On a probabilistic approach to detect Noise Radar random transmit waveforms based on a simple circularity test

Abstract

Noise Radars are electromagnetic systems that use random signals for detecting and locating reflecting objects. Besides high performance with respect to the suppression of range ambiguity in the detection of targets and low range sidelobes, they also present an intrinsic property of low probability of interception by other systems, due to the stochastic nature of their transmit waveforms. Traditional methods and equipment are often ineffective to detect the presence of such kind of radars, both in time and frequency, since they generally adopt an ultra-wide bandwidth (UWB) configuration, spreading its power through a broad portion of the spectrum. Within this context, this paper proposes a probabilistic approach based on a statistical property of random vectors, the circularity, to detect the presence of pulses of this nature in the scenario under study.