EITN90 Radar and Remote Sensing

Lecture 5: Target Reflectivity

Daniel Sjöberg

Department of Electrical and Information Technology

Spring 2020
 Outline

1 Basic reflection physics

2 Radar cross section definition

3 Scattering regimes

4 High-frequency scattering

5 Examples

6 Conclusions

2 / 57
 Learning outcomes of this lecture
In this lecture we will
I Study the properties of electromagnetic waves
I Define the radar cross section (RCS)
I Understand basic scattering and reflectivity physics
I Understand how two or more scattering centers interfere
I Illustrate some high-frequency scattering effects
Target
Transmit signal Antenna

Transmitter T/R

Receive
signal

Receiver
protector
switch Mixer/Preamplifier
Detection and
Signal
Detector A/D measurement
processor
results

Low noise Local

IF amplifier
amplifier oscillator
Receiver (Adapted from Fig. 1-1)

3 / 57
 Outline

1 Basic reflection physics

2 Radar cross section definition

3 Scattering regimes

4 High-frequency scattering

5 Examples

6 Conclusions

4 / 57
 Electromagnetic waves
The time-harmonic Maxwell’s equations in linear media are (where
all fields have time dependence E(x, y, z, t) = E(x, y, z)ejωt )
(
∇ × E = −jωB = −jωµH
∇ × H = jωD = jωE

where  and µ are the permittivity and permeability of the

material. A plane wave propagating in direction k is given by
(where we use ∇e−jk·R = −jke−jk·R , or ∇ → −jk)
( (
E = E 0 ej(ωt−k·R) k × E 0 = ωµH 0
⇒
H = H 0 ej(ωt−k·R) k × H 0 = −ωE 0

The cross products mean (E 0 , H 0 , k) is a right-hand triple like

√
(x̂, ŷ, ẑ). In addition, the equations imply k = ω µ and the
p
ratio |E 0 |/|H 0 | = µ/ = η, where η is the wave impedance.
5 / 57
 Vector analysis, linear algebra
The vectors have three components, one for each spatial direction:
E = Ex x̂ + Ey ŷ + Ez ẑ
In particular, the position vector is r = xx̂ + yŷ + zẑ. Vector
addition, scalar product and vector product are
r1 × r2
r1 r2
r2
r2 ϕ ϕ
r1 + r2 r1 r1
|r 2 | cos ϕ

I Addition: r 1 + r 2 = (x1 + x2 )x̂ + (y1 + y2 )ŷ + (z1 + z2 )ẑ.

I Scalar product: r 1 · r 2 = |r 1 | |r 2 | cos ϕ = x1 x2 + y1 y2 + z1 z2 .
I Vector product: orthogonal to both vectors, with length
|r 1 × r 2 | = |r 1 | |r 2 | sin ϕ, and r 1 × r 2 = −r 2 × r 1 .
x̂ × ŷ = ẑ, ŷ × ẑ = x̂, ẑ × x̂ = ŷ
6 / 57
 Vector analysis, differentiation
(not necessary to understand the book)
The nabla operator is
∂ ∂ ∂
∇ = x̂ + ŷ + ẑ
∂x ∂y ∂z

The gradient, divergence, and curl operations are

∂g ∂g ∂g
∇g = x̂ + ŷ + ẑ
∂x ∂y ∂z
∂ ∂ ∂
∇·E = Ex + Ey + Ez
∂x ∂y ∂z
∂ ∂ ∂
∇×E = x̂ × E + ŷ × E + ẑ × E
∂x ∂y ∂z
 
Ex
Cartesian representation (useful in numerics): [E] = Ey and
Ez
        
0 0 0 Ex 0 0 1 Ex 0 −1 0 Ex
∂ ∂ ∂
0 0 −1 Ey  +  0 0 0 Ey  + 1 0 0 Ey 
[∇ × E] = ∂x ∂y ∂z
0 1 0 Ez −1 0 0 Ez 0 0 0 Ez
| {z } | {z } | {z }
=[x̂×E] =[ŷ×E] =[ẑ×E]
7 / 57
 Misprinted equations

Unfortunately, the vector cross product sign, ×, is sometimes not

printed in the book. The correct versions of the affected equations
are below:
I (6.5) P = 12 Re(E × H ∗ ) W/m2
I (6.6) E total × n̂ = (E inc + E scat ) × n̂ = 0
I Page 216, paragraph 2, line 3: n̂ × E scat = −n̂ × E inc

Comment: the vector product has amplitude

|n̂ × E| = |n̂||E| sin ϕ = |E| sin ϕ

where ϕ is the angle between the unit vector n̂ and the vector E.
Hence, n̂ × E represents the part of E orthogonal to n̂ (tangential
to the surface if n̂ is a surface normal).

