L

NAVAL]-.
MONTEB.

,-;ool

A 93943

NAVAL POSTGRADUATE SCHOOL

Monterey, California

THESIS
ANGULAR TRACKING ERROR IN A PHASE
COMPARISON MONOPULSE TRACKING RADAR,
A CRITICAL REVIEW AND EXTENSION OF THE
PHASE FRONT DISTORTION APPROACH
by

Sopon Bumroongpol

December 1984

Thesis Advisor:

H.

M.

Lee

Approved for public release; distribution unlimited

T221996

SECURITY CLASSIFICATION OF THIS PAGE (Whan Data

Entered)

READ INSTRUCTIONS
BEFORE COMPLETING FORM

REPORT DOCUMENTATION PAGE

1.

4.

REPORT NUMBER

TITLE (and

2.

GOVT ACCESSION NO

Subtitle)

Angular Tracking Error in a Phase Compari

son Monopulse Tracking Radar, A Critical
Review and Extension of the Phase Front
Distortion Approach
7.

AuTHORfsj

3.

RECIPIENT'S CATALOG NUMBER

5.

TYPE OF REPORT

PERIOD COVERED

Master's Thesis;
December 198 4
6.

PERFORMING ORG. REPORT NUMBER

8.

CONTRACT OR GRANT

NUMBERfa.)

Sopon Bumroongpol
9.

PERFORMING ORGANIZATION NAME AND ADDRESS

10.

PROGRAM ELEMENT, PROJECT, TASK

AREA 4 WORK UNIT NUMBERS

Naval Postgraduate School

Monterey, California 93943
II.

CONTROLLING OFFICE NAME AND ADDRESS

12.

December

Naval Postgraduate School

Monterey, California 93943
U. MONITORING

AGENCY NAME

ft

ADDRESSf//

REPORT DATE

13.

19 8

NUMBER OF PAGES

41

different from Controlling Olttce)

15.

SECURITY CLASS,

(ol thla report)

UNCLASSIFIED
15a.

16.

DISTRIBUTION STATEMENT

DECLASSIFICATION/ DOWNGRADING

SCHEDULE

(ol thla Report)

Approved for public release; distribution unlimited

17.

DISTRIBUTION STATEMENT

18.

SUPPLEMENTARY NOTES

19.

KEY WORDS

(ol the abatract entered In

(Continue on reverse aide

Block

30, If different from Report)

neceaaary and Identity by block number)

If

Phase Front Distortion of Monopulse Tracking Radar

20.

ABSTRACT

(Continue on reverse aide

If

neceaaary and Identity by block number)

This thesis studies the inherent angular errors of a phase

comparison monopulse system used for tracking a complex target.
The phase compensation equation is utilized in justifying
Howard's hypothesis on the relationship between the phase front
distortion of the scattered wave from a complex target and
angular tracking errors, in extending this hypothesis to closer
ranges to the target, and in determining the limitations of this
DD

FORM
I

JAN

73

1473

EDITION OF
S

NOV

65

IS

N 0102- LF- 014- 6601

OBSOLETE.
SECURITY CLASSIFICATION OF THIS PAGE (When Data

Bniarad)

SECURITY CLASSIFICATION OF THIS PAGE (Whmi Dttm BntmwO)

Through the phase compensation equation, global

hypothesis.
errors are demonstrated. A local error bound is also
determined for the tracking of a two element target. These
new results are not predicted by Howard's hypothesis.

N 0102-

LF- 014-

6601

SECURITY CLASSIFICATION OF THIS PAGE(TWin Dmtm

Enlttmd)

Approved for public release;

distribution is unlimited.

Angular Tracking Error in a Phase Comparison Monopulse Tracking

Radar, A Critical Review and Extension of
the Phase Front Distortion Approach
by

Sopon Bumrcongpol
Lieutenant Commander, Royal Thai Navy, Thailand
Royal Thai Naval Academy, Thailand, 1970

Submitted in partial fulfillment of the

requirements for the degree of
MASTER OF SCIENCE IN ELECTRICAL ENGINEERING

from the

NAVAL POSTGRAD0ATE SCHOOL

December 1984

-SS3S
S&S8."**"*
ABSTRACT
studies the

This thesis

comparison

phase

complex target.
in

phase front

a complex

extending this

system

used

target

hypothesis

on

distortion of
and

tracking

relationship
the scattered wave
the

angular tracking

hypothesis to closer

ranges to

determining the limitations of

Through the phase compensation equation,
and

for

errors of

The phase compensation equation is utilized

Howard's

justifying

between the
from

mcnopulse

inherent angular

in

this

errors,

in

the target,

hypothesis.

global errors are

demonstrated. A Local error bound is also determined for the

tracking of

two element target.

predicted by Howard's hypothesis.

These new results are not

DUDLEY KNOX LIBRARY

NAVAL POSTGRADUATE SCHOOL
MONTEREY, CALIFORNIA 93943

TABLE OF CCNTEHTS

I.

II-

INTRODUCTION

A.

RADAR FOR TRACKING

B.

NOISE IN TRACKING RAEARS

12

C.

HOWARD'S ASSUMPTION

15

D.

PURPOSE OF THE THESIS

16

THEORETICAL BACKGROUND AND RESULTS FROM

SCATTERING THEORY

III.

