George W. Stimson introduction to Airborne Radar (Se)

Подождите немного. Документ загружается.

Effect of Range and

Doppler Ambiguities on

Ground Clutter

309

n the last chapter, we surveyed the sources of ground

return and became acquainted with its doppler spec-

trum. But we did not consider the profound effects of

range and doppler ambiguities on ground return.

Although we discussed both types of ambiguities in detail

in earlier chapters, the discussions there involved only tar-

get return. If a radar is searching for or tracking a target in

the presence of ground clutter, however, the consequences

of ambiguities in the clutter are quite different from the

consequences of the same ambiguities in the target return.

In the case of a target, we are interested in the target

itself and in the value of its range or doppler frequency.

Since a target such as an aircraft is essentially a point

source, ambiguities simply give the observed range or

doppler frequency more than one possible value. If the

ambiguities are not too severe, we can resolve them

through such techniques as PRF switching.

In the case of ground clutter, on the other hand, we are

interested only in differences in range and doppler frequen-

cy which will enable us to separate the clutter from the tar-

get echoes. Since the sources of the clutter generally are

widely dispersed, ambiguities tend to wash out these dif-

ferences. About all that may be accomplished through res-

olution techniques such as PRF switching is to shift blocks

of clutter about.

In this chapter, after briefly considering the dispersed

nature of ground clutter, we will examine the effects of

ambiguities on the range and doppler profiles for a repre-

sentative flight situation and see how they compound the

problem of separating target echoes from clutter.

PART V Return from the Ground

310

Dispersed Nature of the Clutter

As we saw in Chap. 22, when the antenna beam strikes

the ground, it usually illuminates an area which is extensive

in both range and angle. Furthermore, the antenna invari-

ably has sidelobes through which it radiates an appreciable

amount of energy (Fig. 1).

1. When the antenna beam strikes the ground, it illuminates an

area extensive in both azimuth and elevation. Additionally,

sidelobes illuminate ground in virtually all directions.

2. Representative flight situation. Targets include both low and high closing rate aircraft plus a truck.

Ground return of various amplitudes is thus received

from a great many different ranges and directions. Since the

direction to a point on the ground in large measure deter-

mines the point’s range rate, the return also covers a broad

band of doppler frequencies.

Naturally, any dispersion in range and frequency of the

ground return makes the problem of separating target

echoes from it more difficult. Ambiguities in range and

doppler frequency compound the problem by causing clut-

ter from more than one block of ranges and more than one

portion of the doppler spectrum to be superimposed. The

effect of this can be visualized most clearly by examining

separately the range and doppler-frequency profiles for a

representative flight situation.

We will assume that a radar-equipped aircraft is flying at

low altitude over terrain from which a considerable amount

of ground return is received. The radar antenna is looking

down at a slightly negative elevation angle and to the right

at an angle of about 30˚.

Two targets, A and B, are in the antenna’s mainlobe.

Target A is being overtaken from the rear and so has a low

closing rate. Target B is approaching the radar head-on and

so has a high closing rate (Fig. 2).

Click for high-quality image

For purposes of illustration, Target A has been placed at

a range from which only sidelobe return is being received;

Target B, at a range from which both mainlobe and sidelobe

return are being received (Fig. 3).

Within the ground patch illuminated by the mainlobe is

a truck. It is heading toward the radar and so has a slightly

higher closing rate than the ground it is traveling over.

The flight situation diagram is repeated in Fig. 4, with

the corresponding “true” range profile beneath it. This pro-

file is simply a plot of the amplitude of the radar return ver-

sus the range of its sources relative to the radar—i.e., slant

range, as opposed to horizontal range on the ground.

Sidelobe clutter, you will notice, extends outward from a

range equal to the radar’s altitude. Notice how rapidly it

decreases in amplitude as the range increases.

The echoes from Target A stand out clearly above the

sidelobe clutter. By contrast, the echoes from Target B and

the truck are completely obscured by the much stronger

mainlobe clutter. Even though we know exactly where to

look for these echoes, we cannot distinguish them from the

clutter.

Toward the left end of the range profile, you will notice

two strong spikes. The one at zero range is due to what is

called transmitter spillover—energy from the transmitter

that leaks into the receiver during transmission, despite all

efforts to block it. The second spike is the altitude return.

Range Ambiguities

Range ambiguities arise when all of the echoes from one

pulse are not received before the next pulse is transmitted.

