基于压缩感知的MIMO雷达低空目标测向方法研究

In order to overcome the effect of changeable clutter and complicated multipath echo for low-altitude targets direction finding, the project puts forward the academic thought that the problem can be solved with MIMO radar based on compressed sensing. Firstly, according to the spatial and frequency diversity of MIMO radar, the algorithm about restraining the clutter and multipath effect is derived, also with the spatial and frequency difference. Then, in view of the sharp drop of direction finding precision caused by the space-time distribution, peak and long tail characteristics of low-altitude clutter, the algorithms about clutter distribution lower order moment and sparse reconstruction modeling error are designed. Besides, against the problem of echo signal attenuation and coherent caused by multipath effect, it is realized from the sparse reconstruction and radar diversity to reduce multipath vector effect and use multipath echo energy, respectively. Finally, based on the relationship of radar clutter and multipath echo, the sparse reconstruction model for improving the direction finding accuracy is set up, which provides theoretical basis and technical support for promoting the development of low-altitude target direction finding technology.

针对低空目标测向主要受制于低空环境杂波和多径回波影响实际，本项目提出在压缩感知框架下利用MIMO雷达解决低空目标测向这一学术思想。围绕这一思想，利用不同体制MIMO雷达空间、频率分集特性系统研究杂波与多径的空域、频域差异及相关抑制算法；针对低空杂波的空时分布、尖峰和长拖尾特征引起的测向精度急剧下降问题，从杂波分布的分数低阶矩、重构算法的误差值两个角度进行稀疏重构建模；针对多径效应引起的回波信号衰减和相干问题，采用稀疏集构建、雷达分集来降低多径矢量影响和利用多径回波能量；最终建立多径和杂波复合条件下的稀疏重构模型以实现高精度测向，为促进低空目标测向技术的发展提供理论依据和技术支撑。

低空、超低空突防已成为现代战争中夺取制空权的基本战术手段。针对低空目标测向受制于低空环境杂波和多径效应影响实际，本项目在压缩感知框架下开展MIMO雷达低空目标测向算法研究，为低空目标探测技术的工程化应用提供技术支持。项目的研究成果主要包括：.1. 结合低空目标回波信号模型，基于正交频分复用（OFDM）MIMO雷达参差补偿原理抑制多径效应引起的信号能量剧烈起伏，从多径能量优化视角提出基于熵理论的MIMO雷达检测算法，并建立两步贝叶斯广义似然比检测器以抑制非高斯杂波影响，为后续参数估计奠定理论基础。.2. 针对低空杂波非高斯特性，提出基于对角重加载的MIMO雷达DOA估计算法以抑制非均匀杂波对信号协方差矩阵主对角线元素的扰动影响；针对高斯色噪声（杂波）与有限快拍数造成的测向精度下降问题，提出基于信号子空间的正交投影校正测向算法；针对低空杂波下基于二阶矩或高阶矩的传统测向算法失效问题，提出基于快速稀疏贝叶斯学习的嵌套MIMO雷达目标测向方法。.3. 针对低空多径对消问题，提出基于多径回波能量利用的MIMO雷达低空目标DOA估计算法，利用混合范数优化发射多径矢量，利用峰值优选算法优化接收多径矢量；针对低空多径相干问题，开展基于空间平滑稀疏重构的抑制算法研究，算法具有较高的估计精度和较低的复杂度。.4. 创新性的开展稀疏阵列MIMO雷达低空目标测向算法研究，针对传统稀疏MIMO雷达自由度和互耦不能兼顾问题，通过利用互质扩展因子调整发射和接收阵列的阵元间距建立灵活MIMO雷达结构，并针对传统压缩感知算法混合目标DOA估计引起的高复杂度问题提出一种降复杂度两步压缩感知算法。