Constant False Alarm Rate Processors Based on Range Ordered Statistics

Constant false alarm rate (CFAR) processors are used to detect radar targets in the nonstationary background noise distribution with unknown parameters. In this study, a new CFAR processor based on Range Ordered Statistics (ROS) is proposed. In homogeneous background, for a given probability of false alarm, the probability of detection expression of the ROS-CFAR processor is analytically derived. Also, the detection performance of ROS-CFAR processor is compared with some CFAR processors given in the literature, analytically in homogeneous backgrounds and by simulation in nonhomogeneous (multiple target) backgrounds.