4/26/2005 Wave Propagation present 1/1

III Antenna Fundamentals

Now we will discuss what occurs between the transmitter and
receiver. Recall this region is called the channel, and we
couple an electromagnetic wave to/from the channel using an
antenna.

A. Wave Propagation

We must first review the basics of electromagnetic

propagation in free-space.

HO: EM Wave Propagation in Free-Space

HO: The Poynting Vector

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 Electromagnetic Wave Propagation 1/9

Electromagnetic
Wave Propagation
Maxwell’s equations were cobbled together
from a variety of results from different
scientists (e.g. Ampere, Faraday), whose
work mainly was done using either static or
slowly time-varying sources and fields.

Maxwell brought these results together to form a complete

theory of electromagnetics—a theory that then predicted a
most startling result!

To see this result, consider first the free-space Maxwell’s

Equations in a source-free region (e.g., a vacuum). In other
words, the fields in a region far away from the current and
charges that created them:

∂E ( r,t )
∇xB ( r,t ) = µ 0ε 0
∂t

∂B ( r,t )
∇xE ( r,t ) = −
∂t

∇ ⋅ E ( r,t ) = 0

∇ ⋅ B ( r,t ) = 0

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 Electromagnetic Wave Propagation 2/9

Say we take the curl of Faraday’s Law:

∂∇xB ( r,t )
∇x∇xE ( r,t ) = −
∂t

Inserting Ampere’s Law into this, we get:

∂ ⎛ ∂E ( r,t ) ⎞
∇x∇xE ( r,t ) = − µ ε
⎜ 0 0 ⎟
∂t ⎝ ∂t ⎠
∂2E ( r,t )
= − µ 0ε 0
∂t 2

Recalling that if ∇ ⋅ E ( r ) = 0 then ∇x∇xE ( r )  ∇2E ( r ) , we can

write the following differential equation, one which describes
the behavior on an electric field in a vacuum:

∂2E ( r,t )
∇ E ( r ,t ) + µ 0ε 0
2
=0
∂t 2

This result is none as the vector wave equation, and is very

similar to the transmission line wave equations we studied at the
beginning of this class.

This result means that electric field E ( r ,t ) cannot be any

arbitrary function of position r and time t. Instead, an electric
field E ( r ,t ) is physically possible only if it satisfies the
differential equation above!

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 Electromagnetic Wave Propagation 3/9

Q: So, what are some solutions to this equation?

A: The simplest solution is the plane-wave solution. It is:

(
E ( r ,t ) = E x ˆx + E y ˆy e)j ω (t −z µ 0ε 0 )

For this solution, the electric field is varying with time in a

sinusoidal manner (that eigen function thing!), with an angular
frequency of ω radians/sec . Note this field is a function of
spatial coordinate z only, but the direction of the electric field
is orthogonal to the z-axis.

Q: What does this equation tell us about E ( r ,t ) ? What is this

electric field doing??

A: Lets plot Re {E ( r ,t )} as a function of position z, for

different times t, and find out!

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 Electromagnetic Wave Propagation 4/9

Re {E ( z ,t = 1 )}

t =1
z

Re {E ( z ,t = 2 )}

t =2
z

Re {E ( z ,t = 3 )}

z t =3

Here the red dot indicates plane of constant phase, for this
(
case a phase of zero radians, i.e., φ = ω t − z µ 0ε 0 = 0 . Note)
that this dot appears to be moving forward along the z- axis as
a function of time.

The electric field is moving !

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 Electromagnetic Wave Propagation 5/9

Q: How fast is it moving?

A: Lets see how fast the red dot (i.e., the plane of constant
phase) is moving! Rearranging ω (t − z µ 0ε 0 ) = 0 , we get the
position z of the dot as a function of time t :

t
z =
µ 0ε 0

Its velocity is just the time derivative of its position:

dz 1
vp = =
dt µ 0ε 0

Hey we can calculate this! The electric field is moving at a

velocity of:
1
vp =
µ 0ε 0
1
=
( 4π x10 )(8.854x10 )
-7 -12

= 3x10 8 ⎡ meters ⎤
⎢⎣ second ⎥⎦

Q: Hey wait a minute! 3 x 108

meters/second—that’s the speed of light!?!

A: True! We find that the magnetic field will likewise move in

the same direction and with the same velocity as the electric
field.
Jim Stiles The Univ. of Kansas Dept. of EECS
11/8/2006 Electromagnetic Wave Propagation 6/9

We call the combination of the two fields a propagating (i.e.,

moving) electromagnetic wave.

Light is a propagating electromagnetic wave!

This was a stunning result in Maxwell’s time. No one had linked

light with the phenomena of electricity and magnetism. Among
other things, it meant that “light” could be made with much
greater wavelengths (i.e., lower frequencies) than the light
visible to us humans.

Henrich Hertz first succeeded in creating and measuring this

low frequency “light”. Since then, humans have put this low-
frequency light to great use. We often refer to it as a “radio
waves”—a propagating electromagnetic wave with a frequency in
the range of 1 MHz to 20 GHz. We use it for all “wireless”
technologies !

