George W. Stimson introduction to Airborne Radar (Se)

Подождите немного. Документ загружается.

and the wider the band of frequencies it occupies, the high-

er the angular resolution that can be obtained through

doppler processing.

Sidelobe Clutter

The radar return received through the antenna’s side-

lobes is always undesirable and so is called sidelobe clutter

(SLC). Excluding the altitude return, sidelobe clutter is not

nearly as concentrated (less power per unit of doppler fre-

quency) as mainlobe clutter. But it covers a much wider

band of frequencies.

Frequency and Power. Sidelobes extend in virtually all

directions, even to the rear. Therefore, regardless of the

antenna look angle, there are always sidelobes pointing

ahead, behind, and at virtually every angle in between. As a

result, the band of frequencies covered by the sidelobe clut-

ter extends from a positive frequency corresponding to the

radar’s velocity (f

= 2V

/λ) to an equal negative (less than

the transmitter’s) frequency (Fig. 15).

CHAPTER 22 Sources and Spectra of Ground Return

299

15. Because sidelobes are radiated in all directions, sidelobe clut-

ter extends from a positive frequency corresponding to the

radar’s velocity to an equal negative frequency.

14. As the antenna beam sweeps through its search scan, the mainlobe clutter spectrum moves out to its maximum frequency and squeezes into

a narrow line, then returns again.

PART V Return from the Ground

300

While the power radiated in any one direction through

the sidelobes is relatively slight, the area illuminated by the

sidelobes is extremely large. Moreover, much of the clutter

comes from relatively short ranges. This is so, even out to

the ends of the sidelobe clutter spectrum. As illustrated in

Fig. 16, since the cosine of a small angle is nearly equal to

one, if a radar is at an altitude of 6000 feet, return from a

range of only 4 nautical miles (depression angle = 14°) will

have a doppler frequency only 3 percent less than the maxi-

mum (2V

/λ).

In aggregate, therefore, not only can the sidelobe clutter

power be substantial, but the clutter may be spread more

or less uniformly over a broad band of doppler frequen-

cies.

Impact on Target Detection. The extent to which the

clutter interferes with target detection depends on the fre-

quency discrimination the radar provides. This is illustrated

by the map in Fig. 17 (below).

Plotted on it are lines of constant doppler frequency.

They are called isodoppler contours. Each line represents

the intersection between the ground and the surface of a

cone around the radar’s velocity vector. Since the angle

between this vector and every point on the cone is the

same, the return from every point along the contour has the

same doppler frequency. Just as the distances between con-

tour lines on a relief map correspond to a fixed interval of

elevation, so the distances between isodoppler lines corre-

spond to a fixed interval of doppler frequency.

Let us suppose now that the doppler interval corre-

sponds to the minimum difference in doppler frequency

which can be discerned by a particular radar—i.e., its

doppler resolution. If the radar differentiates between tar-

16. At an altitude of 6,000 feet, sidelobe return from a range of

only 4 nautical miles will have a doppler frequency almost

equal to the maximum sidelobe clutter frequency.

17. Lines of constant doppler frequency—isodoppler contours. Each corresponds to the intersection of a cone about

the radar’s velocity vector with the ground.

Click for high-quality image

gets and clutter solely on the basis of doppler frequency, a

target falling amid the sidelobe clutter must compete with

the return from the entire strip of ground between the con-

tours bracketing the target’s doppler frequency.

And, as is made clear in Fig. 18, depending upon the tar-

get’s range rate, much of this ground may be at substantially

closer range than the target.

To appreciate this point, bear in mind that the strength

of radar return is inversely proportional to the fourth power

of the range from which it is received. For given values of

antenna gain and backscattering coefficient, the power of

the return from a ground patch at a range of, say, 1 mile is

(10/1)

= 10,000 times (40 dB greater than) that of the

return from a ground patch of the same size at a range of 10

miles.

If the radar also provides range resolution (as by range

gating), the target must compete only with that portion of

the clutter passed by the same range gate as the target’s

echoes.

Other Factors Governing Sidelobe Clutter Strength.

The strength of the sidelobe return from any one patch of

ground depends upon several factors besides range. One is

the gain of the particular sidelobe within which the patch

lies. In a representative fighter radar, the two-way gain of

the first sidelobe beyond the mainlobe is on the order of

100 times (20 dB) stronger than that of the weaker side-

lobes.

The strength of the return also varies widely with the

nature of the terrain included in the patch—its backscatter-

ing coefficient.

