ .

The entire function D (θ , φ ) can likewise be plotted in 3-D for

either polar or Cartesian (if you have the proper software!).

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 The Antenna Pattern 5/6

Note that the majority of antenna patterns consist of a

number of “lobes”.
Side
Lobes
Main Lobe

Back
Lobes

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 The Antenna Pattern 6/6

Note these lobes have both a magnitude (the largest value of

D (θ , φ ) within the lobe), and a width (the size of the lobe in
steradians).

* Note that every antenna pattern has a direction(s) where

the function D (θ , φ ) is at its peak value. The lobe associated
with this peak value (i.e., the lobe with the largest magnitude)
is known as the antennas Main Lobe.

* The main lobe is typically surrounded by smaller (but

significant) lobes called Side Lobes.

* There frequently are also very small lobes that appear in

the pattern, usually in the opposite direction of the main lobe.
We call these tiny lobes Back Lobes.

The important characteristics of an antenna are defined by

the main lobe. Generally, side and back lobes are nuisance
lobes—we ideally want them to be as small as possible!

Q: These plots and functions describing antenna pattern

D (θ , φ ) are very complete and helpful, but also a bit busy and
complex. Are there some set of values that can be used to
indicate the important characteristics of an antenna pattern?

A: Yes there is! The three most important are:

1. Antenna Directivity D0 .
2. Antenna Beamwidth .
3. Antenna Sidelobe level.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Directivity and Beamwidth 1/11

Directivity and Beamwidth

One of the most fundamental of antenna parameters is
antenna directivity.

D0  Directivity

Q: Antenna directivity? Haven’t we already studied this?

Isn’t directivity D (θ , φ ) ?

A: NO! Recall that D (θ , φ ) is known as the directivity

pattern (a.k.a. the antenna pattern). Unlike the directivity
pattern D (θ , φ ) , which is a function of coordinates θ and φ ,
antenna directivity is simply a number (e.g., 100 or 20 dB).

Q: But isn’t antenna directivity somehow related to antenna

pattern D (θ , φ ) ?

A: Most definitely!

The directivity of an antenna is simply equal to the largest

value of the directivity pattern:

D0 = max {D (θ , φ )}
θ ,φ

Thus, the directivity of an antenna is generally determined

from the magnitude (i.e., peak) of the main lobe.

Jim Stiles The Univ. of Kansas Dept. of EECS

 11/8/2007 Directivity and Beamwidth 2/11

D0 (dB ) = 6

D (θ = 90 D , φ )
(dB )

Note that directivity is likewise a unitless value, and thus is

often expressed in dB.

Another fundamental antenna parameter is the antenna

beamwidth.
ΩA  beamwidth ⎡⎣steradians ⎤⎦

Just like the “bandwidth” of a microwave device, antenna

beamwidth is a subjective value. Ideally, we can say that the
beamwidth is the size of the antenna mainlobe, expressed in
steradians.

Q: But how do we define the “size” of the mainlobe?

A: That’s the subjective part!

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Directivity and Beamwidth 3/11

Sometimes, we define beamwidth

as the “null-to-null” beamwidth:

null-to-null
beamwidth

But much more common is

the 3dB beamwidth,
defined by the points on
the mainlobe where the
directivity pattern D (θ , φ )
has a value of one half
that of value directivity
D0 (i.e., 3 dB less than
D0 (dB ) ):
3 dB
beamwidth

Jim Stiles The Univ. of Kansas Dept. of EECS

 11/8/2007 Directivity and Beamwidth 4/11

D0 (dB ) = 6

D (θ = 90 D , φ )
(dB )

3 dB φ
Beamwidth

Q: But how do we determine the antenna beamwidth ΩA ?

A: Theoretically, we can use either of the beamwidth

definitions above and integrate over all directions θ and φ
that lie within the mainlobe:

ΩA = ∫∫ sin θ d θ d φ
main
lobe

However, we more often use an approximation to determine

the antenna beamwidth. If the sidelobes of an antenna are
small, then we can approximate its directivity pattern as:

⎧D0 within the mainbeam

⎪
D (θ , φ ) ≈ ⎨
⎪0 outside the mainbeam
⎩

Jim Stiles The Univ. of Kansas Dept. of EECS

 11/8/2007 Directivity and Beamwidth 5/11

approx. D (θ , φ )

D (θ = 90 D , φ )
(dB )

approx. D (θ , φ )

In other words, this approximation “says” that the antenna

radiates its power uniformly throughout the mainlobe, but
radiates no energy in any other direction.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Directivity and Beamwidth 6/11

This of course is a fairly rough approximation, but we can use

it to determine (approximately) the antenna beamwidth ΩA .

