复杂探测环境中基于认知理论的阵列SAR地面动目标检测技术研究

Synthetic aperture radar ground moving target indication (SAR/GMTI), which is seen as an extended technology for SAR, plays important role in both military and civilian task. An efficient approach to achieve SAR/GMTI is the use of multi-channel SAR system, which greatly enhances the detection ability for slowly moving targets submerged in strong ground or sea clutter by space-time processing. With the performance requirement for SAR/GMTI system increasing, the further improvement of the detection performance for slowly moving targets in complex detecting environments becomes a key issue. This project will focus the research on the technology of ground moving target detection (GMTD) for array SAR system in complex detecting environment. Signal processing strategy for SAR/GMTD based on the theory of cognitive radar will be proposed to achieve optimum detection for the detecting environment, in which the detector is chosen adaptively according to the cognition of the environment. The main content is summarized as: (1) statistical analysis of moving target detection test statistics in non-Gaussian background; (2) new types of SAR space-time adaptive processing (SAR-STAP) algorithms for complex environments, including the non-Gaussian, nonhomogeneous, and outlier resistant SAR-STAP algorithms; (3) the design of moving target joint detectors; (4) the design of general GMTD signal processing scheme for array SAR based on the cognitive theory. The research results will provide theoretical support as well as technical reserve for the development of new generation of SAR/GMTI system.

合成孔径雷达地面动目标指示（SAR/GMTI）作为SAR技术的扩展，在军事、民用领域均具有重要地位。实现这一功能的有效途径为采用多通道SAR系统，通过空时级联处理提高强大地、海杂波中慢动目标的检测性能。随着人们对SAR/GMTI系统性能要求的不断提升，如何在复杂探测环境中进一步提高慢动目标检测性能成为了一个关键问题。本项目拟对复杂探测环境中阵列SAR地面动目标检测（GMTD）技术展开研究，设计基于认知雷达理论的SAR/GMTD信号处理方案，通过对环境的认知自适应选取最佳检测器。具体研究内容包括：(1)非高斯背景中检验统计量统计特性分析；(2)复杂环境中新型SAR-STAP算法，具体包括非高斯、非均匀、抗干扰三类SAR-STAP算法；(3)动目标联合检测器设计；(4)基于认知理论的阵列SAR/GMTD总体方案设计。研究结果将为新一代SAR/GMTI系统研制提供理论支撑及技术储备。

针对现有合成孔径雷达地面动目标指示（SAR/GMTI）系统性能提升的进一步需求，本项目开展了复杂探测环境中基于认知理论的阵列（多通道）SAR新型地面动目标检测（GMTD）技术研究。主要完成了如下研究内容：(1)研究了高斯、非高斯杂波及噪声背景中GMTD检验统计量的统计特性，定量分析的结果以闭合解、数值积分或蒙特卡洛实验的方法给出；(2)研究了复杂环境中新型SAR-STAP算法，分别提出了基于alpha稳定分布的SAR-STAP算法（非高斯SAR-STAP），基于对角加载的知识辅助SAR-STAP算法（非均匀SAR-STAP），基于动目标剔除及联合检测的SAR-STAP算法（抗干扰SAR-STAP）；(3)提出了幅度-单脉冲和差比，干涉相位-杂波抑制输出功率等多种形式的动目标联合检测器，克服了单一检测方案在特殊环境中性能下降问题；(4)提出了基于认知雷达理论的阵列（多通道）SAR/GMTD总体方案，进一步提高了SAR/GMTI系统在复杂探测环境中的GMTD性能。研究内容紧密结合了我国SAR/GMTI系统的发展需求，同时，也是信号检测、STAP理论的扩展和完善。.本项目相关研究成果包括学术论文、发明专利以及科技奖励。已发表学术论文18篇，其中SCI收录5篇，EI收录11篇，已申请发明专利4项，已获省部级科技奖励1项。上述研究结果将为新一代SAR/GMTI系统研制提供理论支撑及技术储备，同时可推广应用于中多种新体制SAR，如双基SAR、多输入输出（MIMO）SAR的GMTI处理过程中。