基于实值特征值分解的求根MUSIC算法  

Root MUSIC Algorithm Based on Real-Valued Eigenvalue Decomposition

作　　者：郑春弟[1] 解春维[2] 李有才[1] 

机构地区：[1]海军兵种指挥学院舰艇装备教研室,广州510430 [2]西安科技大学通信与信息工程学院,西安710054

出　　处：《数据采集与处理》2010年第2期154-159,共6页Journal of Data Acquisition and Processing

摘　　要：针对非圆信号DOA估计问题,提出了一种基于实值特征值分解(Eigenvalue decomposition,EVD)的求根MUSIC算法。首先利用非圆信号为实值信号的特点,将阵列上的接收数据及其共轭用欧拉公式转换为实值正弦与余弦数据,然后将正弦与余弦数据进行串联,从而扩展了数据维数。由于采用实值矩阵的EVD,因此在EVD阶段的运算量简化为复值EVD的1/4。根据EVD后获得的信号与噪声子空间的特点,对噪声子空间和导向矩阵进行重构以便于可以使用求根MUSIC算法获取对DOA的估计。仿真实验验证了本文算法的有效性。A computationally efficient real-valued eigenvalue decomposition root MUSIC-like algorithm is proposed to estimate the direction of arrival （DOA） problem of non-circular sources. Firstly, the complex-valued measurements are transformed into real sine and cosine data by using the properties of non-circular sources and Euler equation. Then,by concatenating the sine and cosine data in tandem, a real-valued matrix with doubled dimensions of the row is obtained, thus doubling the number of available sensors. And the computational complexity at the eigenanalysis stage is reduced by the exploiting the eigenvalue decomposition（EVD）of a real-valued covariance matrix. Furthermore ,a complex-valued signal/noise subspace is formed by the exploitation of the real-valued subspace property so as to make it easy to establish a root MUSIC-like algorithm. The algorithm performances are demonstrated by simulation results.

关 键 词：阵列信号处理 子空间类方法 DOA估计 求根MUSIC 

分 类 号：TN911.7[电子电信—通信与信息系统]