34 Design Technology of Synthetic Aperture Radar

PRT S E
B

Figure 2.13 PRF and azimuth sampling.

B and E is the Doppler bandwidth of point T in the radar echo. To meet the Nyquist
sampling theorem,
v
PRF ≥ Ba = (2.31)
𝜌a

2.3.4.2 Data Reception Interspersed with Transmitting Event

The transmitted and received pulses of spaceborne SAR are frequently separated by sev-
eral pulse repetition periods, and the echoes returned to the antenna are likely masked
by the subsequent transmitted pulses, making the echoes lost by the receiver. Therefore,
the echoes within the swath must completely fall into the receiving window, as shown
in Figure 2.14.
Suppose that the echo of the transmitted radar pulse can only be received after the ith
pulse repetition cycle, as in Figure 2.14. The condition that the echo receiving window
is not covered up by the transmitted pulse is
2Rn
i ⋅ PRT + 𝜏p + 𝜏pt ≤ Tnear =
c
2Rf
(i + 1) ⋅ PRT + 𝜏p + 𝜏pt ≤ Tfar = (2.32)
c
where 𝜏 p is the pulse width, PRT is the pulse repetition time, 𝜏 pt is the pulse protec-
tion time, and Tnear and Tfar are the echo delay times in near-end Rn and far-end Rf ,
respectively. From the above equations, the range of PRF is chosen by
i i+1
≤ PRF ≤ (2.33)
2Rn ∕c − 𝜏p − 𝜏pt 2Rf ∕c + 𝜏p + 𝜏pt

τpt

Tnear Tfar

τp

Receiving window
PRT

Figure 2.14 Data reception interspersed with transmitting event.

 Radar System Design 35

τpt

Nadir echo Tnear Tfar

τp
Receiving window
tnadir PRT

Figure 2.15 Schematic of avoiding nadir echo.

2.3.4.3 Avoiding Nadir Echo

The spatial path of the nadir echo is the shortest while the backscattering coefficient is
large when the electromagnetic wave is incoming vertically and the radar echo is inten-
sified. Therefore, even illuminated by side lobe, the intensity of the echo is relatively
strong, which can cause serious ambiguity of the useful signal in the swath. It is required
to avoid interference of the nadir echo by PRF selection, as shown in Figure 2.15.
Suppose tnadir is the delay time of the nadir echo, and the gap between the near-end
echo delay time Tnear and the far-end echo delay Tfar is larger than j times the pulse
repetition time. Then the condition of avoiding the nadir echo is
2Rn
tnadir + j ⋅ PRT + 𝜏p + 𝜏pt ≤ Tnear =
c
2Rf
tnadir + (j + 1) ⋅ PRT − 𝜏p − 𝜏pt ≥ Tfar = (2.34)
c
where tnadir = 2H
c
, and H is the height relative to the ground. From Eq. (2.34), the range
of PRF is given by
j j+1
≤ PRF ≤ (2.35)
2Rn ∕c − 𝜏p − 𝜏pt − 2Hr ∕c 2Rf ∕c + 𝜏p + 𝜏pt − 2Hr ∕c
From the above three requirements, the near end and far end of the swath can be
obtained after the radar viewing angle is defined, and the PRF range at different view-
ing angles, i.e. the zebra map, can be drawn. Figure 2.16 shows an example of the zebra
map of a spaceborne SAR. The dark strips represent the interference of the signal trans-
mitted, and the strip width represents the time of the pulse width plus the protection
time 𝜏 RP , which happens before and after the pulse. The light strips represent the inter-
ference of the nadir signal, and the strip width represents the width of the nadir echo
(normally assumed to be twice the pulse width). The blank areas represent the selectable
range, and the vertical lines are the selected PRF, whose length represents swath width
(Figure 2.16 shows the scope of the viewing angle). As the viewing angle increases, the
length becomes shorter. This is because variation scope of the viewing angle for a fixed
swath width is smaller as the viewing angle becomes larger.

2.3.5 Ambiguity
The ambiguity feature of SAR means that the nonartificial interference echoes from
outside the swath and the useful echoes inside the swath come into the radar receiver
simultaneously. After imaging signal processing, on one hand, SNR of the radar image
36 Design Technology of Synthetic Aperture Radar

50

45
Viewing angle (deg)
40

35

30

25
3200 3300 3400 3500 3600 3700 3800
PRF (Hz)

Figure 2.16 Zebra map of a spaceborne SAR.

