Imaging and tracking of targets in clutter using differential time-reversal

This paper discusses the performance of two radar imaging techniques for detecting and tracking multiple moving targets in cluttered environments based on differential time-reversal (TR) techniques. The first technique classifies existing targets into stationary versus moving ones based on information provided by multistatic data matrices (MDMs) corresponding to successive pulse radar acquisitions (snapshots). Under certain conditions, the singular value decomposition (SVD) of the time-averaged MDM provides information (number and location) on stationary targets, whereas the SVD of the differential MDM provides similar information on moving targets. The second imaging technique is aimed at providing real-time tracking of the moving targets. This technique exhibits superior clutter rejection and requires minimal processing costs. We show that the tracking technique benefits from statistical stability (self-averaging) provided by ultra-wideband (UWB) TR operation.