基于块稀疏的频率分集阵MIMO二维参数估计

Research on estimation of two-dimensional parameters in frequency diverse array MIMO based on block sparse

针对导向矢量时变导致频率分集阵列-多输入多输出(MIMO)雷达参数估计误差增大的问题,提出一种基于块稀疏的距离和角度参数联合估计方法.首先,就导向矢量时变性对参数估计性能的影响进行了理论分析;然后,通过公式推导证明了块稀疏理论在频率分集阵MIMO雷达参数联合估计中的适用性,并提出了基于块稀疏的参数联合估计方法,给出了参数估计的具体流程;最后,通过仿真对参数估计性能进行了分析.理论分析及仿真结果表明:基于块稀疏的参数联合估计方法可实现单次快拍下多目标距离和角度参数的联合估计,估计能力优于传统谱估计方法;在单目标和单快拍条件下,所提算法精度与二维ESPRIT和二维MUSIC算法相比,基于稀疏字典估计的参数精度优于两算法,而基于系数向量估计的参数精度介于两算法之间.

The problems of target’s parameters estimation under time-varying steering vectors were studied and the parameter estimation method based on block sparse was proposed.Firstly,the relationship between the time-varying and estimation error was derived theoretically. Then, the specific transformations of the frequency diverse array-multiple input and multiple output(FDAMIMO) steering vector’s structure was conducted,which can prove that the block sparse theory is suitable for the joint estimation of two-dimensional parameters in frequency diverse array MIMO.The estimation process based on block orthogonal matching pursuit was also put forward in this part. Finally, the estimation performance was analyzed by the simulation. Theory analysis and simulation results show that the proposed algorithm can successfully estimate the two-dimensional parameters of multiple targets by single snapshot, which cannot be achieved by traditional spectrum estimation method. Compared with the two-dimensional ESPRIT and MUSIC algorithms under the circumstance of single target and single snapshot, the parameter estimated by sparse dictionary in block sparse has better performance in the estimated accuracy, while the performance of parameter estimated by coefficient vector fall in between with other two algorithms.