Robust CFAR Detector with Weighted Amplitude Iteration in Nonhomogeneous Sea Clutter

Constant false alarm rate (CFAR) is a desired property for target detection in unknown and nonstationary sea clutter. Analysis of the experimental data shows that gamma distribution is a promising model for sea clutter. A robust CFAR method is proposed for target detection in nonhomogeneous gamma-distributed clutter, using the weighted amplitude iteration of the samples in the reference window as the adaptive threshold. By combining the advantages of cell-averaging CFAR (CA-CFAR), greatest of selection CFAR (GO-CFAR), and ordered statistic CFAR (OS-CFAR), the proposed method shows a similar detection performance as the CA-CFAR in homogenous gamma-distributed environment with a known shape parameter. In a nonhomogeneous environment, the proposed method also works robustly with the appropriate weighting factors, whereas CA-, GO-, and OS-CFAR methods exhibit a serious degradation of the detection probability and an excessive increase in the false alarm rate. The detection performance of the proposed method in gamma-distributed clutter with different shape parameters is also presented by simulation. The superiority of the proposed method, which is applicable to different clutter scenarios with corresponding weighting factors, is investigated and verified by simulations and experimental data.