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 Electromagnetic Wave Propagation 7/9

Given the results above, we can rewrite our plane-wave solution

as:
(
E ( r ,t ) = E x ˆx + E y ˆy e )
j ω (t −z µ 0ε 0 )

( )
= (E ˆy ) e
j ω t −z c
x
ˆx + E y
( )
= (E ˆy ) e
−j ωc z
x
ˆx + E y e j ωt

Now, making the definition:

ω 2π
k0 = = ⎡⎣radians / meter ⎤⎦
c λ

We get:

( )
E ( r ,t ) = E x ˆx + E y ˆy e − jk0z e j ωt

Q: This plane-wave solution reminds me somewhat of the

solution to the telegrapher’s equations, with k0 analogous to β .
Is this just a coincidence?

A: Nope! Since we have voltages and currents along our

transmission line, we must also have electric fields and magnetic
fields. In fact, the voltage and current wave solutions for a
transmission line can likewise be expressed as propagating
electric and magnetic (i.e., electromagnetic) fields.

But, there is one super-huge difference between the

transmission line solutions and the plane wave solution presented
here!
Jim Stiles The Univ. of Kansas Dept. of EECS
11/8/2006 Electromagnetic Wave Propagation 8/9

The propagating wave along at transmission line is constrained

to one of two directions—the plus z direction or the minus z
direction.

ˆ
−z ẑ

In contrast, nothing constrains a plane wave in free space—it

can propagate in any and all directions!
k̂4
k̂1 k̂3

k̂5
k̂2

Although the plane-wave solution shown above propagates in the

ẑ direction, the solution would be equally valid in the −y
ˆ
direction or x̂ direction, or any arbitrary direction k̂ .

The only constraint is that the direction of the electric field

vector be orthogonal to the direction of wave propagation, i.e.:

E ( r ,t )
kˆ ⋅ E ( r ,t ) = 0
k̂

Q: Are there any other solutions to this vector wave equation?

A: Plenty! Since the wave equation is a linear differential

equation, superposition holds.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 Electromagnetic Wave Propagation 9/9

In other words, a weighted sum of solutions is also a solution.

This means that we can (and often do) have multiple waves
propagating simultaneously in all different directions.

Moreover, there are many other solutions besides the plane-

wave solution. The most relevant of these, perhaps, is the
spherical wave:

( e
)
− jk r
E ( r ,t ) = Eθ (θ , φ ) ˆθ + Eφ (θ , φ ) ˆφ
0

e j ωt
r

Note the spherical wave is (most easily) expressed using the

spherical coordinate system (i.e., coordinates r , θ , φ and base
vectors rˆ, θˆ, φˆ ).

The spherical wave propagates outward

from the origin (i.e., in the direction
kˆ = rˆ ). In other words, a sphere of
constant phase (as opposed to a plane of
constant phase) propagates outward from
the origin. Thus, this sphere of constant
phase “expands” as a function of time—
sort of like a balloon being filled with air!

We likewise see from the expression above that the direction

of the electric field is likewise orthogonal to the direction of
wave propagation.
kˆ ⋅ E ( r ,t ) = rˆ ⋅ E ( r ,t ) = 0

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 The Poynting Vector 1/8

The Poynting Vector

Recall that plane waves and spherical waves are electro-
magnetic waves .

In other words, they consist of both electric and magnetic

fields!

Q: You provided us with the electric field representations of

plane and spherical waves, is there some way to use these to
determine the corresponding magnetic field?

A: You bet! Just apply Faraday’s Law:

∂ H ( r ,t )
∇x E ( r ,t ) = − µ 0
∂t

If the electric field of a plane wave is:

( )
E ( r ,t ) = E x ˆx + E y ˆy e − jk0z e j ωt

Then we find the magnetic field must be:

k0
H ( r ,t ) = µ 0
ω
( −E y ˆx + E x ˆy ) e − jk z e j ωt0

µ0
=
ε0
( −E y ˆx + E x ˆy ) e − jk z e j ωt
0

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 The Poynting Vector 2/8

Now, making the definition:

µ0
η0 = = 377 Ohms
ε0

We find the corresponding magnetic field for our plane wave

solution is thus:

( )
H ( r ,t ) = η0 −E y ˆx + E x ˆy e − jk0z e j ωt

The value η0 is know as the wave impedance, or sometimes

called the characteristic impedance of free space.

Q: Why is η0 referred to as an impedance? Does it really

have units of Ohms?