8 / 57
 Frequency and wavelength

x
λ
H
z
y
E

The electric field is E(x, y, z, t) = E0 cos(ωt − kz)x̂. The

wavelength λ is the periodicity in z, determined by (using ω = 2πf
and ωt − kz = ω(t − kz/ω) = ω(t − z/v))
ω 2π
λf = v ⇒ k= =
v λ
The polarization corresponds to the direction of the electric field.
The wave depicted above is linearly polarized in the x-direction.
9 / 57
 Polarization

Vertical Horizontal

Right hand Left hand

See the animation program EMANIM.
10 / 57
 Right-hand rules
Let the propagation direction k be along the thumb. At any time,
E, H, and k are orthogonal to each other.

Linear polarization Circular polarization

rotation
V
H direction

k
k

E oscillating along Horizontal E rotating along Right or Left

or Vertical direction, H along hand fingers, H rotating the
the other. same but at right angle.
11 / 57
 IEEE definition of left and right
With your right hand thumb in the propagation direction and
fingers in rotation direction in a fixed plane: right hand circular.

x x
E(t) right-polarized E(t) left-polarized
forward moving forward moving
z z
y y

x x
E(t) left-polarized E(t) right-polarized
backward moving backward moving
−z −z
y y

12 / 57
 Refractive index and wave impedance
From the expression of a plane wave
k k
E = E 0 ej(ωt−k·R) = E 0 ejω(t− ω k ·R)
where k/k = k̂ is the propagation direction, it is seen that only the
projection of the position vector R in the propagation direction k̂,
k̂ · R, matters, and the speed of propagation is
r
ω 1 c µ
v = = √ = , where n =
k µ n 0 µ0
The refractive index n is the speed of an electromagnetic wave in a
material (v), relative to the speed in vacuum (c). The wave
impedance in the material is given by (denoted by η or Z)
r
|E| µ
= =η=Z
|H| 
Discussion

13 / 57
 Plane waves vs spherical waves

The plane wave is a beautiful theoretical tool, but it typically only

applies locally around the target or transmitter/receiver.

transmitter D target

The wavefront is spherical close to the transmitter, but

approximately plane at the target if the range R satisfies

2D2
R≥
λ

where D is the diameter of the target and λ the wavelength.

14 / 57
 Induced currents
Assume the target is made of metal. The electrons move around so
as to cancel the field inside the metal, quantified through the
boundary condition (zero tangential electric field)
E tot × n̂ = (E inc + E scat ) × n̂ = 0

A more general boundary condition is E tot

tan = ZS n̂ × H
tot
, where
ZS is the surface impedance (ZS = 0 for perfect conductors). This
does not change the general results in this lecture. 15 / 57
 Radiation
Since the incident field is oscillating like ejωt , the induced surface
currents J S and surface charges ρS on the target will oscillate with
the same frequency and radiate a scattered field E scat .

Z  
ρS (Rs )
E scat (Rf ) = −jωµJ (Rs )g(Rf − Rs ) + ∇g(Rf − Rs ) dS(Rs )

−jk|R −Rs |
The function g(Rf − Rs ) = e4π|Rf −R
f
s|
is called the Green’s
function, corresponding to radiation at field point Rf from a unit
point source at Rs .
16 / 57
 Scattering theory definitions
At large range from the scatterer, the scattered field can be written
e−jkR scat
E scat = F (k̂ )
R
scat
where F (k̂ ) is the far field amplitude in scattering direction
scat
k̂ . Given knowledge of scattered electric and magnetic fields E s
and H s on a surface S enclosing the scatterer, this is
Z h
jk i
F (k̂) = k̂ × k̂ × (n̂ × η0 H s ) + E s × n̂ ejkk̂·r dS
4π S

The bistatic scattering cross section is

 scat scat 2 scat 2
   
E (k̂ ) F (k̂ )

inc scat
σ∗ (k̂ , k̂ ) = lim 4πR2   = 4π  inc 2
R→∞  inc inc 2 E 
E (k̂ )

and the monostatic radar cross section at incident direction k̂ is

σ(k̂) = σ∗ (k̂, −k̂)
17 / 57
 Simulations

The scattering theory can be implemented numerically. This

provides excellent tools to compute and visualize the scattered
field.

http://falstad.com/emwave2/ http://www.comsol.com

18 / 57
 Outline

1 Basic reflection physics

2 Radar cross section definition

3 Scattering regimes

4 High-frequency scattering

5 Examples

6 Conclusions

19 / 57
 Radar cross section
In this chapter, power density is denoted by P instead of Q as in
Chapter 1.

