Digital Signal Processing Reference
In-Depth Information
a Doppler-rate filter-bank may be used to compensate the unknown Doppler-
rate of the moving target before applying the reference signal to it.
To estimate the Doppler rate, an algorithm based on the Doppler
centroid was proposed in [5]. First, the cross energy between the received
baseband signal
s
0
(
t
) and a reference signal
s
Ref
(
t
,
h
) = exp{
j
2
p
(
h
/2)
t
2
}
is calculated:
T
|
T
/2
-
T
/2
s
0
(
t
,
f
D
R
)
s
Rref
(
t
,
h
)
dt
|
1
|
e
(
h
)
|
=
(7.26)
where
h
is the Doppler rate to be estimated. After removing the Doppler shift
in (7.15), the received baseband signal can be written as
s
0
(
t
,
f
D
R
)=
exp{
j
2
p
(
f
D
R
/2)
t
2
}, where
2
l
R
0
[(
v
-
v
x
)
2
+
v
y
+
R
0
a
y
]
f
D
R
=
and, thus, we have
|
e
(
h
)
|
=
|
K
T
/2
-
T
/2
h
)
2
}
dt
|
exp{
j
p
(
f
D
R
-
(7.27)
|
e
(
h
)
|
If
f
D
R
-
h
= 0,
the
has
a
maximum
cross-energy.
Thus,
|
e
(
h
)
|
will be the estimated Doppler rate. However, to accurately
estimate motion parameters requires a large number of filters.
˜
= ma
h
7.3.1.3 Parameter Estimation for Doppler and Doppler-Rate
The filter-bank approach matches the Doppler and the Doppler-rate sepa-
rately. If the Doppler and the Doppler-rate can be estimated simultaneously,
a better result of target detection and imaging can be expected.
A parameter estimation approach based on the maximum likelihood
method was proposed in [9]. Estimation of motion parameters based on a
sequence of single-look SAR images was described in [10, 11]. A method
called keystone formatting that uses a unique processing kernel of 1D interpo-
lation of deramped phase history was proposed for SAR imaging of moving
targets in [12].
Search WWH ::
Custom Search