Evaluation of Detection Performance under Employment of the Generalized Detector in Radar Sensor Systems

Abstract

The detection performance of the generalized detector (GD) constructed based on the generalized approach to signal processing in noise is evaluated under homogeneous and non-homogeneous noise. The GD adaptive threshold is derived and defined applying an appropriate noise power estimation using the sliding window technique. The direct close expressions for the GD average probability of detection and probability of false alarm are also derived. Typical constant false alarm rate (CFAR) detectors, namely, the cell averaging CFAR (CA-CFAR) detector, the ordered statistic CFAR (OS-CFAR) detector, the generalized censored mean level (GCML) detector, and the adaptive censored greatest-of CFAR (ACGO-CFAR) detector are compared with the GD by detection performance under both homogenous and non-homogeneous noise conditions, i.e. when the interfering targets are absent or present, respectively. Simulation results demonstrate a superiority of GD in detection performance in comparison with the above mentioned CFAR detectors under both homogeneous and non-homogeneous noise conditions.