The CLEAN type algorithms for radar signal processing

Abstract

A new class of signal processing algorithm based on CLEAN methods primary introduced in radio-astronomy is presented in the paper. The classical radar signal processing is based on match filtering concept, which is optimal in mean square sense in case, when single target echo is detected against white or colour Gaussian noise. Such approach was effective when pulse radar have been widely used. The introduction of pulse-compression technique changes significantly the signal model, but still the match filter have been widely. The introduction of continuous wave radars, and especially noise and passive radars changed dramatically the situation. In such radars all echoes are superimposed and interfere with each other and the simple model no longer fits to that case. The straightforward solution - use of inverse problem mathematical solutions such as solving the set of nonlinear equations to find all echoes in received signal - is usually computationally ineffective and often numerically not stable, so suboptimal methods are being developed to improve detections of weak signals in CW radars. One of possible solution is to use concept of CLEAN technique an remove all strong echoes from received signal. When only weak signals and white noise remains in is possible to use matched filter concept without significant loses of radar sensitivity. In the paper several techniques for radar signal processing utilizing CLEAN concept are shown. © 2008 IEEE.