17

A.

RADAR CROSS SECTION

17

B.

THE ELECTROMAGNETIC SCATTERING

13

PHASE FRONT DISTORTION AND THE ANGULAR

TRACKING ERROR OF

PHASE COMPARISON MONOPULSE

TRACKING SYSTEM
A.

23

RELATION BETWEEN ANGLE ERROR AND PHASE

FRONT

IV.

V.

23

B.

PHASE FRONT OF THE SCATTERED FIELD FROM

C.

TARGET AND THE ANGULAR TRACKING ERROR

PHASE COMPENSATION EQUATION AND THE
ANGULAR TRACKING ERRCRS

AZIMUTHAL ANGULAR ERROR EOR

24

30

TWO-ELEMENT

TARGET

36

CONCLUSION AND REMARKS

39

LIST OF REFERENCES

40

INITIAL DISTRIBUTION LIST

41

LIST OF FIGURES

1.

Conical-Scan Tracking

1.2

Monopulse Antenna Patterns and Error Signal.

1.3

Kavefront Phase Relationships in Phase Comp.

...

11

Monopulse Radar

13

2.1

Bistatic Case

19

2.2

Monostatic Case

22

3.1

Coordinates of a Simplified Phase Comparison

Tracking System

26

3.2

Phase Fronts and

Phase Comparison Tracking

System

29

3.3

Phase Compensation

32

3.4

Local and Global Angular Tracking Errors

34

3.5

Max. Local Angular Error of a Phase Comparison

System
4.1

Angular Tracking Error across a Singularity of

Phase Distortion
4.2

35

37

Maximal Local Error , First and Second Global

Error

38

I.

INTRODUCTION

RADAR FOR TRACKING

A.

tracking-radar

system measures

the coordinates

of a

target and provides data which nay be used to determine the

target path and to predict its future position.
All or only
part of

the available radar

data such as

range

elevation

angle, azimuth angle and doppler frequency shift may be used

in

predicting future target position; that is, a radar might

track in range, in angle,

tion.

Almost any

radar can be considered

provided its output information

it is

the method

that distinguishes

or with any combina-

in dcppler,

by which
what is

radar from any other radar.

tracking radar

is processed properly.

angle tracking

But

is accomplished

normally considered

tracking

It is also necessary to distin-

guish between a continuous tracking radar and a track-whilescan

(TWS)

data

on a

radar.

The former supplied continuous tracking

particular target,

supplies sampled data on one

while the

track-while-scan

or more targets.

In general,

radar employ
different types of equipment and serve different purposes.
For TWS,
the track of a target can be determined with a
surveillance radar
coordinates of the target as
from the
measured from scan to scan.
The quality of such a track
the locating
will depend on the time between observations,

the

continuous tracking

radar

and

the TWS

accuracy of each observation,

and
the number of extraneous
targets that might be present in the vicinity of the tracked
target.
All of these, including prediction and estimation,
are usually accomplished by using

For continuous tracking

radars,

specific computer.
a

popular type

is the

conical scan radar, as shown in figure 1.1 The angle between

jrget crs~.

Figure

1.1

Conical-Scan Trackin;

the

axis of

antenna

rotation (which

reflector)

called the squint

The echo

usually the

is

the axis

and

signal will be modulated

the rotation

freguency of the

the

antenna beam is
target at position A.

of the

Consider

angle.

axis of

at a frequency

equal to

The amplitude

beam.

of the

echo-signal modulation, called the angle error signal,

will
depend upon the shape of the antenna pattern,
the squint
angle,
and the angle between the target line of sight and
the rotation axis.

the angle

The phase

between the

of the modulation depends on

the rotation

target and

axis.

conical scan modulation is extracted from the

and applied to a
servo-control system which
positions the antenna

Note

on the target.

The

echo signal
continually

that two servos

are required

because

dimensional.

Both the rectangular and polar tracking coor-

the

tracking

problem

is

two-

dinates may be used.

When the antenna is on target, and if
the target is located at B of
figure 1.1,
then the line of
sight to the target and the rotation axis coincide,

conical-scan amplitude modulation is zero.

tracking radar requires information from

in

to extract

order

interval during
the train

the angle-error

which
echo

of

number of pulses
the time

conical-scan measurement

is made,

pulses

caused,

section,

cross

degraded.

signal.

contain

must

amplitude-

no

the modulation produced by

If the echo pulse-train contains additional modu-

lation components

target

The conical-scan

In

modulation components other than

scanning.

and the

This

can be

for

example,

by

fluctuating

the tracking accuracy might be

especially severe if the spectral

content of the fluctuation is strong at or near the conicalThe effect of

scan frequency or the sequential-lobing rate.
serious in some
be sufficiently
the fluctuating echo can

applications

to

tracking radars

severely

limit

which require many

the

accuracy

pulses to

before the error signal can be extracted.

of

those

be processed

The

because

conical-scan

the

necessary for the

obtained through a

information

all

the angular

determination of

superior to

technique is

monopulse tracking

error is

stationary for

target appears

single pulse.

Since the

pulse duration,

pulse to pulse variations of the echo signal

performance of

limit the

which

motion,

to target

due

conical-scan radar, have no effect on a monopulse radar.