decreases; on the other hand, it may cause a false image (ghost) in the low-reflection
region of the radar image. The ambiguity characteristic of SAR is not only related to
swath width and the azimuth resolution but also to the antenna area, the shape of the
pattern, and the selection of the PRF. For an airborne SAR system, the echoes are gen-
erally received within one pulse repetition time due to the close operation distance.
Moreover, the PRF is generally several times larger than the azimuth Doppler band-
width. Therefore, the ambiguity issue of airborne SAR is not prominent. However, the
ambiguity issue is very prominent in spaceborne SAR, due to the long operation dis-
tance and the fast moving speed. The ambiguity analysis plays an important role in the
design of spaceborne SAR. It is the key factor that decides the success or failure of the
spaceborne SAR design.
The ambiguity is measured by the ambiguity-to-signal ratio of SAR, that is, the ratio
of the ambiguity power to the power of the useful signals. Usually, the coupling signal
between the range direction and the azimuth ambiguity can be neglected in analyzing
system ambiguity or computing ambiguity-to-signal ratio.

2.3.5.1 Range Ambiguity

Range ambiguity refers to echoes from outside the swath, and it enters the radar receiver
together with the useful echoes inside the swath, leading the radar image quality to
decrease after signal processing. In spaceborne SAR, the echoes of one transmitted pulse
can only be received after one or several pulses, which causes a serious ambiguity prob-
lem. Figure 2.17 shows the range ambiguity, where T P is the sampling period, R1 is the
distance between radar and target, T w is the target recording window, and T w ′ is the
width of the main lobe.
The range ambiguity-to-signal ratio (RASR) is calculated by
∑
N
Sai
i=1
RASR = (2.36)
∑N
Si
i=1
 Radar System Design 37

t = –Tp

t=0
t = +Tp
Track of subastral point

R1
R1–(c/2)Tp
R1+(c/2)Tp

r
Ambiguity zone
Ambiguity
r zone

r
S
Range ambiguity
Tw

–Tp Tw´ Tp t

Figure 2.17 Schematic of range ambiguity.

where Sai is the ambiguity power of the ith sampling point inside the echo recording win-
dow, and Si is signal power of the ith sampling point inside the echo recording window.

2.3.5.2 Azimuth Ambiguity

The Doppler frequency difference between the target echoes in some angles and the
main beam is an integral multiple of the PRF azimuth ambiguity caused by Doppler fre-
quency foldback. Azimuth ambiguity is mainly caused by finite samplings of the Doppler
spectrum at an interval of PRF. Because the echo spectrum of the target is repeated in
PRF, echo signals outside the main spectrum will fold back into the main spectral region,
as shown in Figure 2.18, where BD is the Doppler bandwidth, BP is the signal processing
bandwidth, and f P is the PRF.
The ratio of ambiguity to the expected signal is defined as the azimuth ambiguity-to-
signal ratio (AASR), which is calculated by
∞
∑ Bp∕2
∫−Bp∕2 G2 (fd + mf p )df d
m = −∞
m≠0
AASR = Bp∕2
(2.37)
∫−Bp∕2 G2 (f d )df d

where G is the azimuth antenna pattern, and m is normally −10 to 10. The influence of
m on AASR is very small outside this range. Of course, the range can be enlarged.
The key of ambiguity-to-signal ratio calculation is to compute the relationship
between azimuth round-trip far-field antenna power pattern G2 and f d . For a uniform
38 Design Technology of Synthetic Aperture Radar

Energy of the echo

BD

BP

–fP 0 fP fD

Azimuth ambiguity signal

Figure 2.18 Schematic of azimuth ambiguity.

weighted antenna aperture, the antenna power pattern is

[ [ ] ]2
sin 𝜋 ( a∕𝜆) sin 𝜃
L
G= (2.38)
𝜋 ( La∕𝜆) sin 𝜃
2vs 2vs 𝜆
Assuming fd = 𝜆
sin 𝜃 ≈ 𝜆
𝜃, then sin 𝜃 = f ,
2vs d
substituting it to G, then
[ [ ] ]2
sin 𝜋 ( La∕2Vs) fd
G= (2.39)
𝜋 ( La∕2Vs) fd
and [ ]4
sin[𝜋(La ∕(2Vs ))(fd + mfp )]
G2 (fd + mfp ) = (2.40)
𝜋(La ∕(2Vs ))(fd + mfp )
where La is antenna aperture size, V s is the moving velocity of the satellite, f d is the
Doppler frequency of the pulse, and f p is the PRF.

2.3.6 Beam Position Design

The beam position includes such parameters as beam width in the range direction,
instantaneous signal bandwidth, and position of the swath. Spaceborne SAR has the
geometrical relationship shown in Figure 2.19. It is shown that
𝜃r Rm
Wg ≈
cos 𝜃
R2m = (Re + H)2 + R2e − 2Re (Re + H) cos γ
sin(𝜃 − α)Re
Rm =
sin α
(Re + H) sin α
𝜃 = arcsin
Re
R sin 𝜃
α = arcsin e
(Re + H)