Anderson-Darling Test based CFAR Detection

An Anderson-Darling(A-D) test based CFAR detector feasible for multiple interfering targets and clutter edge scenarios is proposed and referred as AD-CFAR, which exploits K-sample A-D hypothesis test technique to censor clutter blocks needed for power estimation. Thereafter, AD test is employed for distribution test of the resultant homogenous blocks to select the proper detection algorithm from strategy library composed by the ready-made cell-average (CA), great-of (GO), Log-t detectors. Theoretical analyses indicate that the proposed AD-CFAR outperforms conventional CFAR detectors in non-homogenous clutter while maintains similar performance in noise-only scenario as CA-CFAR.