A: Consider the magnitude of both E ( r ,t ) and H ( r ,t ) :

2
E ( r ,t ) = E ( r ,t ) ⋅ E∗ ( r ,t )

( )( )
= e − jk0z e j ωt E x ˆx + E y ˆy ⋅ E x∗ ˆx + E y∗ ˆy e + jk0z e − j ωt

=e e j (ωt −ωt ) (E x E x∗ ˆx ⋅ ˆx + E y E x∗ ˆy ⋅ ˆx
− j (k0z −k0z )

+E x E y∗ ˆx ⋅ ˆy + E y E y∗ ˆy ⋅ ˆy )
2 2
= Ex + Ey

Therefore:
E ( r ,t ) =
2
Ex + Ey
2
V
m

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 The Poynting Vector 3/8

Using the same procedure for the magnetic field, we find:

1 2 2
A
H ( r ,t ) = Ex + Ey m
η0

Note that the magnitude of both the electric field and

magnetic field of a plane wave are constants with respect to
space and time!

Now, let’s take the ratio of these two values:

2 2
E ( r ,t ) Ex + Ey
= = η0
H ( r ,t ) 1 2 2
Ex + Ey
η0

The ratio of the electric and magnetic field magnitudes of a

single plane wave (but only for a single plane wave!) is wave
impedance η0 .

More importantly are the units of this value, which confirms

that it is indeed an “impedance” value.

E ( r ,t ) V
= η0 ⇒ m = V = Ohms
H ( r ,t ) A
m A

Now, let’s (finally!) get to the point (no pun intended) of this
handout—The Poynting Vector.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 The Poynting Vector 4/8

The Poynting Vector is defined as:

1
W (r ) = Re {E ( r ,t ) × H∗ ( r ,t )}
2

Note the Poynting Vector is a real-valued vector!

For our plane wave example, the Poynting Vector is:

1
W (r ) = Re {E ( r ,t ) × H∗ ( r ,t )}
2
1
=
2η0
Re E x x{( ˆ + Ey y ˆ × −E y∗ x ) (
ˆ + E x∗ y
ˆ )}
1
=
2η0
{(ˆ + Ey y
Re E x x ˆ + E x∗ y
ˆ × −E y∗ x ˆ) ( )}
=
1
2η0
{
Re E x
2
+ Ey
2
} zˆ
=
ˆ
z
2η0 (
Ex
2
+ Ey
2
)
ˆ
z 2
= E ( r ,t )
2η0

Q: Great. Do mind telling me what exactly this &%$!@ result

means!?!

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 The Poynting Vector 5/8

A: Let’s again do a dimensional analysis and see what we find.

Since the Poynting Vector is a (cross) product of an electric
field and a magnetic field, the units of the Poynting Vector
will be the product of V/m and A/m:

V A V ⋅ A Watts
⋅ = =
m m m2 m2

The Poynting Vector has units of Watts per square meter—

power per unit area. These are the units of power density.

Thus, the Poynting Vector describes the magnitude

and direction of the power flow associated with a
propagating electromagnetic wave.

This is why the Poynting Vector is a real-valued vector—

power is a real-valued quantity!

* The magnitude of the Poynting Vector (i.e., W ( r ) )

describes the power flow in terms of its spatial density.

For example, say a propagating wave has a power density of

5.0mW/m2 . Consider also a window whose surface area is 2
square meters.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 The Poynting Vector 6/8

If this electromagnetic wave is propagating toward this

window, then we will find that electromagnetic energy is
passing through this window at a rate of 10.0 milli-Joules
every second!

1m

⎛ W ⎞
2 ⎟(
P = ⎜ 5.0 2.0 m 2 )
W ⎝ m ⎠
W (r ) = 5.0 2m
m2 mJ
= 10
sec

* The direction of the Poynting Vector indicates the

direction of this power flow—the direction of the propagating
wave.

Note that the power density of a plane wave is a constant:

W (r ) =
ˆ
z
2η0 (
Ex
2
+ Ey
2
)
ˆ
z 2
= E ( r ,t )
2η0

In other words, the magnitude and direction of a plane-wave

Poynting Vector is identical at every point in the entire
universe!

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 The Poynting Vector 7/8

Q: Is this likewise true for all propagating electromagnetic

waves?

A: Absolutely not! For example, the Poynting vector of a

spherical wave is:

2 2
rˆ Eθ (θ , φ ) + Eφ (θ , φ )
W (r ) =
2η0 r2

Here we will make the definition:

2 2
Eθ (θ , φ ) + Eφ (θ , φ ) Watts
U (θ , φ ) =
2η0 Steradian

Therefore, the power density of a spherical wave is:

rˆ
W ( r ) = U (θ , φ )
r2

Clearly, this power density is not constant, but instead

diminishes (as 1 r 2 ) as we move away from the origin.

Q: So what’s up with this function U (θ , φ ) ?

A: The real, scalar function U (θ , φ ) is called the intensity of a

spherical wave. We will find that it is a very important
function in determining the performance of an antenna.

Jim Stiles The Univ. of Kansas Dept. of EECS

11/8/2006 The Poynting Vector 8/8

Q: Antenna? What does all this have to do with antennas?

A: A radiating antenna in fact launches a spherical wave.

The expression above thus describes the power density
produced by a radiating antenna (when located at the origin).

Jim Stiles The Univ. of Kansas Dept. of EECS