Research on a Novel Clutter Map Constant False Alarm Rate Detector Based on Power Transform

Abstract

A power transform-based clutter map constant false alarm rate (CM/PT-CFAR) algorithm is proposed to improve the detection performance to the weak target in multiple persisting targets situations. In the CM/PT-CFAR detector, the radar dataset obtained at each scan is normalized and multiplied by a scale factor, and then fed to the power transform operation. The transformed dataset is divided into two parts with a numerical value of 1 as the boundary. The part exceeding 1 is fed to update the scale factor, while the other is used for updating the detection threshold. Because transcendental integrals are produced in the derivation process, an accurate analytical expression of detection probability for CM/PT-CFAR is non-existent. Hence a third-order Taylor expansion operation is introduced to approximate the result. The detection performance of CM/PT-CFAR under various conditions is evaluated and compared with those of other CM/-CFAR detectors. The advantage of CM/PT-CFAR is the detection for weak targets, especially in complex detection situations. The proposed algorithm remains a relatively stable computation load in different cases, which is beneficial for practical application.