Asymptotically Optimal Rank Test Detection in Long Tailed Clutter

In this paper, the problem of finding optimum rank test signal detection in long tailed clutter is considered. For deriving the detector two assumptions are accepted. First a large number of reference samples are assumed, this assumption is realistic in a two dimensional radar detector. Besides, it is assumed that the signal to noise power ratio (SNR) is enough to achieve a considerable detection probability. As a result of these assumptions, the detector is optimum only asymptotically. The performance of this detector is compared with other commonly used rank test detectors. The results show that between these rank detectors, the performance of the rank sum test detector is very close to that of optimum detector