As explained in detail in Chap. 12, when return is received

from beyond the unambiguous range, R

, it is impossible

to tell which pulse a particular echo belongs to (Fig. 5).

CHAPTER 23 Effect of Range and Doppler Ambiguities on Ground Clutter

311

4. True range profile of representative flight situation. Target A

can be seen clearly above sidelobe clutter, but target B and

truck are obscured by mainlobe clutter.

5. Range ambiguities occur when return from one pulse is being

received after the next pulse is transmitted. Range corre-

sponding to interpulse period, T, is R

3. Target A is at a range from which only sidelobe return is

received; target B, at a range from which both mainlobe and

sidelobe return are received.

But, far more important from the standpoint of clutter

rejection, the returns from ranges separated by R

are

received simultaneously. The echoes from a target, therefore,

must compete with ground return not only from the target’s

own range but from every range that is separated from it by

a whole multiple of R

Click for high-quality image

PART V Return from the Ground

312

Returns from Ranges Separated by R

. To illustrate the

effect of range ambiguities on the range profile observed at

the output of the receiver, Fig. 6 traces the paths of the

echoes of three successive transmitted pulses from three

points on the ground—“a,” “b,” and “c”—which are sepa-

rated in range by R

. The key points in the figure (called

out by the large lettered arrows) are the following.

(A) An echo of pulse No. 1 is reflected from the most

distant point, “a.”

(B) This echo reaches the next most distant point, “b,”

just as it is reflecting an echo of pulse No. 2.

the near point, “c,” just as it is reflecting an echo of

pulse No. 3. The three echoes travel the remaining

distance to the radar together.

(D) They arrive simultaneously and appear on the radar

display as though received from a single range. All

three points have the same apparent range.

An instant later, the echoes of the same pulses from

points just beyond “a,” “b,” and “c” arrive simultaneously.

An instant after that, so do the echoes from points just

beyond these, and so on.

Thus, the range profile is, in effect, broken into seg-

ments, R

wide, which are superimposed, one over the

other (Fig. 7).

6. Return of echoes of three successive transmitted pulses from

points on the ground which are separated in range by the

unambiguous range, R

7. Echoes from points just beyond “c,” “b,” and “a” are likewise

received simultaneously. So are echoes from points just beyond

these, and on and on.

Range Zones. Now the range of point “a” in Figure 6 was

selected more or less at random. It could have been any

range within the region from which return is received.

Let us assume now that the range of “a” is such as to

place point “c” at zero range (right at the radar antenna).

Then the echoes of all ranges between “c” and “b” would be

first-time-around echoes; i.e., they would be echoes of the

immediately preceding (last) transmitted pulse.

The echoes from all ranges between “b” and “a” would be

second-time-around echoes; they would be echoes of the

pulse before the immediately preceding one.

Likewise, the echoes from all points between “a” and a

point R

beyond “a” would be third-time-around echoes,

and so on.

The particular segments into which the range profile

would in this case be divided (Fig. 8) are called range zones

(or ambiguity zones).

Although the true range profile could similarly be divid-

ed into any number of different sets of contiguous zones R

wide, this particular set was chosen for two reasons. First, it

conveniently starts at zero range. Second, the true range of

every point within any one zone is the point’s apparent

range plus the same whole multiple of R

Now, the higher the PRF is, the shorter R

will be, hence

the narrower the range zones. The narrower the zones, the

greater the number of segments into which the true range

profile will be divided and from which returns will be

received simultaneously.

As shown in Chap. 12, R

very nearly equals 80 nautical

miles divided by the PRF in kilohertz.

Width of range zones =

nmi

If, for example, the PRF were 4 kilohertz, the range

zones would be 20 nautical miles wide. If returns were

received from ranges out to 60 miles, the true range profile

would be broken into 3 zones.

Range Zones Superimposed. The effect of breaking the

true range profile for our representative flight situation into

three range zones is illustrated in Fig. 9. There the returns

from Zones 2 and 3 are placed beneath the return from

Zone 1, and the corresponding ranges within the zones are

lined up. Beneath these plots is the composite profile that

would appear at the output of the receiver.

As you can see, superimposed over the echoes from

Target A are not only the sidelobe clutter from the target’s

own range but the much stronger close-in sidelobe clutter

from the corresponding range in Zone 1 and the still

stronger mainlobe clutter from the corresponding range in

Zone 3.