The amplitude-comparison monopulse employs two identical
but slightly offset antennas

angular

in

error

antenna beams

(figure

cne coordinate.

four feeds

azimuth

error signals

antenna patterns of
(b)

The

both elevation

and

(a)

The sum

pattern

are

used

received with the difference

of the angle error.

The sum pattern

on

reception.

The

signal

pattern provides the magnitude

The sum signal provides the range meas-

urement and is also used as

of the

two

while both the sum pattern and the

is used for transmission,

difference

(c)

of the

in figure 1.2

is shown

and the difference in figure 1.2

reflector or
(A

are desired.)

figure 1.2

overlapping

two

by two adjacent feeds.

used if

may be

to provide the

a)

with a single

may be generated

with a lens antenna illuminated

cluster of

1.2

error signal.

reference to extract the sign

They are

amplified separately

and

combined in a phase-sensitive detector to produce the error

signal characteristic shown in figure 1.2 (d) ,
The useful
region is on the linear portion of the curve.
phase comparison

In a

arrival

(in

the radar)

moncpulse radar,

the angle

of

either the azimuth cr the elevation direction of

is determined by

between the signals from two

comparing the phase difference

separate antennas.

Unlike the

antennas of amplitude comparison trackers, those used in

phase comparison systems are not offset from the axis.
The
individual boresight axes of the antennas are parallel,
causing the
(far- field)
radiation to illuminate the same
volume in space.

The amplitudes of the

10

target echo signal

i9

~-9^
\b\

Figure

Monopulse Antenna Patterns and Error Signal

Left-Hand Diagrams in (a-c) are in Polar
Coordinate: Right-Hand Diagrams are in Rectangular Coordinates, (a) Overlapping Antenna Patterns; (b) Sum Pattern; (c) Difference Pattern; (d) Product (error) Signal.

11

are essentially

the same from

phases are different.

each antenna beam,

radar which operates with

tracking

but the

phase information is similar to an active interferometer and

might be called an "interferometer

radar". In figure 1.3 two

The distance
antennas are shown separated by a distance d.
target is R and is assumed large compared with the
to the

antenna separation d.

The line of sight to the target makes

an angle 6 to the perpendicular bisector of the line joining

the two antennas.

The distance from antenna

Eg = R

+ d.

to the target is

sinG/2

and the distance from antenna

RA

= R

- d.

to the target is

sine/2

The phase difference between the echo signals in the two

antennas is approximately

A0

= 21fd.sine/x

For small angles where SIN 6 = e

the phase difference

linear function of the angular error and may be used to

is a

position the antenna via

B.

servo-control loop. Ref.1J

NOISE IN TRACKING RADARS

The target-noise

or target-scintillation is due

to the

physical complexity and manuevers of the target.

Because of
the size of the
target compared to the wavelengths being
used,

the amplitudes of radar

echoes from moving targets of

practical interest such as an aircraft will fluctuate widely

and rapidly as target aspect changes. In general, there will
be rapid

fluctuation caused by pitching and yawing of the

airframe about the line of flight,
and slow variations
caused by change in the average aspect of the air frame
relative to the radar.

12

r
,/

,/
,w

w//

///
/'//
/

//
/
/

Figure 1.3

Wavefront Phase Relationships in Phase

Comparison Monopulse Radar.

13

The noise sources which contribute to tracking noise may

lie

separated into four components, i.e.,

1.

Servo noise

is the

servome-

the tracking

error of

chanism which results from backlash and compliance in

and structures of the mount.

the gears, shafts,

The

magnitude of this noise is essentially independent of

and will thus be independent of range.

the target
2.

Eeceiver noise is the effect on the tracking accuracy

thermal noise generated

radar due to

of the

receiver and any spurious hum

in the

which may be picked up

by the circuitry.
3.

(angle scintillation

Angle noise

tracking error

tions in

introduced into

the apparent angle

or glint)

the radar

is the

by varia-

of arrival of

the echo

This effect is

from a complex target of finite size.

caused by variations in the phase front of the radiation

from

multiple- point target

changes its aspect.

The magnitude

the

as

target

of angle noise is

inversely proportional to the range of the target.

4.

Amplitude

noise

effect on the

(amplitude scintillation)

radar accuracy of the

the amplitude of

is

the

fluctuations in

the signal returned by

the target.

These fluctuations are caused by any change in aspect

of the target and must

be taken to include

propeller

rotation and skin vibration.

The first two (No. 1,2)
of these components originate in
the radar itself.
The second two components
are
(No. 3,4)

called target noise.

Angle noise is the

close range.
[Eef.2]
Superiority
scan

of monopulse

tracking radar

appears stationary
pulse to

pulse.

most important

tracking

is based

over a pulse

on

error at
radar over

that

the fact

duration but

Monopulse tracking radar

14

medium to

conical
a

target

varying from

still encounters

angle errors for

each pulse.

This may

result in increased

pulse to pulse variations in

causes

predicted target locations and

the servo
system.
This target
noise

problems in

should be dealt

bandwidth

with through

which determines

proper choice

the

of the

sensitivity

of the

servoservo

system.
C.