To see how, first recall that the average directivity of any

antenna (averaged over 4π steradians) is:

2π π
1
1. 0 = ∫ ∫ D (θ , φ ) sin θ d θ d φ
4π 0 0

Inserting our approximation into this integral, we find:

1 1
1.0 = ∫∫ D (θ , φ ) sin θ d θ d φ + ∫∫ D (θ , φ ) sin θ d θ d φ
4π main 4π side
lobe lobe

1 1
= ∫∫ D0 sin θ d θ d φ + ∫∫ 0 sin θ d θ d φ
4π main 4π side
lobe lobe

D0
= ∫∫ sin θ d θ d φ
4π main
lobe

Look! Recall the integral above is the beamwidth of the

antenna:
ΩA = ∫∫ sin θ d θ d φ
main
lobe

And so:
D0 D0
1. 0 = ∫∫ sin θ d θ d φ = ΩA
4π main 4π
lobe

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Directivity and Beamwidth 7/11

Rearranging, we find an important result:

D0 ΩA = 4π

This says that the product of the antenna directivity and

antenna beamwidth is a constant (i.e., 4π).

Q: So what?

A: This means—yet again—that we cannot “have our cake and

eat it too”! If we increase the directivity of an antenna, then
its beamwidth must decrease.

Conversely, if we increase antenna

beamwidth, its directivity must diminish
proportionately.

This of course makes sense; we can increase

directivity only by “crushing” the available power
into a smaller solid angle (i.e., the main lobe beamwidth ΩA ).

Moreover, the expression above allows us to determine—given

beamwidth ΩA —the (approximate) value of antenna
directivity:
4π
D0 =
ΩA

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Directivity and Beamwidth 8/11

Note from this equation we can “define” antenna directivity as

the ratio of the beamwidth of an isotropic radiator (4π) to
the beamwidth of the antenna ( ΩA )!

4π beamwidth of isotropic radiator

D0 = =
ΩA beamwidth of antenna

Likewise, we can—given antenna directivity D0 —determine the

antenna beamwidth:

4π
ΩA =
D0

Thus, by simply determining the maximum value of function

D (θ , φ ) (i.e., D0 ), we can easily determine an approximate value
of antenna beamwidth (in steradians) using the equation shown
above!

Q: Now, ΩA tells us the size of the mainlobe solid angle (in

steradians), but it does not tells its shape. Didn’t you say
that solid angles with different shapes can have the same
size ΩA ?

A: That’s exactly correct!

Jim Stiles The Univ. of Kansas Dept. of EECS

 11/8/2007 Directivity and Beamwidth 9/11

Recall that our 3-D beam pattern D (θ , φ ) is often plotted on

two, orthogonal 2-D planes. We can define the beamwidth on
each of these two planes in terms of radians (or degrees).

For example, we might plot D (θ , φ ) on the x-y plane (i.e.,

D (θ = π 2 , φ ) ) and find that its (2-D) 3dB beamwidth has a
value (in radians) that we’ll call βφ .

D0 (dB ) = 6

D (θ = 90 D , φ )
(dB )

φ
βφ

We could likewise plot D (θ , φ ) on the x-z plane (i.e.,

D (θ , φ = 0 ) ) and find that its (2-D) 3dB beamwidth has a
value (in radians) that we’ll call βθ .

We find that antenna beamwidth is often expressed in terms

of these two angles ( βθ and βφ ), as opposed to the value of
the solid angle ΩA in steradians.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Directivity and Beamwidth 10/11

Finally, a third fundamental antenna parameter that we can

extract from antenna pattern D (θ , φ ) is the peak sidelobe
level.

This provides a measure of the magnitude of the sidelobes, as

compared to the directivity of the mainlobe. Say we define
the largest value of D (θ , φ ) found in the sidelobes (i.e.,
outside the mainlobe) as the peak sidelobe directivity:

Dsl  max {D (θ , φ )}
side
lobes

Peak Sidelobe

We can then normalize this value to antenna directivity D0 .

This value is known as the peak sidelobe level, and is typically
expressed in dB:

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2007 Directivity and Beamwidth 11/11

⎡D ⎤
peak sidelobe level  10 log10 ⎢ sl ⎥
⎣ D0 ⎦

Sidelobes are generally considered to be a non-ideal artifact

in antenna patterns. Essentially, sidelobe levels represent a
waste of energy—electromagnetic propagation in directions
other than the desired direction of the mainlobe.

Thus, we generally desire a peak sidelobe level that is a small

as possible (e.g., < -40 dB).

Jim Stiles The Univ. of Kansas Dept. of EECS