IEEE - Aerospace and Electronic Systems - October 2021 - 7

Chan et al.
Figure 2.
LFMCW based on sawtooth modulation. (a) Transmitted signal and return echo based on statics target. (b) Beat frequency of statics target.
mixing process will be additional added with the Doppler
shift frequency caused by the movement. The beat frequency
of the static target can be written as
fb ¼ a
2R
c
:
GBSAR IMAGE FORMATION
For the SAR radar uses LFMCW system, its image formation
algorithm typically utilized the improved frequency
scaling (FS) algorithm or range-Doppler algorithm (RDA).
Among them, the classic range-Doppler (RD) imaging algorithm
is based on the assumption of the stop and go of the
antenna. First, the echo signal is compressed in a range direction,
and transforms the compressed signal into the RD
domain. Interpolation is carried out to eliminate the effect of
range migration [14], [15]. Finally, the focus processing in
the azimuth direction took placed to form a 2-D SAR image.
The RD algorithm can be applied to wideband chirped con(2)
tinuous
wave GBSAR, but the computation load of its interpolation
will increase rapidly as the migration magnitude
become larger. FS algorithm is a SAR imaging algorithm
proposed by Mittermayer et al. [16] in 1999 based on the
conventional chirp scaling algorithm. This algorithm uses a
frequency scale conversion function, which can accurately
correct the distance migration without performing interpolation
operations, and further improve the calculation efficiency
of imaging. It is an imaging algorithm that can
effectively process demodulated signals and is very suitable
for processing FM continuous wave signals.
RDA is the most widely used satellite SAR image formation
algorithm. This algorithm was developed for processing
SEASAT SAR data in 1976-1978 [17], [18]. The basic RDA
implementation is shown in Figure 3. Both the range and azimuth
compression are accomplished via frequency-domain
operation to improve the processing efficiency. RDA can be
implemented via pipelining structure as all the processes
within this algorithm can be performed in simple 1-D operations.
Range cell migration correction is employed in the RD
domain via interpolation technique.
However, the computation load will be increased due
to the interpolation operation. For low squint angle acquisition,
the azimuth phase history typically is a linear frequency
modulated (LFM) signal response. However when
the squint angle increase, its azimuth phase history will no
longer be an LFM-like signal. This effect will greatly
influence the precision of the image and the quality of the
SAR image generated. Therefore to tackle this issue,
enhanced processing such as secondary range compression
[17], [19] is proposed. Finally, the output of the RDA is
typically single look complex (SLC) image.
PRINCIPLE OF GROUND-BASED SAR DIFFERENTIAL
INTERFEROMETRY
Figure 3.
Implementations ofRDA.
OCTOBER 2021
Differential interferometry can be used to measure the
microdisplacement deformation of objects with high precision.
The ground-based InSAR system combines SAR
complex images of the same target area obtained at different
times, and obtains the microvalue of the target
IEEE A&E SYSTEMS MAGAZINE
7

IEEE - Aerospace and Electronic Systems - October 2021

Table of Contents for the Digital Edition of IEEE - Aerospace and Electronic Systems - October 2021