A method using influence function for evaluating robustness of CFAR detectors

A method for evaluating the robustness of constant false alarm rate (CFAR) detectors is presented; it is based on the powerful methodology of influence function (IF) developed in the literature on robust statistics. The robustness of different kinds of CFAR detectors can be evaluated and compared by calculating the first derivative of the false alarm probability (FAP) and detection probability (DP) at an underlying distribution, which are named IF-FAP and IF-DP. The two IFs are compared among some kinds of CFAR detectors. It is concluded that the robustness of a detector can be asymptotically represented by the IF of the clutter power estimator. Finally, according to a robust measure drawn from the IF of the clutter power estimator, the "most robust" detectors in three ordered-statistic-based groups are presented.