IEEE Aerospace and Electronic Systems Magazine - June 2020 - 47

Credit: Image licensed by Ingram Publishing

savings and reducing the computational burden of the
adaptive algorithm. We accomplish this while considering
the practical waveform design constraints of limited bandwidth and constant modulus. The negative impact of the
waveform design on the ambiguity function is illustrated,
but solutions are not addressed.
Straightforward models of the PU transmissions and
target responses are utilized to ensure adequate reproduction of the signals when implementing the framework on
an experimental hardware platform. This provides for
repeatability and consistency between the simulation and
over-the-air results.
The scenario described in this work relies on the application of waveform diversity [14]-[17] in two respects. First in
the form of spectral coexistence, aimed at allowing multiple
RF systems to operate in the same geographical locality and
spectral region. Second in the design of transmit waveforms
matched to the target response to achieve an increase in
SNR. Haykin et al. [18] recognize the connection between
the PAC found in cognitive radar and optimal waveform
design based on matching the waveform to some characteristic of the sensed environment on an on-going basis. This has
become a central theme explored in cognitive radar research.
Waveform design has attracted interest over a long
period. Bell [19] considers waveform and filter design for
optimum detection of extended targets, and contrasts the
result with a mutual information design scheme for information extraction. Setlur et al. [20] expand on the mutual information approach taken by Bell to create a two transmission
epoch design process, which results in an improved distribution of the available energy. In [21], the authors address the
problem of waveform design in the context of a spectrally
crowded environment. Selesnick et al. [22], [23] describe
iterative algorithms for creating notched and multiple
notched chirp-like waveforms allowing transmissions on
selected frequencies to be controlled. Spectral notching in
the FM noise radar in the presence of narrowband noise is
analyzed in [24], and experimentally demonstrated in [25].
Aubry et al. [26] employ an optimization-based approach to
waveform design in the spectral coexistence context, the
perception element of the cognitive process relying on
JUNE 2020

external environment knowledge provided by a radio environment map (REM). The authors provide a useful summary
of the merits and drawbacks of various optimization-based
waveform design approaches. Huang et al. [27] introduce
the notion of bandwidth quality to the coexistence problem,
again using a REM to provide the cognitive sensing element
of the architecture.
The performance cost of introducing spectrally disjoint radar waveforms is considered in [28], and in [29],
the impact of spectral notching on beamforming performance in phased array architectures is assessed.
The principles of waveform design for target-matched
illumination (TMI) are described by Gjessing [30]. The
authors in [31] derive performance bounds for the optimization of transmit waveform, channel, target, and receiver.
In [32], specifically designed radar waveforms are compared
with LFM waveforms for target detection and discrimination. Orthogonal frequency division multiplexing waveform
design is examined in [33] for adapting to an extended target
response and nonstationary interference, which maximizes
the mutual information between the target response and the
received signal. The joint implementation of spectral coexistence with TMI is investigated in [34] for the maximization
of target SINR while adhering to external spectral costraints,
and the authors in [35] undertake the design of transmit
waveforms by considering the transmit signal-target
mutual information in a spectral coexistence context. The
use of TMI, in addition to being applied to improve
target detection, is also applied to the target identification
objective [36]-[39].
Aubry et al. [40] extend the scenario described in [21]
to include signal dependent interference. Zhang et al. [41]
consider the joint transmit waveform and associated receive
filter design in signal-dependent interference, employing an
iterative algorithm to maximize the target SINR. The development of a practical testbed is described in [42] and used
to experimentally demonstrate an improvement in the signal-to-clutter ratio given an a priori estimate of target
response; in addition, interference avoidance is addressed
given a priori environmental spectral knowledge. In [43]
and [44], knowledge of the target SNR and mutual

IEEE A&E SYSTEMS MAGAZINE

IEEE Aerospace and Electronic Systems Magazine - June 2020

Table of Contents for the Digital Edition of IEEE Aerospace and Electronic Systems Magazine - June 2020