基于功率中值和归一化采样协方差矩阵的自适应匹配滤波检测器

Adaptive Matched Filter Detector Based on Power Median and Normalized Sample Covariance Matrix

在非均匀海杂波环境中,参考单元中的异常单元限制了采样协方差矩阵(SCM)的估计性能,从而影响了传统自适应匹配滤波(AMF)检测器的检测性能。而考虑删除异常单元的方法在参考单元数目有限时可能导致矩阵的奇异性。鉴于此,在保持参考单元数目不变条件下,该文设计一种基于功率中值与归一化采样协方差矩阵(MNSCM)的估计方法,并将其与匹配滤波器(MF)相结合,构造一种新型的自适应匹配滤波检测器。与传统的自适应匹配滤波相比,该文设计的检测器在实测和仿真海杂波数据条件下均具有明显的性能优势。

In nonhomogeneous sea clutter, abnormal cells included reference cells constrain the performance of the Sample Covariance Matrix （SCM）, and then influence the detection performance of the traditional Adaptive Matched Filter （AMF） detector, while censoring abnormal ceils may cause singularity of the covariance matrix in the case of limited reference cells. Without changing number of the reference cells, this paper devises the median and normalized covariance matrix estimator and uses in the detection scheme of the AMF. Compared with the traditional AMF, the newly devised AMF obtains better performance in both measured and simulated clutter.