现有的杂波谱稀疏恢复空时自适应处理(STAP)均假设不同距离单元的杂波训练样本满足平稳性条件,但实际系统中存在很多非理想因素,如杂波内部运动 (ICM) ,会造成杂波模型失配从而导致杂波协方差矩阵(CCM)估计不准确,进而使得STAP 处理算法的性能急剧下降。文中针对ICM 情况下的非平稳杂波抑制问题,提出了一种变分贝叶斯推断的非平稳杂波抑制STAP 算法。首先,通过一种非平稳杂波模型将多个杂波训练样本之间的非平稳性模型化;然后,引入变分贝叶斯推断方法对杂波空时功率谱进行稀疏恢复;最后,估计CCM 实现非平稳杂波的有效抑制。仿真实验表明,在存在ICM 的非平稳杂波环境情况下,文中方法效果优于已有的稀疏贝叶斯学习STAP 方法。

It is assumed in existing clutter sparse recovery space-time adaptive processing (STAP) that clutter training samples from different range cells satisfy the stationarity condition. However, there are many non-ideal factors in the actual systems, such as the internal clutter motion (ICM) , which will cause clutter model mismatch, resulting in inaccurate estimation of the clutter-plus-noise covariance matrix (CCM), and then the performance of STAP algorithm will be greatly reduced. In this paper, a variational Bayesian inference STAP algorithm for non-stationary clutter suppression is proposed to address the problem of non-stationary clutter suppression under ICM. First, a non-stationary clutter model is used to model the non-stationarity among the clutter training samples. Then, the variational Bayesian inference method is introduced to perform sparse recovery of the clutter space-time power spectrum. Finally, the CCM is estimated to achieve effective suppression of non-stationary clutter. Simulation experiments show that the proposed method is better than the existing sparse Bayesian learning STAP method in the case of non-stationary clutter environment with ICM.

TN959.73

中央高校基本科研业务费中国民航大学资助专项(3122019048)