基于杂波稀疏性的机载TS-MIMO雷达空时自适应处理方法研究

The airborne transmit subaperturing multiple-input multiple-output (TS-MIMO) radar, which combines the advantages of the MIMO radar (such as waveform diversity) with the advantages of the phased-array radar (such as coherent processing), opens a new avenue in MIMO radar developments. Compared with airborne phased-array radar, the requirements of independent and identically distributed training samples for airborne TS-MIMO radar increase dramatically due to the multiple extension of spatial degrees of freedom, therefore the performance of conventional space-time adaptive processing (STAP) methods degrade significantly in the real clutter environment. Recently, more research efforts are devoted to improving the performance of the clutter suppression in complex environment by exploiting the sparsity of clutter in space-time plane..With the characteristics of the airborne TS-MIMO radar system being considered, the theoretical and methodological studies will be conducted in this project to obtain some innovative progresses in sparse recovery clutter suppression. First, we study the theory of atom norm compressive sensing, and propose the gridless sparse recovery STAP method to mitigate the off-grid effect. Second, we conduct the joint optimization model of array amplitude/phase error and sparse coefficient, and achieve the real-time self-calibration of atom error to reduce the performance loss brought by array error. Last, we analyze the inherent structure of space-time dictionary, and present the novel sparse recovery STAP method based on tensor decomposition to achieve the significant reduction of computational load. Therefore, the study will provide the theory and key technology support for the development of the airborne TS-MIMO radar.

机载TS-MIMO雷达既有MIMO雷达波形分集优势，又兼具相控阵雷达相干处理得益，是MIMO体制走向实用化的一条重要可行途径。相较于机载相控阵雷达，其空域自由度成倍扩展导致对独立同分布样本需求急剧增长，使得传统STAP在实际环境下性能显著下降。利用杂波在空时平面的稀疏性可显著降低样本需求，进而改善复杂环境下杂波抑制能力，是国内外研究机构正在尝试的重要方向。.本申请结合机载TS-MIMO雷达体制特点，开展理论和方法研究，争取在稀疏恢复杂波抑制处理方面取得创新性进展。首先，拟研究原子范数压缩感知理论，提出无网格稀疏恢复STAP，彻底消除网格失配问题；拟构建阵元幅相误差及稀疏系数联合优化模型，完成原子误差的实时自校正，解决误差导致的性能损失问题；最后，拟分析空时字典内在结构特性，并提出张量分解稀疏恢复STAP，实现运算量大幅降低。本申请可为发展TS-MIMO体制机载雷达提供理论依据与技术支撑。

相较于传统机载相控阵雷达，机载发射子孔径-多输入多输出（TS-MIMO）雷达空域自由度成倍扩展使得自适应处理对均匀训练样本的需求成倍增长，进而导致传统空时自适应处理（STAP）方法在实际非均匀杂波环境下性能显著下降。稀疏恢复技术利用杂波在空时平面的稀疏性可实现小快拍条件下杂波协方差矩阵的精确重构，因此可显著改善复杂杂波环境下的STAP性能。然而，现有稀疏恢复STAP技术面临网格失配和阵列流形失配带来的性能下降难题，同时其巨大运算复杂度问题也亟待解决。.本项目重点研究了适用于机载TS-MIMO雷达的稳健和低复杂度稀疏恢复STAP方法，包括：基于粒子群优化的无网格稀疏恢复STAP方法、基于深度学习的误差稳健稀疏STAP方法、知识辅助块稀疏贝叶斯学习STAP方法和张量分解STAP方法。此外，本项目还研究了其他新体制预警雷达杂波抑制和目标跟踪方法，包括：降维三维距离依赖杂波抑制STAP方法、天基预警雷达低自由度非平稳和距离模糊杂波抑制STAP方法、无人机集群预警探测杂波抑制STAP方法及单重频距离解模糊方法、机载预警雷达海面慢速运动目标跟踪方法等。上述研究形成了一系列原创创新研究成果，包括：发表学术论文22篇，其中在IEEE Trans. AES、IEEE Trans. AC、IEEE GRSL、SP和IET RSN等国际学术期刊发表学术论文12篇；在雷达学报、电子学报、CIE国际雷达年会和IET国际雷达年会等EI源刊及EI收录国际会议发表学术论文10篇；申请中国发明专利11项。本项目的研究不仅可提升机载TS-MIMO雷达在现实非均匀杂波环境中的杂波抑制性能，同时还可为其他新体制预警雷达的研制提供理论和关键技术支撑。