A CFAR algorithm for close-spaced targets

A modified automatic censored mean-level detector (MLACMLD) is proposed to achieve robust detection performance when two or more targets fly in a formation. In the noise only environment (without clutter), the noise standard deviation σ estimated from the last sliding window can be used as the threshold because the noise standard deviation in the nearby windows varies little. The current reference cells whose amplitude are greater than T σ (T is a scaling constant) are regarded as the cells generated by the interfering targets and are rejected to calculate σ. It is shown by the results of the simulation that the close-spaced targets can be detected automatically with robust performance by the MACMLD algorithm. The computational requirements are much less than those in the algorithm of automatic censored mean-level detector (ACMLD)