The IEEE definition of RCS is

def |E scat |2
σ = lim 4πR2
R→∞ |E inc |2

which is motivated by intercepted power σP inc and isotropically

radiated scattered power density P scat = (σP inc )/(4πR2 ).
20 / 57
 Factors affecting the RCS
The bistatic RCS of a target depends on the following factors:
I Target geometry and material composition.
I Direction of transmitter relative to target.
I Direction of receiver relative to target.
I Frequency or wavelength.
I Transmitter polarization.
I Receiver polarization.

In a monostatic setting, transmitter and receiver are co-located.

Discussion
21 / 57
 Polarization scattering matrix
It is sometimes necessary to keep track of polarization. In general,
the polarization scattering matrix (PSM) S is defined by
scat scat inc inc
E scat (k̂ ) = S(k̂ , k̂ ) · E inc (k̂ )
scat inc
where the arguments k̂ and k̂ are often left out to write
scat inc
E = S · E . The PSM can be represented in linear
polarization
 scat     inc 
EV SVV SVH EV
scat = S
EH SHH inc
EH
HV

or circular polarization
 scat     inc 
ER SRR SRL ER
=
ELscat SLH SLL ELinc

22 / 57
 Converting between polarizations
The relation between LP and CP in transmission is
t
    t  t
   t 
ER 1 1 j EH EH 1 1 1 ER
=√ ⇔ =√
ELt 2 1 −j t
EV t
EV 2 −j j ELt

and in reflection we have (due to different propagation direction)

r
    r  r
   r 
ER 1 1 −j EH EH 1 1 1 ER
=√ ⇔ =√
ELr 2 1 j r
EV r
EV 2 j −j ELr

Hence, the polarization scattering matrix in LP and CP are related

by
     
SRR SRL 1 1 −j SHH SHV 1 1
=
SLR SLL 2 1 j SVH SVV −j j

Note that the sense of rotation of circular polarization changes

with each reflection.
23 / 57
 Outline

1 Basic reflection physics

2 Radar cross section definition

3 Scattering regimes

4 High-frequency scattering

5 Examples

6 Conclusions

24 / 57
 Sphere scattering

In the following slides, scattering from an oblong object is shown.

Note the outmost spherical shell corresponds to a material layer
absorbing outgoing radiation, and does not correspond to a
physical region. 25 / 57
 Low frequency: Rayleigh scattering

f = f0 /8

f = f0 /4

f = f0 /2

Incident field Scattered field Total field

26 / 57
Intermediate frequency: resonant scattering

f = f0

f = 2f0

f = 4f0

Incident field Scattered field Total field

27 / 57
 High frequency: optical scattering

f = 8f0

f = 16f0

f = 32f0

Incident field Scattered field Total field

28 / 57
Scattering mechanisms

29 / 57
 Outline

1 Basic reflection physics

2 Radar cross section definition

3 Scattering regimes

4 High-frequency scattering

5 Examples

6 Conclusions

30 / 57
 Superposition from several scatterers

E 0 e−jk·R k̂ · R1
R1
k̂ · R2
R2

R3
k̂ · R3

When several scatterers are subjected to an incident wave

E 0 e−jk·R , the backscattering is (complex addition)
 2
N
X √ −j2k·Ri 
σtot =  σi e


 
i=1

31 / 57
 Phasor addition

Two scatterers spaced by 2λ. Strong interference when scatterers

have equal amplitude, dominated by the strong scatterer when
they are very different.

32 / 57
 Specular scattering
When the surface normal n̂ of a relatively flat surface points
toward the radar, there is little variation of k̂ · R over the surface.
Hence, the phase does not change much, and we have coherent
addition:

A2cp
σspecular = 4π 2 , Acp = area of constant phase
λ
r
Rc λ
Lcp ∼ length of constant phase, Rc = radius of curvature
2
σ = πRc1 Rc2 double curved surfaces
2π
σcyl = Rc L2 cylindrical surface Discussion
λ 33 / 57
 End-region scattering

Scattering from a metal plate shows significant side-lobes. At

off-specular directions, only the edges are scattering in the
back-direction.