Similarly, superimposed over the echoes from Target B

and the truck are not only the mainlobe clutter from their

own ranges but the sidelobe clutter from the corresponding

ranges in Zones 1 and 2.

As the PRF is increased and the range zones narrow, the

amount of clutter that may be superimposed over any one

CHAPTER 23 Effect of Range and Doppler Ambiguities on Ground Clutter

313

8. If point “c” in the preceding example had been moved in to

zero range and points “b” and “a” had been moved equally,

the true range profile would have been broken into segments

called range zones.

1. In kilometers, R

equals 150

divided by the PRF.

9. Result of R

being one-third of the maximum detection range.

Range profile for representative flight situation is broken into

three range zones. Mainlobe clutter blankets composite profile

seen by radar.

Click for high-quality image

11. True doppler frequency profile of representative flight situa-

tion. Target B and truck, having higher closing rates than the

ground, appear in clear.

12. Each element of radar return has sideband frequencies separat-

ed from the doppler-shifted carrier by multiples of the PRF, f

PART V Return from the Ground

314

target’s echoes increases (Fig. 10). If the PRF is increased

without limit, a point is ultimately reached where the radar

transmits continuously.

A target’s echoes must then compete with the ground

clutter received from all ranges.

Clearly, the more deeply the clutter is piled up, the less

able the radar will be to isolate target echoes from the clutter

on the basis of differences in range, and the greater the

extent to which the radar must depend on other means, such

as differences in doppler frequency, for clutter rejection.

Doppler Profile

The flight profile is shown with the true doppler fre-

quency profile beneath it in Fig. 11. This profile is a graph

of the amplitude of the radar return versus doppler fre-

quency. In plotting it, no attempt was made to differentiate

the return received at one point in the interpulse period

from the return received at another. The profile represents

the return from all ranges.

Sidelobe clutter extends from zero doppler frequency out

in both positive and negative directions to frequencies cor-

responding to the radar’s full velocity (f

= ±2V

/λ). The

spike at zero frequency is the transmitter spillover. The

broad hump under it is the altitude return. The narrower

hump near the maximum positive sidelobe-clutter frequen-

cy is mainlobe clutter.

Target A is being overtaken, so its doppler frequency falls

below that of the mainlobe clutter, in the band of frequen-

cies blanketed by sidelobe clutter. Because a good deal of

this clutter comes from shorter ranges than the target’s, the

target echoes barely protrude above the clutter. If the target

were smaller or at much greater range, its echoes might not

even be discernible.

Since Target B and the truck are approaching the radar

and are nearly dead ahead, they have higher doppler fre-

quencies than any of the clutter.

Doppler Ambiguities

As we learned in Chap. 16, when transmission is pulsed,

each element of the radar return has sideband frequencies

separated from the doppler shifted carrier frequency by

multiples of the pulse repetition frequency, f

(Fig. 12).

Thus, the portion of the altitude return which has zero

doppler frequency also appears to have doppler frequencies

of ± f

, ± 2f

, ± 3f

, ± 4f

etc.

Similarly, the portion of the return having a true doppler

frequency of, say, +100 hertz also appears to have doppler

frequencies of (100 ± f

), (100 ± 2 f

), (100 ± 3 f

), (100 ±

4 f

), etc.

10. The higher the PRF, the narrower the range zones and the

more deeply the clutter is piled up.

Click for high-quality image

The same is true of the return received from every other

point in the true doppler frequency profile (Fig. 13). The

entire profile is, therefore, repeated at intervals equal to f

above and below the carrier frequency of the transmitted

pulses. Because it is made up of doppler-shifted return from

the central spectral line of the transmitted signal, the true

profile is commonly referred to as the central line return.

Repetitions of the spectrum, or of portions of it such as the

mainlobe clutter, are then referred to as PRF lines.

If f

is sufficiently high, the repetitions of the doppler

spectrum will in no way affect the ability of the radar to dis-

criminate between target echoes and ground clutter. In fact,

if the radar’s doppler passband is no more than f

hertz

wide, the repetitions will not even be seen by the doppler

filter bank (Fig. 14).

CHAPTER 23 Effect of Range and Doppler Ambiguities on Ground Clutter

315

14. If f

is high enough, the nearest sidebands will be entirely out-

side the passband.

