Monochromatic Radiation is Always 100% Polarized

Polarization Ellipse y y Right-Hand Polarization x Propagation vy 3 Parameters Specify Ellipse e.g. a, b, Also, (need + or a, , to right or left elliptical) vx, vy,
Lec13a.3-1 1/12/01

a x E(t)

b vx z

V1

Polarization of Narrowband Radiation Let E( t ) = xv x ( t ) cos[t + ( t )] + yv y ( t ) cos[t + ( t ) + ( t )]

vx(t) and vy(t) are slowly varying and random; vx, vy, and may be non-zero

Stokes Parameters
I So

[ v (t) +
2 x 2 x

v y (t)

Q So

[ v (t) v (t) ]
2 y

2 o

[W m ] total power

-2

2 o

x-ness

U S2

2 v x ( t ) v y ( t ) cos ( t ) 2 o
2 v x ( t ) v y ( t ) sin ( t ) 2 o

45-ness

V S3
Lec13a.3-2 1/12/01

circularity
V2

100% Polarized Narrowband Waves Let E( t ) = xv x ( t ) cos[t + ( t )] + yv y ( t ) cos[t + ( t ) + ( t )]

vx(t) and vy(t) are slowly varying and random; vx, vy, and may be non-zero

( t ) = o and v x v y ( t ) = constant fixed ellipse, variable size

y E(t) x

2 2 2 Also : So = S1 + S2 + S3 2

Therefore, any 3 Stokes parameters specify polarization

Lec13a.3-3 1/12/01

V3

Partially Polarized Narrowband Radiation

Stokes Parameters I So Q S1 [v x (t) + v y (t)] 2o
2 [v x (t) 2 v y (t)] 2 2

[W m-2] total power x-ness 45-ness circularity

2o U S2 2vx (t) vy(t) cos (t) 2o V S3 2vx (t) vy(t) sin (t) 2o
A+B A B

Note: For 2 uncorrelated waves superimposed (A+B), we have Si = Si + Si where i = 0, 1, 2, 3 For 0% polarization, Stokes: So; S1 = S2 = S3 = 0 Therefore, for partially polarized wave: [So, S1, S2, S3] = [Su, 0, 0, 0] + [So Su, S1, S2, S3]
2 2 2

Lec13a.3-4 1/12/01

where (So Su)2 = S 1 + S2 + S 3 So Su Define percentage polarization = 100% So = m, 0 m 1

V4

Coherency Matrix J
E x E x 1 J= o E E y x E x E where E(t) = xR E ( t )e jt + yR E ( t )e jt y x y e e E y E where E x ( t ), E y ( t ) vary slowly y

e.g.

X-polarization

1 0 Jx = 2So 0 0 1 j JRC = So j 1 1 0 Ju = So 0 0
V5

RCP (right-circular)

Unpolarized
Lec13a.3-5 1/12/01

Finding Orthogonal Polarization JRC

e.g. X-polarization RCP (right-circular) Unpolarized
Note : Jx + Jy = 2Ju

1 0 Jx = 2So 0 0 1 j JRC = So j 1 1 0 Ju = So 0 0 JRC + JLC = 2Ju

JA JB , and If then JA + JB = 2Ju Tr JA = Tr JB Therefore, we can find orthogonal polarization JB = 2Jn JA

Lec13a.3-6 1/12/01

V6

Polarized Antennas
Far Fields

Define

Gij (, ) G(, )

A ij (, ) A(, )

Ei E j E x E + E y E x y

e.g. {i, j} = {x, y}, {r, }, {a, b} (b a )

A xx A= A yx

t 1 A xy for incident ; claim Prec = 2 Tr A Jinc [W ] A yy plane wave [m2] [Wm2]

Lec13a.3-7 1/12/01

W1

Polarized Antennas
A xx A= A yx
t 1 A xy for incident ; claim Prec = 2 Tr A Jinc [W ] A yy plane wave [m2] [Wm2]

1 So Prec = [A11J11 + A12J12 + A 21J21 + A 22J22 ] 2

for incident uniform plane wave J on antenna A

1 A (, )Jt (, ) d For s 0 : Prec = Tr 2 4
Lec13a.3-8 1/12/01

W2

To Measure Polarization
Measure 4 powers; use 4 antennas e.g.
A11a Ma M 1 A11b b = Mc 2 A11c A M d 11d

A12a A12b

A 21a

A 22a J11 J12 J21 A11d J22

1 = 2A 1M J is estimate M = A J , so J 2 Is A singular?

Lec13a.3-9 1/12/01

W3

To Measure Polarization
1 = 2A 1M J is estimate M = A J , so J 2 Is A singular? For x, y, RC, LC POL:
1 0 0 0 0 0 A= 1 j j 1 j j

)
0 1 1 1

For x, 45, RC, LC: 1 0 0 1 1 1 A= 1 j j 1 j j

0 1 1 1

Can not distinguish

det A = 0

vs

det A 0 " ok"

Lec13a.3-10 1/12/01

W4

Example of a Polarimeter
Right Circular Diplexer 201.5 MHz 231.5 MHz Local Oscillator 30 MHz 25.0 MHz Local Oscillator [Cohen, Proc. IRE, 1, 1958] Left Circular 201.5 MHz

