Polarization Diversity Detection and Localization of a Target With Energy Spillover

Abstract

In this paper, we address the problem of polarimetric detection of a target with energy spillover in Gaussian environment. We adopt full polarized channels and use diverse criteria to design detectors, resulting in five new adaptive detectors. Performance evaluation reveals that the derived detectors gain higher probabilities of detection (PD) and better localization abilities, compared to the detectors which only use single-channel or dual-channel polarized data. A knowledge-aided detector is also proposed to give a further boost in detection and localization performances, especially under small number of secondary data. Besides, all the derived detectors guarantee the bounded constant false alarm rate (CFAR) properties. Experiments with real data from IPIX radar validate the superior performance and the practicability of the proposed detectors.