15. If f

is less than the width of the true doppler spectrum, repeti-

tions of the spectrum due to sideband frequencies will overlap

and actually merge to from the single composite profile shown

at bottom of the figure.

13. Spectrum of a portion of the altitude return. Elements shown

here are separated in doppler frequency by 100 hertz. Each

element has sidelobes separated from the doppler shifted car-

rier frequency by multiples of the PRF.

However, if f

is less than the width of the true doppler

profile (as it often must be made to reduce or eliminate

range ambiguities), the repetitions will overlap, and the

observed doppler frequencies will be ambiguous. This con-

dition is illustrated in Fig. 15 for a value of f

that is only

one-half the width of the true profile. In this case, the true

profile is overlapped by the repetitions immediately above

and below it. For clarity, each repetition is plotted on a sep-

arate baseline. In actuality, they would all merge into a sin-

gle composite profile. The central portion of the composite

profile is shown at the bottom of the figure.

Any overlapping of the repetitions of the profile, such as

that just illustrated, can result in a target’s echoes and

ground clutter passing through the same doppler filter(s),

even though the true doppler frequencies of the target and

the clutter may be quite different. Examples of this are

Target B and the truck, in Fig. 15. Whereas the true doppler

frequencies of these targets are actually higher than those of

any of the clutter, in the composite profile, both targets are

nearly obscured by sidelobe clutter.

Because the sideband frequencies are separated from the

central-line frequencies by multiples of the PRF, any one

segment of the composite spectrum is identical to every

Click for high-quality image

16. As PRF is decreased, repetitions of mainlobe clutter spectrum

move closer together, leaving less room in which to detect

targets.

PART V Return from the Ground

316

other segment of the same width on either side. As noted in

previous chapters, it is for this reason that the passband of

the doppler filter bank need be no more than f

hertz wide.

The repetitions of the true doppler profile overlap

increasingly as the PRF is reduced (Fig. 16). From the stand-

point of clutter rejection, reducing the PRF has two main

effects. First, more and more sidelobe clutter piles up in the

space between successive mainlobe clutter lines. Second,

and more important, the mainlobe clutter lines move closer

together. Since the width of these lines is independent of the

PRF, reducing the PRF causes the mainlobe clutter to occupy

an increasingly larger percentage of the receiver passband

and causes the altitude return and other close-in sidelobe

clutter to pile up increasingly in the space between.

As the percentage of the passband occupied by mainlobe

clutter increases, it becomes increasingly difficult to reject

even the mainlobe clutter on the basis of its doppler fre-

quency without at the same time rejecting a large percent-

age of the target echoes. If carried to its extreme, the over-

lap would ultimately reach a point where the mainlobe

clutter completely blanketed the receiver passband.

Clearly, the lower the PRF, the more severe the effect of

doppler ambiguities on ground clutter.

Summary

Since ground clutter is widely dispersed, range and

doppler ambiguities greatly compound the problem of iso-

lating target echoes from the clutter.

In effect, range ambiguities break the range profile into

zones, which are superimposed on one another. Because of

this superposition, a target’s echoes may be received simul-

taneously with clutter not only from the target’s own range

but from the corresponding range in every other range

zone. Increasing the PRF narrows the range zones and

increases the number of zones that are superimposed,

thereby making it increasingly difficult to isolate the target

echoes.

Doppler ambiguities cause successive repetitions of the

doppler profile to overlap. Because of this, a target’s echoes

may have to compete with clutter whose true doppler fre-

quency is quite different from the target’s. Increasing the

PRF moves successive repetitions of the mainlobe clutter

spectrum farther apart, thereby making it easier to isolate

the target echoes. Thus the relationship between PRF and

doppler ambiguities is just the opposite of that between

PRF and range ambiguities. The lower the PRF, the more

severe the effect of doppler ambiguities on ground clutter;

and the higher the PRF, the more severe the effect of range

ambiguities.

Click for high-quality image

317

Separating

Ground-Moving

Targets from Clutter

n earlier chapters, very little was said about the detec-

tion of moving targets on the ground—cars, trucks,

tanks. Except that low PRFs are generally required

for air-to-ground operation, it was tacitly assumed

that separating ground moving targets (GMTs) from compet-

ing ground return on the basis of differences in doppler

frequency is no different than separating airborne moving

targets from ground return.

This assumption, however, is only partially true. For the

radial component of the velocity of many ground moving

targets is so low that the returns from them are embedded

in mainlobe clutter and cannot be separated from it by

conventional moving target indication (MTI) techniques.