HOWARD'S ASSUMPTION
The "slope" of the phase frcnt of the echo signal from a

complex target of finite size is claimed by Howard [Ref.U]

to be identical
to the angular errors caused in a tracking
radar by angular scintillation or target angle noise.
No

definition on "slope" was given,

the statement was provided.

however,

and no proof of

Since phase comparison tracking

systems are essentially phase front positioning devices, it

can be argued
that the target angle noise is determined by
the echo

from

signal due

being

to the distortion

sphere.

Howard

and

phase front

of its

Dunn

[Ref

5]

further

proposed, by erroneously assuming that an antenna is a power

measuring device instead of a field strength probe, that

deviations in the
the Poynting

from a

direction of the echo signal

vector.

complex target

They argued
can be

is given by

angular error
apparent
that the

that the

so large

source falls many target spans away from the actual target
location.
Howard further claimed without proof that the

relationship between phase frcnt distortion and angular

tracking error also applies to the amplitude comparison
These claims
tracking of a target and to the search radars.
have been proved to be false. It has been found that angular

tracking error

studies based

en

Howard's

lead to exaggerated error estimates

15

assumption could

D.

PURPOSE OF THE THESIS

In this thesis a detailed study is carried out to demon-

strate the true relationship between

tered

field

and the

angular

tracking

comparison monopulse tracking radar.

16

the phase of the scat-

error in

phase

II. THEORETICAL BACKGROUND AND RESULTS FROM SCATTERING

THEOEY
A.

RADAR CROSS SECTION

For most

between a

practical situations there is

the target

radar and

relative motion

being tracked.

The radar

"sees" a dynamic cross section. Considerable effort has been

expended in

statistics of

investigating the

dynamic cross

particularly in the case of aircraft.

It can be
shown that for a target consisting essentially of a large
sections,

number of point
radar

scatterers in random relative

cross section

density.

For a

will have

relatively

motion,

an exponential

complex

target,

the

probability
such

as

an

experimental results agree

well with this conclusion (see detail in Ref.4) .
That is we

aircraft at

high frequencies,

can use the static cross

secticn in our calculation instead

of using dynamic cross section.

However,
this

for most

thesis the

of the calculations and

target may

be assumed

to be

analysis of
stationary,

That is,
the
because a monopulse tracking radar is used.
radar cross section is a static cross section within a pulse
duration.
There exists a wide variety of analytical tech-

niques for

computing radar cross

sections.

Some

of these

although approximations are usually reguired to obtain numerical results even

A typical
from what was originally an exact formulation.
lead,

in principle,

example of

to exact solutions,

this type of technique

is the formulation

of a

scattering problem as a boundary-value problem with the

resultant formally exact solution appearing as an infinite
series.
To obtain numerical results, the sum of this series
must usually be approximated in some fashion.

17

Other techniques involve various simplifying assumptions

to arrive at the basic formulation so that even if the math-

only an
ematical analysis can be carried through exactly,
approximate solution is obtained.
A typical example of this

type is the physical optics technique for which it is gener-

integration can be
performed exactly, only an approximate solution is obtained.
The definition of radar cross section is as follows:

if the

even

true that

ally

= lim 4irk*|E:/Ei
s

required

fc-+oo

field intensity scattered from

where E_ is the electric

the target and observed at a distance

the field

Ej is

incident field is
the limit of

intensity incident

from the target, and

upon the

assumed to be a uniform
as R>oo is

target.

The

plane wave,

and

taken to ensure that

the true

far-field value of 4 is measured. In practice, it is usually

t
considered sufficient if R is greater than about 2D/A, where
the largest dimension of

D is

the target in

transverse to the direction from which

is the radar wavelength.

for some purposes

recommended

is observed and X

It should be noted, however,

I

the 2D/A criterion is

far-field

Es

any direction

distance

may be

that

inadequate and the

several

times

as

great. [Ref .5]

Note that, the radar cross section also depends upon the
frequency, shape and composition of the scattering object.
B.

THE ELECTROMAGNETIC SCATTERING

In general,

for bistatic radar the scenario is shown in

Figure 2.1
where
is the

direction of

the propagation

vector of

incident plane wave.

Ej

Hj

are the fields of the incident wave.

18

the

Figure 2.1

Bistatic

19

Case.

Scatter
(Radar target)

the distance

is

Tj

transmitter to

from the

the scat-

terer.
s is the radial direction unit vector

from the scatterer

to the receiver
Es

are the far fields of the scattered wave.

Hs,

is the distance from the scatterer to the receiver.

The scatterer is

currents

Js

matallic

which can

object with induced surface

solved for

be

example,

by

the

magnetic field integral equation (MFIE)

Js

= 2nxHj (r) + n x

(r)

((

(r)X V G. ds

G =

{ex p. (-jkR)}/2llR

R =

|f-r|

and

k = 2TT/A.

Scattered

for f and r both points on surface

fields

can

be

obtained

from

surface

the

current. If
=

f(s)

\\

J s (r).{exp-

(jksrJJ.ds

then

E
H

(r

(f s )

,-r

A.

