IEEE - Aerospace and Electronic Systems - November 2022 - 41

Gogineni et al.
Figure 32.
Range-Doppler plot from the 26th CPI in scenario 1.
Figure 34.
Range-Doppler plot from the 26th CPI in scenario 2.
SCENARIO 2
Figure 33.
Cluster of buildings included as part of scenario 2.
weaker and farther away from the other three targets is not
visible in this plot. Also, from this plot, we can clearly
observe littoral nature of the simulated environment as we
can see regions of water within the ground clutter as water
has weaker reflectivity compared to land. As the radar
platform moves along and drags its beam along, we present
the range-Doppler plot for the 26th CPI in Figure 32.
Now, returns from the weaker target are also picked up by
the radar. The other targets are still visible in the rangeDoppler
plot, but they have moved around from their positions
in the previous plot (Figure 31). Since the dataset
also includes the true channel impulse responses corresponding
to multiple pulses and CPIs, this dataset can be
used to test fully adaptive radar optimal waveform design
algorithms where waveforms are changed from pulse to
pulse or from CPI to CPI. From the provided true channel
impulse responses, data cubes can be generated for any
transmit waveform using the model described in
" Stochastic Transfer Function Model " section.
NOVEMBER 2022
In scenario 1, while we had ground clutter and couple of
strong clutter discretes, they were spaced relatively far
from the targets of interest. Hence, the detection and estimation
oftargets is not so difficult. Now, we move along to
a more challenging dataset. In addition to the targets and
strong clutter discretes present in scenario 1, this scenario
contains several (150) clutter discretes in the form ofsmall
buildings (30m 30m 6m) arranged in a cluster very
close to Target No. 1. This makes this scenario more challenging
than scenario 1. The target locations and radar
parameters remain the same as scenario 1. 150 buildings
were added next to Target No. 1 arranged in a 50 3 grid.
This is indeed a region populated by buildings as can be
seen in satellite images (see Figure 33). While all the buildings
in this simulation were approximated to be ofthe same
size, in more advanced datasets, each building can be modeled
to be ofthe exact shape and size as in reality.
In Figure 34, we have the range-Doppler plot for the
26th CPI in scenario 2. We observe that the cluster ofbuildings
is much stronger compared to Target 1 and the target is
not clearly distinguishable from the cluster of clutter discretes.
Note that we have added this cluster ofbuildings all
along the road on which this target was moving. This is a
challenging scenario where CoFAR techniques for waveform
design are needed to suppress the dominating clutter
discretes. It is our hope that users ofthis dataset will devise
advanced CoFAR techniques that can mitigate the effects
of these clutter discretes using adaptive signal processing
techniques and optimal waveform design.
CONCLUDING REMARKS
In this article, we have reviewed advanced modeling and
simulation techniques for modeling ground clutter using a
IEEE A&E SYSTEMS MAGAZINE
41
IEEE - Aerospace and Electronic Systems - November 2022

Table of Contents for the Digital Edition of IEEE - Aerospace and Electronic Systems - November 2022
