基于自动删除算法的分布式CFAR检测算法

A Distributed CFAR Detection Algorithm Based on Automatic Censoring Algorithm

研究分布式恒虚警(CFAR)检测系统在非均匀干扰背景中进行优化检测。针对多传感器分布式恒虚警检测系统在非均匀干扰背景中容易出现检测概率下降或者虚警率提高的问题,提出了一种基于自动删除算法的分布式恒虚警检测算法。算法是一种基于局部检测统计量的分布式CFAR检测算法,充分利用了局部检测器的观测信息,提高了检测性能,同时采用自动删除算法可以自动删除或者接受参考单元的样本信息,无需背景环境的先验知识,达到提高在非均匀背景中的鲁棒性。进行计算机仿真对算法的有效性和鲁棒性进行了验证。

The detection rate of distributed constant false alarm（CFAR） of multisensor in nonhomogeneous background is studied.The multisensor distributed CFAR detector tends to degrade detection probability or increase false alarm rate in nonhomogeneous background,a distributed CFAR algorithm is proposed based on automatic censoring algorithm.This algorithm is a distributed CFAR detection algorithm based on local test statistics in which the information of local observations is sufficiently used,so the performance is improved.It automatically deletes or acceptes samples in reference cells without requiring any priori knowledge about the background environment,so the robustness in nonhomogeneous background is improved.The validity and robustness of this algorithm is tested by computer simulation.