Automatic dual censoring cell-averaging CFAR detector in non-homogenous environments

Abstract

Constant false alarm rate (CFAR) detectors are designed to regulate the false alarm rate against the noise and clutter background whose mean power is unknown and (or) varying. This is accomplished using local threshold estimate of the background observations. When background observations contain irrelevant samples, censoring is warranted to improve detection performance and to regulate the false alarm rate. In this paper, a new CFAR detector composed of an automatic dual censoring algorithm and cell-averaging CFAR detector is proposed. This new detector is named automatic dual censoring cell-averaging (ADCCA)-CFAR. The proposed detector does not require any prior knowledge about the background observation. It uses fuzzy membership function to determine and censor the unwanted samples in the reference window. The performance of the proposed detectors is evaluated and compared with those of the cell-averaging (CA), order statistic (OS) and automatic censoring cell-averaging based on ordered data variability (ACCA-OVD) CFAR detectors. The simulation results show that the ADCCA-CFAR detector not only provides low CFAR loss in homogenous background but also performs robustly in the presence of interfering targets and clutter edge situations.