A generalized smallest of selection CFAR algorithm [radar signal processing]

A generalized smallest of selection CFAR (constant false alarm rate detection) algorithm (TMSO), based on the trimmed mean (TM) method, is proposed in this paper. It takes the smallest local estimation of either the leading or lagging window, which applies the trimmed mean method as a noise power estimation to set an adaptive threshold. Thus, the smallest of selection (SO), the generalized ordered statistic smallest of (GOSSO), or the ordered statistic smallest of (OSSO), is the special case of TMSO. It is shown that the performance of TMSO in homogeneous background and in multiple target situations is improved over that of GOSSO or OSSO.