30 MHz 25.001 MHz Local Oscillator

5 MHz ( )2 1 KHz

4.999 MHz Phase Comparator

dt
Lec13a.3-11 1/12/01

2 r 2 r 4 measurements 4 Stokes parameters

W5

Antenna Phase Errors

Systematic antenna phase errors:
phase front 1) 2) 3) poor design and fabrication gravity, wind, thermal (gravity and thermal limits near 1 arc minute) feed offset

Random antenna phase errors:

1) 2) 3) machine tolerances, surface roughness adjustment errors feed offset

Lec13a.3-12 1/12/01

X1

Examples of Antenna Phase Errors

Random antenna phase errors:
1) 2) 3) matching tolerances, surface roughness adjustment errors feed offset

300-ft parabolic reflector antenna at NRAO, Greenbank, West Virginia 1)

2)

systematic sag fix backup; footprints on mesh

steamrolled mesh long waves ~ new panels: B > 0.5 1 arc minute

3)

Lec13a.3-13 1/12/01

X2

Types of optical and radio propagation phase errors

Systematic:
h T(h) (h) velocity of light c c < c c ionosphere c

Earth

Earth

Random phase: + amplitude?

~ RMS < 2 , weak fluctuations RMS >> 2 , strong fluctuations

vs Thin screen (constant amplitude)

pathlength = L n interference and nulls Thick screen

Lec13a.3-14 1/12/01

X3

Effect of Phase Variation on Directivity

x aperture x

For x-polarization: ~ E x ( x, y ) E x ( x , y )
z

y
y

RE x ( )

E x ()

D(), G()

D( f , , ) = (1 + cos )

x d y A RE x ()e ] 2 E x ( x, y ) dxdy A

2 ( ) d

Lec13a.3-15 1/12/01

X4

Effect of Phase Variation on Directivity

D( f , , ) = (1 + cos )

x d y A RE x ()e ] 2 E x ( x, y ) dxdy A

2 ( ) d

E{D( f , , )} =

(1 + cos )
2

A E x ( x, y )

dxdy

E RE x ( ) e
A

2 ( ) d

x d y

E x r E x (r )dr

() A j (r ) Eo (r ) e

Therefore E RE x ( ) = REo ( ) E e j (r ) j (r )

Spatial stationarity : E e j (r ) j (r ) = E e j(o ) j (r )

Lec13a.3-16 1/12/01

} {

}
X5

Definition of Characteristic Function

It is the Fourier transform of probability distribution p(x) (also called the moment-generating function)
E e j x

[ ]

p( x )e jx dx = F.T.[p( x )]

= ( ; x) = characteristic function of p(x) One use of the Fourier transform of p(x) is when we seek p(x1 + x 2 + ... + xn ) =
n p(x1) p(x 2 ) ... p(xn ) = F.T. F.T.[p(xi )] i=1
Lec13a.3-17 1/12/01

X6

Computation of E{ RE x () }
Thus (1, 2 ; (0), ( )) = E e j1(0 )+ j2 ( ) Recall : If 1, 2 are JGRV , then

}
1 1 2 2 2 2

( 1, 2 , 1, 2 )

1 [12 ] =e 2

1 1 2 1

Here, 1 = (0) , 2 = ( )

Therefore : E e j(0)- j ( ) = (1 = 1, 2 = 1; (0), ( ))

Lec13a.3-18 1/12/01

X7

Computation of E{ RE x () }
(1, 2 , 1, 2 )
1 [12 ] =e 2

Here, 1 = (0) , 2 = ( )

1 1 2 1

1 1 2 2 2 2

Therefore : E e j(0)- j ( ) = (1 = 1, 2 = 1; (0), ( )) (0) (0) = (0) (0) ( ) = ( ) Since: ( ) ( ) = (0) by stationarity

(1 = 1, 2 = 1; (0), ())

1 [11] =e 2

= e( ) (0 )
Lec13a.3-19 1/12/01

(0) ( ) 1 ( ) (0) 1

Therefore E RE x ( ) = REo ( ) e( ) (0 )

X8

Computation of Expected Directivity

2 -j i (1 + cos )2 ( ) (0) E {D( f, , )} = e REo ( ) e d x d y 2 2 E ( r ) da A A
= (0 )

()

()
0 L
1

correlation length L of phase irregularities

(0 ) 0

()

(0 )
2

( ) ( 0 )

1 e

= B( )

=
0

0
Lec13a.3-20 1/12/01

X9

Solution to Expected Directivity

1 e

= B( )

e( ) (0 ) =
0

+
L 0

2 j (1 + cos )2 2 + B( ) REo( ) e d x d y e E{D( f , , )} = 2 2 E(r ) da A A

E{D( f , , )} = e

Do ( f , , ) + B() Do ( f , , )
sidelobe increase

B()

gain degradation
Lec13a.3-21 1/12/01

0 /L

X10

Examples of Random Antenna Surface

Let b = RMS surface tolerance of reflector antenna On-axis gain of random antenna G = Goe o
2

= Goe

(2b2 )2

= Goe

(b 4 )2

If b = 4 Go e 1 b = 16 Go 0.54 b = 32 Go 0.9
log G Any aperture antenna, fixed illumination -2 ~minimum useful wavelength log (power shifts to sidelobes)

new

log G = 2 log + log 4 A e

Lec13a.3-22 1/12/01

X11