Performance Studies on the GLRT, LLRT and CFAR Schemes for Detection in the Presence of Compound-Gaussian Clutter and Noise

摘要

The generalized likelihood ratio test (GLRT) formulation and its approximation are recently proposed methods for detecting signals in non-Gaussian clutter and noise. In this paper, the performance of the GLRT, the log-likelihood ratio test (LLRT), and the cell averaging CFAR detectors are studied in detail with the objective of comparing the performance in different situations of radar target detection. The target signal to be detected is considered as non-fluctuating, but unknown in amplitude in the presence of K-Distributed clutter and additive white Gaussian noise (AWGN). Significant differences in performance are observed from experimental studies.