分布式MIMO雷达目标定位方法研究

Widely distributed multiple-input multiple output (MIMO) radar systems observe the target from diverse directions and jointly deal with received signals from multiple independent paths. It utilize the spatial diversity and signal diversity to combat the target radar cross section (RCS) scintillation and thus it benefits from its stable target detection capability and high target localization accuracy. Widely distributed MIMO radar is believed to be is a promising radar network development with great potential. However, new technical challenges of information obtainment and processing for distributed MIMO radar need to be dealt with. Combing the theoretical research and the engineering application, this project focuses on two key issues of widely distributed MIMO radar: direct target localization and indirect target localization. For the direct target localization: the deghosting algorithm will be studied for target localization using maximum likelihood method, and the target localization problem under off-grid situation and computational complexity for sparsity-aware target localization will be studied. For the indirect target localization: the closed-form solution of weighted least square (WLS) needs to be researched when SNR is small, and the globally optimal solution using nonlinear optimization will be studied. It is expected that the research results could be able to promote the development and practicality of distributed MIMO radar systems.

分布式MIMO雷达采用空间分集和信号分集技术，能够从不同角度探测目标，在接收端合并多个独立观测路径的回波信号，克服由于目标RCS闪烁导致的雷达性能下降的缺点，提高了目标检测能力和定位精度，是未来雷达网络的发展方向。然而，分布式MIMO雷达在信息获取与处理方面也面临新的技术挑战。结合理论研究和工程应用，本项目拟针对分布式MIMO雷达直接目标定位和间接目标定位两个关键技术开展深入研究。直接定位方面：针对最大似然的目标定位，研究虚假目标消除处理方法；针对基于稀疏重构的目标定位，研究网格失配下的定位问题和降低计算复杂度问题。间接定位方面：针对加权最小二乘定位方法，研究低信噪比下的闭合解获取方法；采用非线性优化理论，进而实现间接定位的全局最优解。项目研究成果将为分布式MIMO雷达系统的实用化奠定理论基础。

分布式MIMO雷达采用空间分集和信号分集技术，能够从不同角度探测目标，在接收端合并多个独立观测路径的回波信号，克服由于目标RCS闪烁导致的雷达性能下降的缺点，提高了目标检测能力和定位精度，是未来雷达网络的发展方向。然而，分布式MIMO雷达在信息获取与处理方面也面临新的技术挑战。结合理论研究和工程应用，本项目针对分布式MIMO雷达协同探测基础理论、正交波形设计与处理、延时与相位参数估计、多目标检测与直接定位和混合型MIMO雷达开展研究，提出了基于相位编码信号和离散频率编码信号的正交波形设计与相应的信号处理方法，提出了互相关法和基于ARMA模型的状态空间表示的参数化方法对延时和相位参数进行估计，针对多目标直接定位问题，结合图像膨胀算法提出了IE-MUSIC和IE-MVDR两种多目标直接定位方法，可在目标数量未知的条件下一次性估计全部目标的位置，将相控阵雷达和分布式MIMO雷达相结合提出了混合型MIMO雷达的理念，利用Stein准则和相对熵的概念，从理论推导和仿真实验两方面分析了混合型MIMO雷达的检测性能，项目研究成果能够为分布式MIMO雷达系统的实用化奠定理论基础。