{-jkyaiir s }.{exp. (-jkr s )}. {f-(f.s)s}

{1Ao)s E s

(r

linearity results in J$ being proportional to

Ej

and a

"complex scattering amplitude function" can be defined as:

S(i,s)

{-jkiLo/VffEj }. f- (f .s)s}

then
E

(r s )

= E; S(i

s). [exp. {-jkr s )}/r s

A A
where S(i,s) is polarization dependent.

More generaly,

polarization effects
the "scattering matrix" as:

20

can be included in

jkrs
e
J

or,

r.

>v

Si

= (i,s) .E,-. {exp. (-jkr s

The subscripts

and

}/r s

represent polarizations for both

the incident and the scattered fields.

For

monostatic scattering,

the scenario

is shown

in

figure 2.2 Usually the same antenna is used for transmitting

and receiving in a radar.
For the monostatic case, r
{

21

= r =
s

r.

Trans

M
;

r =r =r

Scatter
(Radar target)

?T ~

Figure 2.2

Monostatic Case.

2 2

III. PHASE FRONT DISTORTION AN! THE ANGULAR TRACKING ERROR

OF A PHASE COMPARISON HQNCPULSE TRACKING SYSTEM

A.

RELATION BETWEEN ANGLE ERROR AND PHASE FRONT

The phase compensation equation is found to describe the

mechanism with which the direction of the tracking axis of

phase

comparison

tracking system

front distortion

Howard's phase

determined

is

technique is

[Ref.7].

shown to

be

applicable to such a system and is an approximation to the

phase compensation equation. In what follows, global angular
tracking errors are

demonstrated and an upper

error is given.

local tracking

The antenna

the essential parameter in

recognized to be

bound to the
separation is
the determina-

tion of both kinds of errors.

Using

the

fields from

far

field approximation

two-element target,

Howard

the

of

scattered

[Ref.U] reasoned

tracking radar is determined by

phase front of the scattered field

that the angular error of a

the distortion

from a target.

of

the

The significance of this

the angle noise of a tracking

from the fact that it relates

radar
target.

directly to
Hence the

tracking radar need

observation stems

scattering characteristics of the

parameter of the antenna system in a
studies in angle
not be considered and

the

noises are greatly simplified.

A

recent investigation

[Ref.6]

revealed

that for

an

amplitude comparison system, the tracking radar can be right

technique
Howard's phase front distortion
on
target yet
This brings about the
predicts an infinite error Ref. 1] .
noise predictions based on
question of whether angle
Howard's assumption have been exaggerated for some system
In this thesis,

while inapplicable to others.

23

the relation

the phase

between

front

angular tracking error of

field and

scattered

of the

the

phase comparison tracking system

is examined carefully. The validity and restrictions of this

established for

are

relationship

tracking

local

errors.

Global tracking errors are found to exist when the system is

track while no error signal is

wrong

on a

companion work

generated.

error of

the angular tracking

In a

an amplitude

comparison system are studied Ref.8,9J.

B-

PHASE FRONT OF THE SCATTERED FIELD FROM

TARGET AND THE

ANGOLAR TRACKING ERROR.

field,
them.
at

components

the three

From

the scattered

of

electric

an antenna will pick up only a linear combination of

ccncern for an antenna located

The received field of

(r,6,0)

can be written as

F(r,6,0) {exp[iD(r,6,0)]} {exp(ikr)}/r

where k = 21\/\ is
and F (r,6,0)

system

the target.

at

is the wavelength

the spherical coordinate

The origin of

0.

is located

the wave number,

(1)

the

In

limit that

approaches infinity,
F(r,6,0)

>f (8,0)

D(r,6,0)

>j (6,0)

Assume the

(2)

incident field

zero phase at the origin. Then

of

the target and 0(6,0)

Note that
f (6,0) {exp|i<$(6,0)

is

the

far

|}

to be
4*flf

(9,0)

of unit

strength and

is the cross section

is the phase shift of the target.

{exp (ikr) }/r

field expression

for

equation

(1)

and

F(r,6,0) and D(r,6,0) are slowly varying functions in r. For

large r the phase of the scattered wave is

24

D(r,e,0)

$(r,6,0) = kr

and

constant.

(3)

is a surface

phase front

on which

$(r,e,0)

is

is the phase distortion which keeps

Thus D(r,6,0)

the phase front from being a sphere centered at the target.

consider only the case when there is no

For simplicity,

error in the elevation angle.

Only the plane containing the

center of the target (i.e.

origin of the spherical coordithe antennas
nate)
and the pair of effective centers of

measuring azimuth

angle has

tc be

considered.

It

can be

modeled as a system of only two antennas.

the

As shown in figure 3.1,

angle

from the negative

radial direction -r T to the tracking axis pointing direction

t (the caps above the characters denote unit vectors) is the
angular tracking error. The antennas A and B, centered at (rA
read the phases $ A and $ B
respectively,
#*0a) an ^ r B'?'^8^
(

according to equation
#a = kr A

(3)

O(r A ,j,0A

Note that this statement is based on the assumption that

extended

angle

the

the

by

target

is

negligibly

small

compared to the beam width of each of the antennas.

