基于张量分解的互质阵MIMO雷达目标多参数估计方法

Co-prime MIMO Radar Multi-parameter Estimation Based on Tensor Decomposition

该文提出了一种基于双基地互质阵列(CPA)多输入多输出(MIMO)雷达的多目标波离方向角(DOD)、波达方向角(DOA)和多普勒频率估计算法。收发阵列各由两个满足互质结构的稀疏均匀子线阵组成。时域的快拍序列同样由两个互质的稀疏均匀采样构成。算法利用张量因子分解得到分别包含DOD,DOA和多普勒频率信息的3个流形矩阵,再从中构造出具有范德蒙德矩阵结构的虚拟流形矩阵。为了提高估计精度,还提出了一种基于特征值分解的误差抑制算法,并通过旋转不变子空间算法(ESPRIT)求取各目标的3个待估参数。与传统算法相比,该算法通过构造均匀虚拟阵列和虚拟快拍提高参数估计性能,且不会产生模糊,避免了谱峰搜索和额外的配对过程。仿真实验验证了该算法有效性。

A novel algorithm for estimation of Direction Of Departure（DOD）, Direction Of Arrival（DOA）, and Doppler frequency based on bistatic MIMO radar with Co-Prime Array（CPA） is presented. The transmit and receive arrays are both composed of a pair of sparse uniform subarrays. Similarly, a pair of snapshot sequences with co-prime intervals constitutes the sampling of temporal. Three manifold matrices which contain multi-targets＇ DODs, DOAs and Doppler frequencies respectively are estimated through tensor decomposition. From which a group of Vandermonde matrices of virtual manifold are constructed. To improve the estimation accuracy, an error depressing algorithm based on eigenvalue decomposition is proposed. Finally, the above three parameters are estimated by an Estimation of Signal Parameters via Rotation Invariant Techniques（ESPRIT） algorithm. The proposed algorithm offers better performance through virtual array and virtual snapshot without parameter ambiguous. It requires neither peak searching nor pairing processes, and the simulation results are presented to verify the effectiveness of the proposed algorithm.