The effect is due to truncation of currents on the flat plate.

34 / 57
Metal plate at different orientations

Scattering from
edges is stronger
than from corners.

35 / 57
 Edge diffraction
When a wave is incident on an edge, a line source current is
induced. At oblique incidence it radiates in a cone.

Monostatic return only at normal incidence, σ ≈ L2 /π.

36 / 57
 Multiple bounces
When two specular reflections combine at 90◦ angle, strong
backscattering occurs, so called corner reflectors.

37 / 57
 Corner reflections

Incident field Scattered field Total field

(Corner reflector, cat eye)

38 / 57
 Multiple bounces

The specular reflections do not need to be at flat surfaces. Many

different combinations can occur.

When designing stealthy objects, it is important to find shapes

with as little corner reflections as possible.
39 / 57
 Outline

1 Basic reflection physics

2 Radar cross section definition

3 Scattering regimes

4 High-frequency scattering

5 Examples

6 Conclusions

40 / 57
 Metal plate, different methods
Diffraction important at low levels of scattering.

41 / 57
A7 aircraft

42 / 57
 A7 aircraft

Discussion
43 / 57
A7 aircraft

44 / 57
Stovepipe aircraft

45 / 57
Stovepipe aircraft

46 / 57
Stovepipe aircraft

47 / 57
Stovepipe aircraft

48 / 57
Stovepipe aircraft

49 / 57
Stovepipe aircraft

50 / 57
Stovepipe aircraft

51 / 57
 Outline

1 Basic reflection physics

2 Radar cross section definition

3 Scattering regimes

4 High-frequency scattering

5 Examples

6 Conclusions

52 / 57
 Conclusions

I We have reviewed basic scattering theory and how it relates to

RCS.
I Three different scattering regimes: Rayleigh, resonance,
optical.
I Interaction between multiple targets.
I Scattering mechanisms: dipole, surface waves, specular,
multiple bounces, end regions, edge diffraction, discontinuities.

53 / 57
 Discussion
Pair up the different materials with the corresponding relative
permittivity!
1. Vacuum A. 2.1
2. Teflon B. 5000
3. Rubber C. 7
4. Barium titanate D. 1

Go back 54 / 57
 Discussion
Pair up the different materials with the corresponding relative
permittivity!
1. Vacuum A. 2.1
2. Teflon B. 5000
3. Rubber C. 7
4. Barium titanate D. 1
Answer:
1. Vacuum D. 1
2. Teflon A. 2.1
3. Rubber C. 7
4. Barium titanate B. 5000
A rule of thumb is that permittivity is higher the denser the
material. BaTiO3 is a ferro-electric material, which has
exceptionally high permittivity and is used in capacitors.
Go back 54 / 57
 Discussion (tough one!)
The scattering in the forward direction is significantly larger than
in other directions. Why?

Go back

55 / 57
 Discussion (tough one!)
The scattering in the forward direction is significantly larger than
in other directions. Why?

Answer: For a large object, we expect a shadow. The scattering in

the forward direction needs to cancel the incident field in order to
create a shadow, E tot = E inc + E scat ≈ 0. Since E inc = E 0 e−jk·R
does not decrease with distance but E scat does, it needs to be
sharply focused to maintain the shadow at large distance.
Go back

55 / 57
 Discussion
The face below is illuminated from the front and below by a
60 GHz radar (λ = 5 mm). Numerical mesh on the left, phase
distribution on the right. Identify some of the strongest specular
reflections!

Go back

56 / 57
 Discussion
The face below is illuminated from the front and below by a
60 GHz radar (λ = 5 mm). Numerical mesh on the left, phase
distribution on the right. Identify some of the strongest specular
reflections!

Answer: Between the eye brows, on the nose tip, on the upper lip,
and on the chin.
Go back

56 / 57
 Discussion
Why is the RCS so sensitive to angle?

Go back
57 / 57
 Discussion
Why is the RCS so sensitive to angle?

Answer: Many different specular reflections and multiple bounces

are interfering to form the total RCS
N 2
X √ 
−j2k·Ri 
σtot =  σi e


 
i=1

Go back
57 / 57