New nonparametric detectors under K-distributed sea clutter in radar applications

Based on modified rank test, this paper proposes two new nonparametric detectors by using the K-distributed sea clutter model. These two detectors are Modified-Rank-based Generalized Sign detector (MRGS) and Modified-Rank-based Mann-Whitney detector (MRMW). The new detectors are achieved by designing the inverse-normal-score functions of the rank statistics. They are CFAR (Constant False Alarm Rate) for the data with continuous CDF (Cumulative Distribution Function) and with PDF (Probability Density Function) bing symmetric about zero. The performance of the new detectors is analyzed under the Swerling II target model by Monte-Carlo simulation. We compare the results of our proposed detectors with two classical nonparametric detectors: Generalized Sign (GS) detector and Mann-Whitney (MW) Detector. The most important conclusion is that the detectability of nonparametric detector is insensitive to the change of the scale parameter b but just depends on the shape parameter v of the K-distributed clutter for the Swerling II target model. Moreover, nonparametric detector especially for the proposed MRMW detector in this paper is quite suitable for low observable targets detection under spiky K-distributed sea clutter.