基于相对马氏距离的非均匀杂波抑制方法

由于天基雷达覆盖范围广、探测背景复杂，雷达回波会呈现出明显的非均匀特性，从而导致杂波协方差矩阵（clutter covariance matrix, CCM）估计与实际情况出现偏差，从而恶化天基雷达对空中动目标的检测性能。针对上述问题，提出了一种基于相对马氏距离的非均匀样本抑制方法，该方法首先计算所有样本的广义内积（generalized inner product, GIP），在此基础上选取一个合适的样本作为参考。通过比较场景中所有样本与参考样本的马氏距离，将相对马氏距离大于判定门限的样本作为非均匀样本剔除。用筛选后的样本进行杂波协方差矩阵估计，从而提高天基雷达在复杂环境下对空中动目标的检测性能。理论分析和实测数据结果表明，所提算法能够在非均匀环境下有效检测出空中动目标。

A heterogeneous clutter suppression method based on relative Mahalanobis distance

The radar echoes of space-based radars show obvious inhomogeneous characteristics due to the wide coverage and complex detection background. It leads to the deviation between the clutter covariance matrix (CCM) estimation and the actual situation, which reduces detection performance for air moving targets. To solve the above problem, a heterogeneous sample screening method based on relative Mahalanobis distance is proposed. Firstly, the generalized inner product (GIP) of all samples are calculated, and an appropriate sample is chosen as the reference sample. By comparing the Mahalanobis distance between all samples in the scene and the reference sample, the samples whose relative Mahalanobis distance is greater than the decision threshold are eliminated as non-uniform samples. The processed samples are used to estimate the clutter covariance matrix, so as to improve the detection performance of SBR in complex environment. Theoretical analysis and actual data results show that the proposed algorithm can effectively detect air moving targets in heterogeneous environment.