IEEE - Aerospace and Electronic Systems - November 2022 - 40

High Fidelity RF Clutter Modeling and Simulation
dataset. The main purpose ofthe dataset described here is to
provide radar researchers with a common dataset to benchmark
their results and compare with existing algorithms.
Along with ground clutter induced by the terrain, we have
also included few clutter discretes in the form of buildings.
This dataset can be used to test radar detection and estimation
algorithms along with CoFAR concepts for radar waveform
design. The data were generated using the signal
model presented in " Stochastic Transfer Function Model "
section.
Given this signal model, CoFAR research can be
broadly classified into two tasks:
Using known channel impulse responses, designing
optimal radar transmit waveforms to maximize target
detection and estimation performance.
Estimating the channel impulse responses from
measured data.
Both these tasks are equally important for practical
CoFAR systems. Ultimately, in a CoFAR system, allocating
resources between these two tasks is a tradeoff. One would
like to allocate resources to channel estimation to obtain
accurate estimates of the channel impulse responses while
not compromising too much on the primary radar objective
oftarget detection and estimation. Some basic ideas for optimal
waveform design using impulse-response-based modeling
have been discussed in the previous section whereas
approaches for channel impulse response estimation have
been discussed in [44] and [45]. It is our hope that the
CoFAR community can develop more advanced waveform
design algorithms, further incorporating practical constraints
on the waveforms. Our dataset can be used to test the performance
of these algorithms for pulse-to-pulse waveform
design as well as CPI-to-CPI waveform design. The dataset
presented here can be used to perform both the waveform
design and channel impulse response estimation tasks. This
dataset contains data cubes as well as the corresponding true
channel impulse responses for two scenarios, each ofwhich
is described in this section.
This challenge dataset was initially distributed at the
inaugural High Fidelity RF Modeling and Simulation Workshop
in August 2020 [53]. The dataset can be accessed/downloaded
by all readers by creating a free trial account at [35].
SCENARIO 1
The first scenario (Figure 30) is supposed to be a beginner
dataset with few targets, ground clutter, and couple of
clutter discretes (like buildings). This scenario involves an
airborne monostatic radar flying over the Pacific Ocean
near the coast of San Diego looking down for ground
moving targets. The data span several coherent processing
intervals as the platform is moving with constant velocity
along the coastline.
40
Figure 30.
Scenario 1 of the challenge dataset with four targets and two clutter
discretes.
Along with the simulated data for this scenario, we
have provided the true channel impulse responses for clutter
and targets. These data span 30 CPIs, 32 spatial channels,
64 pulses, and 2334 range bins. Basic beamforming
and delay-Doppler processing of the data cube gives 30
range-Doppler plots, one for each CPI. Along with the
data, a reference video containing 30 frames is also provided.
Note that we used standard delay-Doppler processing
and also assumed that a fixed waveform was
transmitted. The goal is for the readers to test their own
algorithms and optimally designed waveforms to improve
target detection and estimation performance and obtain
better results than the plots we demonstrate here with
basic signal processing. The details of all the parameters
chosen for this simulation are described in a user guide
provided along with the challenge dataset.
For example, in the sixth CPI, we obtain the plot in
Figure 31. Three targets and two clutter discretes can be
identified from this plot. The other target, which is much
Figure 31.
Range-Doppler plot from the 6th CPI in scenario 1.
IEEE A&E SYSTEMS MAGAZINE
NOVEMBER 2022
IEEE - Aerospace and Electronic Systems - November 2022

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