Censored Maximum Likelihood CFAR Detector in Heterogeneous Gamma-distributed Radar Clutter

Abstract

In practical radar signal detection, Constant False Alarm Rate (CFAR) algorithm is of great importance for target detection embedded in a non-stationary clutter. For Gamma distributed clutter with a known shape parameter, maximum likelihood (ML) CFAR detector is investigated in this paper under type II censoring sample case. The proposed decision rule for the CML-CFAR algorithm (Censored Maximum Likelihood CFAR) is provided in terms of the ML estimations of the scale parameter. The detection performances of CML-CFAR detector are compared through Monte-Carlo simulations against those obtained by CA-CFAR (Cell averaging), ML-CFAR and OSCFAR (Order Statistic) algorithms. Several tests show the favorable and robust performance of the CML-CFAR in the presence of interfering targets.