Analysis and application of the excision CFAR detector

Abstract

The presence of interfering signals in the sample set used by a cell-averaging CFAR detector can cause a drastic degradation in its performance. A detector that alleviates the problem by excising strong signals before the cell-averaging operation is proposed and analysed. This detector suffers almost no degradation in performance, in comparison with a conventional cell-averaging detector, when it operates in an environment of homogeneous noise such as thermal noise with the possible addition of wideb and jamming. The proposed CFAR detector can be used in both radar and communications receivers. The paper presents a mathematical analysis of the operation of the excision CFAR detector resulting in explicit formulas for probabilities of detection and false alarm. An application that includes a binary postdetection integrator for discrimination against randomly occurring pulse interferences is presented, and its performance is evaluated numerically. Procedures for the determination of the detector's parameters are discussed.