Performance analysis of a weighted max CFAR processor

In this paper, we study the performances of a novel CFAR in homogeneous environment. This processor consists to take maximum power of cells in the reference windows and multiply it by a coefficient alpha (alpha < 1) to obtain the background level estimate. We assume that the target is embedded in a white Gaussian noise and fluctuate according to a Swirling I model. Closed form expressions of the probabilities of detection and false alarm are determined and the performance of the proposed detector is investigated.