一种快速DOA估计算法

A Fast Algorithm for DOA Estimation

MUSIC算法需要将天线阵列接收数据的协方差矩阵进行特征分解,并在全空域进行谱峰搜索。该算法具有很高的分辨力、估计精度及稳定性,但是运算量巨大,难以实时实现。通过对等距线阵特点及MUSIC算法的研究,提出了一种无需特征分解和在全空域进行谱峰搜索的快速算法,算法采取降维处理的方法快速估计信号子空间,然后根据基于阵列一次快拍的FFT算法粗略估计的局域信号空间进行谱峰搜索,从而有效降低了算法的计算量,理论分析和计算机仿真结果证明了该算法的有效性。

MUSIC algorithm requires the eigendecomposition of covariance matrix, and makes spectral peak searching on a whole spectrum range. It is famous for its statistic efficiency, high resolution and good accuracy, but the computational complexity limits its real-time application. In the study on ULA （Uniform Linear Array） and MUSIC algorithms, a fast algorithm without eigendecomposition and spectral peak searching on a whole spectrum range has been proposed. As the dimension-reduction method is employed, the signal subspace can be obtained fast, and then the spectral peak searching is implemented in several local subspaces based on approximate estimation of DOA with FFT algorithm. It can reduce the computational complexity. The theoretical analysis and simulation results show that the algorithm is effective.