A new rank test nonparametric procedure for radar detection in multiple target situations

This paper deals with the detection performance evaluation of a new nonparametric procedure. The test statistics of the proposed processor have a simple statistical form which is derived from that of the suboptimum rank test procedure [1]. The performance of the new distribution-free detection technique is obtained through Monte-Carlo simulations and compared with that of the suboptimum scheme when detecting nonfluctuating target embedded in Gaussian noise and in the presence of interfering target returns among the reference samples. The new non parametric detector is shown to give a relative improvement over the suboptimum processor. especially, when the background environment contains a numerous-number of extraneous targets.