IEEE Aerospace and Electronic Systems Magazine - September 2020 - 7

We are on the threshold of being able to develop and field noise radars which match the performance of existing small radars,
such as low-power marine radars and battlefield surveillance radars, but which uniquely meet customer needs to be covert and to
use the spectrum efficiently. Significant obstacles still remain to be overcome, however, before noise radars can cross the
threshold of practical utility.
Perhaps the most serious of the remaining issues is that of the range sidelobe levels. One of the features of any continuouswave radar, such as noise radar, is that it must cope with handling the transmitted signals and the received signals at the same
time. Even for low-power radars, the mean transmitter power will be of the order of 10 dBm. The integration time will be of the
order of 10 ms, determined by the dwell time. This will give a practical noise floor of about -150 dBm, so the radar needs to be
able to handle a dynamic range of the order of 160 dB. If the maximum practical signal bandwidth is 1 GHz then 10 ms integration
time will give a time bandwidth product of 70 dB, leaving a shortfall of 90 dB in the dynamic range. Of course, most of this is provided by the transmitter-receiver isolation and the rest must come from novel design techniques. State-of-the-art systems achieve
transmitter-receiver isolations approaching 70 dB and achieve the remaining 20 dB by waveform design, which is discussed
below, and by signal processing.
We note that in the illustration above, the required isolation is actually independent of the integration time, which
"cancels out" from the calculation. This is an example of the mathematical elegance that often occurs within the design of
noise radars. On the other hand, the random nature of the signals means that the conclusions of the analysis of noise radar
performance are often stubbornly resistant to changes in the radar's design. For example, the random nature of the signals
means that the dynamic range cannot be increased by time gating of the signals, as is done in pulsed radars or by analog processing as can be done in frequency-modulated continuous wave radars. In addition, the dynamic range of CW noise radar can
be improved by modifying the characteristics of the waveform, so that it evolves from (pseudo) random to a quasi-random
signal, retaining the desirable properties of random waveforms but with the ability to "tailor" their properties in a deterministic manner. As an alternative, time-gating may also be used to form pulse-coherent noise radar waveforms, and analog
processing of the returns may be implemented with stepped-frequency noise radar.
Another reason for "tailoring" the peak to mean power ratios of a noise waveform are relatively high (theoretically infinite) so the power added efficiency of the transmitters is low unless the peak levels of the waveform can be controlled.
So important are these issues of tailoring the waveforms and optimizing the dynamic range that these will form the "subtopic"
of a second special issue on noise radar, which will appear in this magazine in the next few months.
The articles in this special issue are as follows.
"Noise Radar-Overview and Recent Developments" by K. Savci, A. G. Stove, F. de Palo, A. Y. Erdogan, G. Galati,
K. A. Lukin, S. Lukin, P. Marques, G. Pavan, and C. Wasserzier, which addresses many of the issues raised in this editorial
and, in addition, discusses the issues which arise in the characterization of the performance of noise radars.
"Detection of Marine Noise Radars with Spectral Kurtosis Method" by N. Cotuk, A. H. Kayran, A. Aytun, and
K. Savci discusses some practical issues associated with the interception of noise radars-an important topic since low
detectability is a key "pull" for noise radar research.
"Sidelobe Mitigation in Noise Radar Using Sparse Signal Processing," by L. de Martin, W. van Rossum, D. Ribeiro, and
L. Anitori looks at a novel way of addressing the dynamic range issue by using ideas and techniques from compressive
sensing to find large targets and suppress their sidelobes, allowing weaker targets also to be seen.
"Real-Time Signal Generator for Noise Radar" by M. Ispir and A. Yildrum addresses the real-time generation of
pseudo-noise signals. For a practical radar, this is often a better done digitally than by using analog techniques. It is
necessary to perform this task in real time because precalculating the signals and downloading them would put an
excessive load on the radar's data busses and this must be performed continuously since the signal is the nonrepetitive.
It is also very important that the covert nature of the signals is not compromised by weaknesses in way they are
generated, and the article also addresses this issue.
We hope you will enjoy the articles in this first special issue and will find them both interesting and useful.
SEPTEMBER 2020

IEEE A&E SYSTEMS MAGAZINE

IEEE Aerospace and Electronic Systems Magazine - September 2020

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