As explained previously, as the grazing angle increases,

the backscattering coefficient also increases. Consequently,

even though a radar is closer to the ground at low altitudes,

sidelobe clutter may be most severe when flying at moder-

ate rather than low altitudes.

Significance. Clearly, the extent to which sidelobe clutter

is a problem depends upon many things.

• Frequency resolution provided by the radar

• Range resolution provided by the radar

• Gain of the sidelobes

• Altitude of the radar

• Backscattering coefficient and angle of incidence

Also, as already noted, certain man-made objects can be

immensely important sources of sidelobe clutter. (They will

be discussed separately at the end of this chapter.)

CHAPTER 22 Sources and Spectra of Ground Return

301

18. If a radar differentiates between target echoes and ground

return solely on basis of doppler frequency, sidelobe clutter

can present a serious problem.

Click for high-quality image

PART V Return from the Ground

302

This return is called the altitude return, since generally its

range equals the radar’s absolute altitude.

Relative Strength. The altitude return is not only much

stronger than the surrounding sidelobe clutter but may be

as strong or stronger than the mainlobe clutter.

For, the

area from which it comes not only may be very large but is

often at extremely close range.

This point is illustrated in Fig. 20. A radar is at an alti-

tude (h) of 6000 feet over flat terrain. As you can see, the

slant range to the ground at an angle of incidence

equals

h/cos

. The cosine of a small angle being only slightly less

than one, even when

is as much as 22° the slant range to

the ground is only 500 feet greater than the altitude (verti-

cal range).

If the slant range at an angle of incidence of 22° is rotat-

ed about the vertical axis, it traces a circle on the ground

having a diameter of roughly 5000 feet (Fig. 21).

A circle this size contains nearly 20 million square feet.

Thus, the radar receives all of the return from a 20-million-

square-foot area, at a range of just 1 nautical mile, in the

round-trip transit time for a range increment of 500 feet.

That time is only 1 microsecond.

Furthermore, at near vertical incidence, the backscatter-

ing coefficient tends to be very large. Over water, the coeffi-

cient is enormous. Little wonder, then, that the altitude

return appears as a sharp spike in a plot of amplitude ver-

sus range.

Doppler Frequency. The altitude return peaks up in a

plot of amplitude versus frequency, too, but not sharply.

The reason can be seen from Fig. 22 at the top of the facing

page.

20. At an altitude of 6,000 feet, even at an angle of incidence

of 22˚, the slant range to the ground is only 500 feet greater

than the altitude.

21. Yet, at an altitude of 6,000 feet, an angle of incidence of 22˚

encompasses a circle having an area of 20 million square feet.

Altitude Return

Beneath an aircraft, there is usually a region of consider-

able extent within which the ground is so close to being at a

single range that the sidelobe return from it appears as a

spike on a plot of amplitude versus range (Fig. 19).

19. Altitude return comes from a large area, often at very close

range.

2. In the case of an altimeter, the

mainlobe return is the altitude

return.

The projection of the radar velocity, V

, on the slant

range to the ground equals V

sin

. Unlike the cosine, the

sine of an angle changes most rapidly as the angle goes

through zero. While the doppler frequency of the clutter is

zero when θ is zero, it increases to nearly 40 percent of its

maximum value (2 V

/λ) at an angle

of only 22°. Return

from a circle of ground that produces a sharp spike in a

plot of amplitude versus range, therefore, produces only a

broad hump in a plot of amplitude versus doppler fre-

quency (Fig. 23).

CHAPTER 22 Sources and Spectra of Ground Return

303

23. Since sine

changes most rapidly as

goes through zero,

altitude return is spread over a comparatively broad band of

doppler frequencies.

24. Doppler frequency of altitude return is normally low, but may

be quite high in a dive.

Normally, the doppler frequency of the altitude return is

centered at zero. However, if the altitude of the radar is

changing—as when the aircraft is climbing, diving, or flying

over sloping terrain—this will not be so. For a dive, the

doppler frequency will be positive (Fig. 24); for a climb, it

will be negative. Even though the frequency is generally

fairly low, it can be considerable. In a 30° dive, for instance,

the altitude would be changing at a rate equal to half the

radar velocity.

Significance. Despite its strength, the altitude return is

usually less difficult to deal with than the other ground

returns. Not only does it come from a single range, but its

range is predictable; and, as we have seen, its frequency is

generally close to zero—though that unfortunately is the

doppler frequency of a target pursued at constant range

(e.g., a tail chase with zero closing rate, R

⋅

= 0).

