一种没有特征值分解的MUSIC算法  

MUSIC algorithm without eigendecomposition

作　　者：黄磊[1] 吴顺君[1] 张林让[1] 

机构地区：[1]西安电子科技大学雷达信号处理重点实验室,陕西西安710071

出　　处：《系统工程与电子技术》2005年第12期1988-1990,共3页Systems Engineering and Electronics

基　　金：国家自然科学基金资助项目(60102005)

摘　　要：提出一种没有特征值分解的MUSIC算法。对于相干信源情况,该方法的噪声子空间估计只需要多级维纳滤波器的空间平滑的前向递推,不需要估计样本协方差矩阵和对其作特征值分解。从而使得该方法具有小运算量和低复杂度的特点,易于实时处理。仿真结果证明了本方法的有效性。A novel MUSIC algorithm without eigendecomposition is developed. For coherent signals, the proposed method only needs spatially smoothed forward recursions of the multi-stage Wiener filter （MSWF） to obtain the noise subspace, does not involve any estimation of the covariance matrix or its eigendecomposition, thereby indicating that the presented method is computationally attractive. Numerical results are included to illustrate the performance of the proposed method.

关 键 词：波达方向 空间谱 降维 多级维纳滤波器 

分 类 号：TN911.7[电子电信—通信与信息系统]