An error signal is generated in the tracking system if

The tracking axis is conditioned to point

to a direction so as to null the error signal. Therefore the
tracking error is determined by the equation :
differs from

k(r A -r B

<j>

D(r 3 ,,0 B

The quantities rA
of r T

T,

d,

D(rA ,,0 A
$A

<J

e+

terms
B can be expressed in
:

(5)

r6 = r T - r T d.sin

d/U

(4)

through the following equations

rA = r T + r T d.sin
1

rp

+d/4

(6)

25

Target

Figure 3.1

Coordinates of

Simplified Phase Comparison

Tracking System and Its Target.

2 6

rA .sin (0e -0 T +0 A )

= rT

r g .sin(0 e -0 T +0 B )

= r T .sin

Where
-rf
#

# T

midpoint

is the

Together with equations

T

d/2

(7)

- d/2

(3)

e +

is the separation

term of r

.sin

between the antennas

between the

to

(5)

(8)

d.sin

equations

(7)

and

(8)

=
=
=

e in

to the lowest order in

d,

lead to

(6)

lead to

()

A -<P B )cos

i<t>

(1/rA
d.

cos

+ -

(0e

-Vrgjr^sin

cos

T)

e
+

(1/rA -1/rg

d/2

order,

and assuming that

the region between

D(r,a,0)

the antennas,

signal nulling condition as given by equation

k(rA -r s ) = kd.sin

(4)

varies

the error

reduces to

= D(r 8/ ^,0
B
=

gives

(4)

- 03 = d.cos 0g

To the same

smoothly over

and B.

sin (0e-0/0 A ) -sin ((^-0/0 B ) = 2. sin

or

equation

(r-

and d

With the assumption that r T^>

the ratio d/r T equations (5) and
rA - r B =

antennas

and

D(rA ,|,0A

^- rA)^ D ^T'^r) +

(0B

-0 A )^D(r T

,l,0 T )

=-dCsin0c.^: D(r
/
#
| / T )}
| # rTT )+ccs0e(1/r)^D(r
}r *r*'
^0

'

(9)

or

tan0 =-{(1/kr)^-D(r T ,,0 T )}/{1+ (1/k) D

e

27

(r

T ,*,0T )}

(10)

Equation

Howard's

to

reduces

(10)

expression for

the limit when r T approaches infinity so

is replaced with <$(![, r ).

complex target in

that D(r T ,,0 T )

The above results

error

elevation

be extended

can

to form

are utilized

antennas A,B,C,D

system. One pair of antennas, A at

,0 fe )

(rA ,6 A

coplanar

four

tracking

,0 A

and

when

antenna
at

(r

B ,98

measure the azimuth angle of the target location rela-

The other pair of antennas, C at

tive to the tracker.

,0C

that

Assume

present.

is

the case

to

c ,9 c

(r

D/ 6 p# 0p), measure the elevation angle.

Assume that the antennas C and D are arranged along a

and

at

(r

and mutually bisecting with

line perpendicular to

antennas

and B r

and the

located at the common

midpoint

connecting the
center

T is

pairs of

both

antennas.

azimuth angle

Assume

further

is determined by phase

the line

antenna system
(r T# 9 T

that the

D(r B ,e B ,0 B )

target

- D(r ,9 ,0 A )
A
A

which is just an extension cf equation

^2 plane to

of

comparison technique.

The azimuth error signal nulling condition leads to

k(rA - r 5

,0 T

(11)
(4)

from the

plane containing the target and the antennas

This plane and the

direction of the tracking axis

cannot be determined until the mechanism for determining the

and B.

elevation angle is specified.

is also used

phase comparison technique

If

to determine the target

elevation angle,

elevation error signal nulling condition leads to

k(rc - rp

Dfr^e^p)

Note that equations

dently and though
between

and

(11)

B and

D(r ,ec ,0 c )
c

and
=

Equations

(12)

(11)

mine the pointing direction of

28

and

the

(12)

are obtained indepen-

there is no connection
(12)

together deter-

the tracking axis.

X
Target
Figure 3.2

Phase Fronts and

29

Phase Comparison Tracking System

3.2 shows

Figure

shaded

surface

plane,

which

1 ,

the antenna system center

contains the antennas A,B,C,D and

The tracking axis t is normal to surface 1; a phase front

another
surface 2 passes through the antenna pairs A and B;

T.

phase front surface

D.

Both surface

surface

If the

passes through the antenna pairs C and

and surface 3 are

when one

1,

system.

looks into the

presence of

which points

front of

blocked regions of surface

the mean value

are smooth,

partically blocked by

the antenna

and surface

theorem of calculus assures the

normal on each surface in the blocked region

in the direction cf t.
Since D(r,6,0)
is a

slowly varying function of r for large

r,

together with the

varies slowly over the solid angle

extended by the blocked region,
the paricular
normals on
surface 2 and surface 3 in the direction of t can be approximated by the normal at T of the phase front passing through

assumption that D(r,G,0)

(not shown in figure 3.2)

with this approximation,

t = -V$(r T ,e ,0 T )/lV$(r ,e T ,0 T
T
T
=

-{?+(1/k)VD(r T ,0 T ,0 T )}/|r+(1/k)VD(r T ,e T ,0 T

For a

phase comparison tracking system,

reduces to equation
elevation angle.
If amplitude
and equation

PHASE

(13)

(10)

when there

is no

comparison technique is used

the elevation angle,

C.

e^uatior^ (12)

(13)

equation

error in

(13)

the

to determine

will have to be replaced

will not apply.