Relation of Clutter Spectrum to Target Frequencies

Having become familiar with the characteristics of main-

lobe clutter, sidelobe clutter, and altitude return individual-

ly, let us look briefly at the composite clutter spectrum and

its relationship to the frequencies of the echoes from repre-

22. Doppler frequency of return from a given angle of incidence,

θ, is proportional to projection of radar velocity onto slant

range at that angle, hence to the sine of θ.

PART V Return from the Ground

304

sentative airborne targets in typical operational situations.

Again, we will assume that the PRF is high enough to avoid

doppler ambiguities.

Figure 25 illustrates the relationship between target and

clutter frequencies for a nose-aspect approach. Because the

target’s closing rate is greater than the radar’s velocity, the

target’s doppler frequency is greater than that of any of the

ground return.

Figure 26 (below) shows the relationship for a tail chase.

Because the target’s range rate is less than the radar’s veloci-

ty, the target’s doppler frequency falls within the band of

sidelobe clutter. Just where, depends upon the range rate.

26. Doppler frequency of target falls within sidelobe clutter.

27. Target is buried in mainlobe clutter. Fortunately, a target will

attain such a relationship only occasionally and usually will

remain in it fleetingly.

28. Target is buried in altitude return.

25. Doppler frequency of target is greater than that of any ground

return.

In Fig. 27, the target’s velocity is perpendicular to the

line of sight from the radar; the target has the same doppler

frequency as the mainlobe clutter. Fortunately, a target will

attain such a relationship only occasionally and usually will

remain in it fleetingly.

In Fig. 28, the target’s closing rate is zero; the target has

the same doppler frequency as the altitude return.

TAIL CHASE, ZERO CLOSING RATE (V

= V

)

Click for high-quality image

Figure 29 shows two opening targets. Target A has an

opening rate that is greater than the radar’s ground speed

), so this target appears in the clear beyond the negative-

frequency end of the sidelobe clutter spectrum. On the

other hand, Target B has an opening rate less than V

. So

this target appears within the negative frequency portion of

the sidelobe clutter spectrum.

With these situations as a guide, the relationship of the

doppler frequencies of target return and ground return for

virtually any situation can easily be pictured (Fig. 30). Bear

in mind, though, that at lower PRFs doppler ambiguities

can occur which may cause a target and a ground patch

having quite different range rates to appear to have the

same doppler frequency.

The consequences of such ambiguities will be discussed

in detail in the next chapter. The signal processing com-

monly performed to separate target echoes from ground

clutter in the various operational situations presented here,

as well as when operating at PRFs which make doppler fre-

quencies ambiguous, will be described in Chaps. 26, 27,

and 28.

CHAPTER 22 Sources and Spectra of Ground Return

305

30. Relationships between target doppler frequency and ground-clutter spectrum for various target closing rates. (Doppler frequencies assumed

to be unambiguous.)

29. If range rate is more negative than –VR, target appears in

clear (A) below sidelobe clutter spectrum; otherwise, it

appears in negative frequency half of the spectrum (B).

Click for high-quality image

PART V Return from the Ground

306

Return from Objects on the Terrain

The return from certain man-made structures can be

very strong. Viewed straight on by an X-band radar, a

smooth, flat metal sign only 4 feet on a side, for example,

has a radar cross section on the order of 320,000 square

feet (Fig. 31) compared to 10 square feet or less for a small

aircraft in some aspects.

This may sound absurd, but not if you stop to think.

Most of the power intercepted by the sign when viewed

straight-on will be reflected back in the direction of the

radar. The sign is a specular (mirrorlike) reflector. It acts, in

fact, just like an antenna that is trained on the radar and

reradiates all of the transmitted power it intercepts, back in

the radar’s direction. At X-band frequencies, for example,

an antenna with a 16-square foot aperture has a gain of

around 20,000. Multiply the area of the sign by this gain

(16 x 20,000) and you get a radar cross section of 320,000

square feet.

In principle, the radar return from a flat reflecting sur-

face, such as a sign, is directly comparable to the intense

reflections one frequently gets from the windshield of a car

or the window of a hillside house when it is struck from

just the right angle by the early morning or late afternoon

sun (Fig. 32).

