A modified quasi-best weighted order statistics CFAR algorithm with greatest of selection

In order to improve the performance of OSGO or GOSGO, a CFAR algorithm based on quasi best weighted (QBW) order statistics CFAR method is proposed in this paper. The analytic results show that the performance of QBWGO is evidently superior to that of OSGO or GOSGO both in a homogeneous background and in multiple target situations. If the number of interfering targets is 4, the CFAR loss of QBWGO is 4.659, the CFAR loss of OSGO is 5.357, and an improvement of nearly 1 dB by QBWGO is obtained. In the special case of M/sub 1/=M/sub 2/=0, N/sub 1/=N/sub 2/=0, QBWGO reduces to GO, and in the case of M/sub 1/=N/sub 1/=0, QBWGO reduces to MX-CMLD.