非高斯杂波下基于子空间的距离扩展目标检测器

Range-Spread Target Detector Based on Subspace in Non-Gaussian Clutter

在辅助数据缺失的非高斯杂波背景下,采用两步法设计策略研究了距离扩展目标检测方法.首先,在杂波纹理分量已知的条件下,对待检测数据进行高斯化,利用高斯背景下杂波协方差矩阵和目标散射点幅度的合适估计,建立检验统计量.其次,利用待检测数据在信号子空间正交补上的正交投影,估计杂波纹理分量,提出了基于子空间的距离扩展目标自适应检测器,并证明了其对杂波纹理分量的恒虚警率(CFAR,Constant False Alarm Rate)特性.仿真结果表明,在典型非高斯背景下,所提检测器的CFAR特性和检测性能均优于对比检测器;另外,阵元数、目标距离单元数或杂波尖峰的增加,能不同程度改善检测性能.

In non-Gaussian clutter environment,the range-spread target detection without secondary data is addressed, by utilizing two-step design procedure. Firstly, with known clutter texture component, the data under test is transformed into the Gaussian ones. And then the test statistic is derived by using the appropriate estimates of unknown clutter covariance ma- trix and target scatterer amplitudes. Secondly, by orthogonally projecting the data under test onto the orthogonal complement of signal subspace, the unknown clutter texture components are estimated. Consequently, the range-spread target detector based on subspace is proposed,and its constant false alarm rate （CFAR） property is also analyzed. It is showed that, under the classical non-Gaussian background, the proposed detector outperforms the compared ones, in terms of CFAR property and detection performance. In addition, its detection performance enhances to different extents, as the number of sensors, the number of target range cells or the clutter spikiness increases.