Adaptive censoring maximum likelihood CFAR detector in Weibull clutter

Abstract

An adaptive censoring maximum likelihood constant false alarm rate detector (ACML-CFAR) based on ordered data variability (ODV) is researched in spiky Weibull clutter environment. ^{1 2} The ACML-CFAR employs ODV-based censoring algorithm to censor several reference samples, and then employs the rest to estimate the parameters of Weibull clutter through maximum likelihood estimator. Censoring efficiency of the ODV-based censoring algorithm in Weibull clutter and interfering targets environments is evaluated. And then the detection performance of ACML-CFAR is evaluated and compared with that of ML-CFAR in Weibull clutter and interfering targets environments through computer simulations. It shows that the detection performance of ACML-CFAR is slightly better than that of ML-CFAR in Weibull clutter. In interfering targets environments, when the number of censoring samples is less than the number of interfering targets, the detection performance of ML-CFAR degrade severely, whereas ACML-CFAR can censor interfering targets adaptively and effectively, and doesn't need any prior information about the number of interfering targets, and therefore performs robustly. It can be used in actual environments that include unknown interfering targets.