An Improved VI-CFAR Detector

The modified trimmed mean (MTM) scheme is robust in multiple targets background. It takes the sum of two local estimations of leading and lagging reference windows which apply TM method as the total clutter power estimation. In order to improve the robust performance of the variability index (VI) scheme in multiple targets situation where the interfering targets are present in both the halves of the reference window, an improved VI constant false alarm rate (CAFR) detector based on MTM (VIMTM) scheme is proposed. It employs a composite approach based on the cell averaging CFAR (CA-CFAR), greatest-of CFAR (GO-CFAR), and modified trimmed mean CFAR (MTM-CFAR) detectors. The analytic expression of its threshold multiplier factor TMTM is also derived. Under SwerllingⅡassumption, the performance of the VIMTM detector in different clutter environments is compared with the VI and OSVI (order statistic VI) through Monte-Carlo simulation and analysis. Results show that the VIMTM detector has better detection performance and robustness in homogeneous environment and multiple targets background. In clutter edge environment, the false alarm performance of the VIMTM detector is almost consistent with the VI and OSVI. In addition, the sample sorting time of VIMTM detector is less than that of OSVI.