Fast OMP algorithm for 2D angle estimation in MIMO radar
Fast OMP algorithm for 2D angle estimation in MIMO radar
复制标题
DOI：
10.1049/el.2010.3023
复制
发表时间：
2010-03
影响因子：
1.1
通讯作者：
Y. Liu;M. Wu;Shunjun Wu
中科院分区：
文献类型：
--
作者：
Y. Liu;M. Wu;Shunjun Wu
高维稀疏信号通常应重新排列为长一维信号，以便通过正交匹配追踪（OMP）（一种有效的压缩感知算法）来恢复。然而，显然，重新对齐的长信号将导致 OMP 中的大量计算。如果字典中的每个原子都可以表示为两个向量的克罗内克乘积，则可以将该字典分解为两个子字典。利用这一特性，提出了一种针对此类二维稀疏信号的快速 OMP 算法，并将其应用于 MIMO 雷达中的二维角度估计。仿真结果验证了其良好的近似OMP的重建质量，并大大提高了计算效率。
A high-dimensional sparse signal usually should be realigned as a long 1D signal to be recovered by orthogonal matching pursuit (OMP), an efficient algorithm for compressed sensing. Clearly, however, the realigned long signal will result in a large amount of computation in OMP. If each atom in the dictionary can be expressed as the Kronecker product of two vectors, it can possible to decompose this dictionary into two sub-dictionaries. By exploiting this property, a fast OMP algorithm for 2D sparse signals of this kind is presented, and applied to 2D angle estimation in MIMO radar. Simulation results verify its good reconstruction quality approximate to that of OMP and greatly improved computational efficiency.