Whereas a single flat surface such as a sign must be

viewed from nearly straight-on to reflect the incident ener-

gy back to the radar, two surfaces forming a 90° corner

will do so over a wide range of angles in a plane normal to

the intersection of the surfaces. They are what is called

retroreflective. If a third surface is added at right angles to

the other two (forming a corner reflector), the range of

angles over which the surfaces will be retroreflective may be

increased to nearly a quarter of a hemisphere (Fig. 33).

This, incidentally, is the way bicycle reflectors work.

Portions of a large building may act like corner reflectors,

and a vehicle such as a truck may look like a group of cor-

ner reflectors (Fig. 34).

Because of their enormous radar cross sections, retro-

reflective objects on the ground can produce sidelobe

return as strong or stronger than the echoes from distant

aircraft received through the mainlobe. Furthermore,

because the objects are of limited geographic extent—they

are discrete as opposed to distributed reflectors—all of the

return from one of them has very nearly the same doppler

frequency and comes from very nearly the same range. The

return may appear to the radar, therefore, exactly as if it

came from an aircraft in the mainlobe.

Naturally, since these objects are virtually all man-made,

they are much more numerous in urban than in rural areas.

31. Viewed straight-on, a smooth

flat plate can have an immense

radar cross section.

3. The surface variations are

small in comparison to a

wavelength.

32. Radar echoes from certain man-made objects on the ground

are comparable to reflection from a window struck from just

the right angle by the sun.

33. Corner formed by two flat surfaces is retroreflective over wide

range of angles; that formed by three surfaces, over wider

range still.

34. Portions of a building may act like corner reflectors; a truck

like several corner reflectors.

Nevertheless, they may be encountered almost anywhere.

In the little populated region of Southern California’s

Antelope Valley, for example, the long, low corrugated

metal sheds of the turkey ranchers (Fig. 35) return tremen-

dously strong echoes.

Depending on the use of the radar, special measures may

be required to reduce or eliminate sidelobe return of this

sort. Mainlobe return from such objects is usually not a

problem, since its doppler frequency is generally different

from that of targets of interest. But if the objects are moving

or have extraordinarily large radar cross sections, the main-

lobe return, too, can be a problem.

Summary

The same factors govern the power of ground return as

that from an aircraft. Backscattering from the ground, how-

ever, is expressed in terms of an incremental coefficient,

which must be multiplied by the area of a ground patch to

obtain its radar cross section,

. The coefficient

varies

with angle of incidence, frequency, polarization, electrical

characteristics of the ground, roughness of the terrain, and

nature of the objects on it. Generally, the variation with

angle is due primarily to foreshortening of the ground area

as viewed from the radar.

The most important ground return—and the only return

of interest for ground mapping etc.—is that received

through the antenna’s mainlobe. When the antenna is look-

ing straight ahead, the doppler frequency of this return cor-

responds to the radar’s full ground speed. As the look angle

increases, the frequency decreases and spreads over an

increasingly broad band. Both the center frequency and the

width of this band increase directly with the radar’s velocity

and are inversely proportional to wavelength.

Ground return received through the sidelobes, though

comparatively weak at any one frequency, extends from a

positive frequency corresponding to the radar’s full velocity

(2V

/λ) to an equal negative frequency. The portion

received from directly below—the altitude return—is espe-

cially strong, particularly over water. It appears as a spike

on a plot of amplitude versus range and as a broad hump

on the doppler spectrum. Its center doppler frequency is

normally zero.

Man-made objects on the ground may be highly retro-

reflective and can produce sidelobe return as strong as tar-

get echoes received through the mainlobe.

If the PRF is high enough to eliminate ambiguities,

whether a target’s echoes and the ground return have differ-

ent doppler frequencies—as is desirable for ground clutter

rejection—depends upon the target’s closing rate. As long

CHAPTER 22 Sources and Spectra of Ground Return

307

35. Even in little populated regions, there may be numerous struc-

tures having tremendous radar cross sections.

PART V Return from the Ground

308

as it is greater than the radar’s velocity, the echoes will lie

outside the ground return. Otherwise, they must compete

with sidelobe return. Only if the target is flying at right

angles to the line of sight from the radar, will its echoes

have the same frequency as the mainlobe return. Only if the

closing rate is zero, will the target echoes have the same fre-

quency as the altitude return.

However, as we shall see in the next chapter, doppler

ambiguities can cause a target and ground patches, having

quite different range rates, to appear to have the same

doppler frequency, thereby greatly compounding the prob-

lem of separating target echoes from clutter.