Modified rank-sum nonparametric CFAR in multiple-target situation

Since the well-known rank-sum (RS) nonparametric constant false alarm rate (CFAR) detector performs robustly in various and complex background clutter, it plays a very important role in radar detection fields. In order to elevate the detection performance of the classical RS nonparametric CFAR in the multiple-target situation, a modified rank-sum nonparametric CFAR based on the mean ratio of the samples in the leading and lagging windows and the smallest-of selection logic (RS-MRSO) is proposed. The analytical expressions of the false alarm rate and the detection probability of the RS-MRSO nonparametric CFAR in a homogeneous background and in the multiple-target situation are derived, and a comparison to the performance of the RS nonparametric CFAR, the cell-averaging (CA) CFAR, the smallest-of (SO) CFAR and the ordered-statistic (OS) CFAR is made. The obtained results show that the RS-MRSO nonparametric CFAR performs similarly to the RS nonparametric CFAR in a homogeneous background, and the detection performance of the RS-MRSO nonparametric CFAR is greatly improved relative to the RS nonparametric CFAR in the multiple-target situation.