A New CFAR Detector Based on the EM Algorithm

This paper proposes a simple Constant False Alarm Rate (CFAR) approach relying on the expectation-maximization algorithm to deal with clutter edges. The possible signal backscattered from a target is modeled as an incoherent pulse train and the newly-introduced CFAR technique is assessed in conjunction with an energy detector. Natural competitors are the well-known Cell-Averaging (CA), Greatest Of (GO), and Smallest Of (SO) CFAR techniques. The performance analysis shows that the proposed approach outperforms the state-of-the-art competitors in terms of false alarms control while maintaining satisfying detection performance.