未知杂波背景下恒虚警检测门限获取方法

摘要

针对未知杂波背景下的雷达恒虚警检测问题，提出了一种基于核密度估计的 CFAR(K-CFAR)检测方法。该方法利用核密度估计方法处理参考单元数据，实时准确地估计杂波背景的概率密度函数，进而根据分布函数与上分位点之间的关系确定检测门限。该方法不仅成功地解决了未知杂波背景下检测性能下降、检测器失去恒虚警特性的问题，而且相比已有的单样本非参量 CFAR 检测器，检测性能大幅提升。且对各种未知背景的杂波具有通用性的特点。最后，利用计算机仿真数据验证了文中所提方法的有效性。%In view of radar detection under unknown clutter background,we propose a new CFAR detec-tor based on kernel density estimation(K-CFAR).Firstly,the K-CFAR exploits the kernel density estima-tion to estimate the probability density function(PDF)of the clutter background.Then,in a given constant false alarm rate,we calculate the detection threshold based on the relationship of distribution function and points on the site.This method not only successfully solves the problem of a lower probability of target de-tection and a higher probability of false alarm in unknown clutter background,but also improves the detection performance in comparison with the existing single sample parameter CFAR detectors.And the K-CFAR has the generality to all sorts of unknown clutter background.Finally,we verify the effectiveness of K-CFAR de-tector using the computer simulation data.