Dual estimators for clutter-map CFAR detection on non-Gaussian background

Abstract

A new clutter-map constant false alarm rate (CFAR) detection algorithm for log-normal distribution is proposed. A combination of two recursive estimators (dual estimators) is considered for the estimation of the mean and standard deviation of the log-normal interference. Compared with the estimation method based on digital exponential filtering technique, the dual estimators converge faster while giving smaller steady state errors. Detection analysis based on the log-t test shows that the use of the dual estimators can improve the detection sensitivity by a few dBs.