COMPENSATION EQUATION

AND

THE ANGULAR

TRACKING

ERRORS
If the problem

of determining the target

elevation and

azimuth angles are set aside, equations (4), (11),

and (12)
carry the same message concerning the manner the tracking
axis is determined by a phase comparison
tracking system.

30

Take equation

example.

as and

(11)

phase distortion

The

introduces different amounts of phase shift at the

two
antennas.
If
the antennas are aligned
so that the
tracking axis is pointing toward the target direction -r T as
D(r,e,$)

shown in

figure 3.3

will not

be equal

phases read by

the

(a),

the antennas

signal is generated.
The
to move its axis off
this direction to

and an error

tracking system has

create a difference in the distances traveled by the signals

arriving at antennas A and B.
This difference in r A and r B
introduces into

the signal

chase difference

- kr s

of kr A

which compensates the difference in phase distortions at the

antennas as shown in figure 3.3

(b)

Thus equation

will

(11)

be referred to as the phase compensation equation.

Let q be the angle from

the target direction -r T to t
which is in the direction of
the projection of the tracking

axis t on the plane containing

the target

and the

(origin)

Since k(r A - r B )
can always be approximated by kd.sin $ e as long as r>?d r
and the antenna system
can only measure phase difference to
within 21T,
the right-

antennas

and B.

hand side of equation

and

1T

That is, for

kd.sin
without

-IT

rd,

D(r ,e ,0g )
B
3

$e =

any

will always be a number between

(11)

assumption

D(rA ,6 A ,0 A )}

of the

modulo

smoothness

of

2Tf

(14)

the

phase

distortion over a region containing both antennas. For d>*/2

equation
|$ e

Note

(14)

always has a solution and

= sin

j>

is bounded

(V2d)

that if

d is

(15)

increased, $ e
|

can

be limited

to

within any desired value.

For d>Ar

equation

(11)

may exceed the right-hand-side of

k(rA -r B )
by 2lf.
If the tracking axis is pointing along

this global error direction,

there

generated in the tracking systen.

31

will be no error signal

Figure 3.4

(a)

shows the

Target

D(r.,e.,(f
A)
A' A ,r A

D(r R ,e R

(a)

Ob)

D(r R ,e p

,/.

)-D(r A ,6 A ,K)
A' A' ^A
1

kCr.-r^

Figure 3.3 Phase Compensation. (a)Error Signal is Generated when the Azimuth Plane Tracking Axis Points to a Target with uneven Phase Distortions
at Antennas (b)Uneven Phase Distortions are
Compensated by offsetting the Azimuth Plane
Tracking Axis

32

situation when antennas

The angular error

and

are on the same phase front.

in this situation will be

called

local

error and is given by equation (11) or (14). Figure 3.4 (b)

shows an example of the global error when the antennas are
on different phase fronts with
Errol:

signals will

<j)

<f)&

2Tf.

be generated by the

tracking system

to keep flhe antennas staying

on separate phase fronts until

target track is lost.

the angular separation between

the

global error

Since

directions decreases

with increasing

d,

efforts to reduce local error bound will increase the possibility of committing a global error and should be examined
carefully.

Figure 3.5

shows the maximum

local angular error

phase comparison system (normalized by

antenna separation.

33

TT

of a

as a function of

(a)

Target

^A^R +27r

(b)

Figure 3.4

Local and Global Angular Tracking Errors

(a)

A Local Angular Tracking Error.

(b)

A Global Angular Tracking Error.

If

r
Ls.

1.0

75

(Radian)

7T

50

25

..

,-

1-0

Figure 3.5

d/x

3.0

2.0

4.0

Maximum Local Angular Error of

5.0

6.0

Phase Compa-

rison System as a Function of Antenna

35

Separation

AZIH5THAL ANGlHfAl ERROR FOR

IV.

and the phase shift

across which equations

tered

in the

TWO^ ELEMENT TARGET

consideration of which the

One simple
D(r,6,0)

possess

d (6, 0)

and (13)

(10)

phase comparison

phase distortion
a

singularity

do not apply is encoun-

tracking

of a

two-element

The two reflecting elements appear to be two dipole

target.

radiators with

the tracking antennas

Assume

phase.

180 degrees

out of

are polarized

in the

magnitude but

the same

z-direction. The target can be modeled as two delta-function

current elements located on the y-axis.

The current element

has a phase of zero degree while the

at y = 1/2

-1/2 has a phase of -180 degrees.

one at

y =

As r approaches infinity,

the z-component of the electric field strength becomes

E z (r ,6,0)

(Wy*t/2

<

rT)

sine. sin (kl. sine. sin0/2)

{exp (ikr)

/r

Thus
o*0*ir

io
-IT

Assume that

interval {-%1f)

-Tf^O

assigns phases to the

The phase fronts of the wave and a pair of

the tracking system

antennas with separation

d =

2"X

are shown in figure 4.1

Figure 4.2 shows the maximal local error (curve 1),

the
first global error when the two antennas are separated by 2*ff
(curve

2)

and a higher order global error

antennas are separated

by

4*rr

It

(curve

3)

when the

can be seen that

as d/A

although the local error decreases,

the separation between local and global errors also decreases. Thus it

increases,

is difficult to justify whether

it is desirable to increase

antenna separation to reduce the local error.