In this chapter, after briefly examining the problem, we

will be introduced to two highly effective techniques for

detecting such targets. One is called Classical DPCA, for

displaced phase center antenna. The other and newer tech-

nique is called notching or clutter nulling. We’ll take up

Classical DPCA first; then, notching; and, finally, a combi-

nation of the two. In closing we’ll touch on the adaptation

of these techniques to precise angle measurement.

Problem of Detecting “Slow” Moving Targets

The chief problem in detecting moving targets in low-

PRF modes—whether on the ground or in flight—is sepa-

rating the target returns from mainlobe clutter. As

explained in detail in Chap. 26, by employing a reasonably

long antenna and flying at comparatively low speeds, we

can reduce the width of the mainlobe clutter spectrum and

spread its repetitions far enough apart to provide a fairly

WHAT AN ANTENNA PHASE CENTER IS

Every antenna has a phase center. It is that point

in space where, if a hypothetical omnidirectional

point-source radiator were

placed . . .

. . . the signals received by

it from any source within

the field of regard of the

real antenna

would have the same radio

frequency phase as signals

from the same source re-

ceived by the real antenna.

If weighting of the antenna

(for sidelobe reduction) is

symmetrical, the position

of the phase center will be

same for all look angles.

But if the weighting is

nonsymmetrical—as in a

half aperture of a mono-

pulse antenna—the position

of the phase center will be

a function of the look angle.

Phase

Center

1. Doppler spectrum of ground moving targets. Special tech-

niques are needed to detect targets whose

true

doppler fre-

quencies fall within mainlobe clutter (MLC).

PART V The Problem of Ground Clutter

318

wide clutter-free region in which to detect moving targets

(Fig. 1). Moreover, any target whose apparent doppler fre-

quency falls within the mainlobe clutter can be periodically

moved into this region by switching the PRF among several

different widely separated values.

However, if the radial component of a target’s velocity is

so low that its true doppler frequency lies within the main-

lobe clutter, no amount of PRF switching will move the tar-

get’s returns out of the clutter. Consequently, in many appli-

cations a special “slow-moving-target” indication capability

is needed. Conceptually, the simplest is Classical DPCA.

Classical DPCA

This technique takes advantage of the fact that the

doppler shift in the frequency of the returns received from

the ground is due entirely

to the aircraft’s velocity.

Specifically, this shift—which is manifest as a progressive

pulse-to-pulse shift in the phase of the returns from any

one range—is the result of the forward displacement of the

radar antenna’s phase center (defined in the panel on the

preceeding page) from one interpulse period to the next.

For any two successive pulses, therefore, the shift can be

eliminated by displacing the antenna phase center by an

equal amount in the opposite direction before the second

pulse of the pair is transmitted. The second pulse will then

be transmitted from the same point in space as the first.

And how does one displace an antenna’s phase center?

Generally, the radar is provided with a two-segment side-

looking electronically steered antenna. The aircraft’s velocity

and the radar’s PRF are adjusted so that during each inter-

pulse period the aircraft will advance a distance precisely

equal to that between the phase centers of the two antenna

segments (Fig. 2).

Successive pulses then are alternately transmitted by the

two segments:

• Pulse (n) by the forward segment, pulse (n + 1) by

the aft segment;

• Pulse (n + 2) by the forward segment, pulse (n + 3)

by the aft segment, and so on.

As a result, the pulses of every pair—e.g., (n) and (n + 1) —

are transmitted from exactly the same point in space.

The returns of each pulse are received by the antenna

segment which transmitted the pulse. When the return

from any one range, R, is received, of course, the phase

center of that segment will have advanced a distance equal

to the aircraft velocity, V, times the round-trip transit time,

, for the range R.

2. In Classical DPCA, radar transmits alternate pulses with for-

ward and aft segments. By adjusting velocity, V, and PRF so

radar advances distance between segments’ phase centers

during interpulse period, pulse n and (n + 1) are transmitted

from the same point in space.

1. Some shift may also be due to

so-called “internal motion” of

the clutter scatterers, e.g.,

wind-blown trees.

T =

PRF

Pulse

(n)

Pulse

(n + 1)

V T

DPCA TRANSMISSION

Target

Time

PRF

Doppler Frequency, f

MLC

Return from a moving target

Moving

targets on

the ground