线性检波下CA-CFAR的两种近似方法性能分析

Analysis of Performances of Two Approximate Methods for CA-CFAR Under Linear Detection

由于瑞利分布情况下难以得到CA-CFAR性能分析解析式,通常采用多个2概率密度的混合近似瑞利分布和基于中心极限定理的CA-CFAR两种近似方法分析其检测性能.为此对上述两种近似方法近似程度问题进行了研究.首先基于平均判决阈值(ADT)的定义导出了两种近似方法相对CFAR损失比解析式,然后对两种近似方法与理论CA-CFAR处理之间的相对CFAR损失进行了详细地仿真和分析,最后给出了相应的结论.

Since it is hard to get the analytic formula of CA-CFAR performance due to Rayleigh distribution its detection performance is generally analyzed by using two CA-CFAR approximate methods of approximate Rayleigh distribution with multi-Х^2 probability density mixing and the central limit theorem. Therefore, the approximate extent of the above-mentioned two methods was studied. First, based on the definition of ADT, the analytic formula of loss ratio of two approximate methods with respect to CFAR were derived, then the relative CFAR loss between the two approximate methods and the theory of CA-CFAR was simulatively analyzed in detail, finally the corresponding conclusions were drawn.