36

1>

-\

v-

-1

[Angular Tracking Error|

Figure 4.1

Sin (1/4)

14.4

Angular Tracking Error across a Singularity

of the Phase Distortion.

37

l.o

7T

Radian)

75

50

25

d/a
1.0

3.0

2.0

4.0

5.0

6.0

Figure 4.2 Maximum Local Angular Error and Global Error

Direction for a Two Element Target as a Function of Antenna Separation.
1) Maximal Local Error.
2) First Global Error.
3) Second Global Error.

38

CONCLOSION

V.

Howard's phase

MD

REMARKS

front distortion technique is

proved to

be applicable only to a phase comparison tracking system and

only if the phase shift of the scattered field of the target

is smooth across the solid angle containing the antennas and

if the antennas are far from the target so that only the 1/r

term in the far field is retained.

are extensions

of Howard's techniques

the target

long as

Equations

still extends

(10)

to closer

and (13)

ranges as

negligibly small

only a

angle to each of the antennas compared to their beam widths.

All these equations are approximations

to the phase compen-

sation equation which describes the mechanism of determining

the direction of the tracking

axis.

equation shows that the rate of

tion

across

the

The phase compensation

change of the phase distor-

finite separation

between

the

antennas

governs the tracking axis direction. Rapid variations in the

phase distortion between the antennas are averaged out.

Howard's techniques

and its extensions,

equation

(10)

approximate this rate of variation with one over

midpoint between
an inf iuitesimally
small distarce at the
the antennas. Hence angular tracking errors will be exagger-

and

ated

(13)

if

the

phase

distortion

varies

rapidly

over

the

antennas separation, which can happen with a complex target

large compared to the wavelength being used.
Through the phase compensation equation,
global errors
are found to
exist for a phase comparison tracking system.
For the local error an upper bound is obtained.
The antenna
separation is an important parameter

in all these findings.

angular error in a phase comparison tracking

system should be investigated thoroughly.

Its effect on

39

LIST OF REFERENCES

1.

Skolnik.
Merrill,
Introduction to Radar
McGraw-Hill, McGraw-HiIT7~fl'ew~Yofk~,~T980

2.

D.D.Howard,
and A.M.King "Phenomena of
J.H.Dunn,
Noise
in
Radar Tracking
Scintillation
Systems"
Proceeding of the IRE, 1959

3.

R.J.Sims and E.R.Graf "The Reduction of Radar Glint by

Diversity Techniques",
IEEE, Transactions on Antennas
Vol. At- 19,
mo. 4, juiy
and Propagation,
9 / 1 ,
pp~
3

Systems.

462-468.

4.

D.D.Howard "Radar Target Angular Scintillation in

Tracking and Guidance Systems Based on Echo Signal
Phase Front Distortion" Proceeding Nat'l Electronics
Conference , Vol. 15, Octobe r 959, pp. 840-849.
1

5.

J.H.Dunn

"Radar Target
Amplitude,
and D.D.Howard
Angle,
and Doppler Scintillation from Analysis of the
Echo Signal Propagating in Space", IEEE Transaction on
Microwave Theory and
Vol.MTT-l6,No.9
Techniques^
September 968.
1

6.

Lee,Hung-Mou, "One Term Correction to Meade's Analysis

of a
Two Element Target",
paper submitted to IEEE
Transactions on Aerospace and Electronic System.

7.

Lee,Hung-Mou,
"Phase Front Distortion and the Angular
a
Tracking
Error
of
Phase
Comparison Monopuise
Tracking System", paper submitted to IEEE Transactions
on Aerospace and Electronic System.

8.

Lee,Hung-Mou,
"Analysis of Angular Tracking Error of
an Amplitude Comparison Monopuise
System",
paper
submitted to IEEE Transactions
on Aerospace and
Electronic System.

9.

Park,
Dae Hyun Inherent Angular Tracking Error of an
Amplitude Compari~so~n Monopuise System ,
M. S. Thesis,
naval postgraduate Scnool, Monterey, California,
in
preparation.

40

INITIAL DISTBIEUTION LIST

No.

Copies

1.

Library, Code 0142

Naval Postgraduate Shool
Monterey, California 93943

2.

Department Chairnan, Code 62Rr

1
Department of Electrical and Computer Engineering
Naval Posgraduate Shool
Monterey, California 93943

3.

Professor H-M. Lee, Code 62LH

2
Department of Electrical and Computer Engineering
Naval Posgraduate Shool
Monterey, California 93943

4.

Professor M.A. Morgan, Code 62MW

1
Department of Electrical and Computer Engineering
Naval Posgraduate Shool
Monterey, California 93943

5.

LCDE.Sopon Bumrcongpol
Electronics Department
Naval Dock Yard
Royal Thai Navy, Bangkok,

6.

Thailand

Defense Technical Information Center

Cameron Station
Alexandria, Virginia 22314

41

13

3 7

>:y?52
Thesis
B8U72
c. 1

Bumroongpol
Angular tracking
error in a phase comparison monopulse
tracking radar , a critical review and extension of the phase
front distortion approach.