摘要
在阵列信号处理中,诸如MUSIC和ESPRIT等高分辨率空间谱估计算法都要对阵列输出数据的协方差矩阵进行特征值分解,其计算量较大,不适合实时处理。因此,文中提出MCA(次分量分析)高效迭代算法,用来逼近噪声子空间,该算法无需进行特征值分解,计算过程相对简单,具有自组织特性,算法稳定收敛,适合于神经网络来实现。通过仿真实验证实了所提算法的优良性能及其可实施性。
In the array signal processing, such MUSIC and ESPRIT as high-resolution spatial estimation algorithm need to decompose covariance matrix of the array output data, which involves large amount of computation and is not suitable for real-time processing. Therefore, the MCA(Component Analysis) efficient iterative algorithm is proposed in the paper to approximate the noise subspacc.This algorithm needs no eigenvalue decomposition, is relatively simple in the calculation process, is of self-organizing feature. It is stable in convergence and good for implementation by neural networks. Simulation results have proved the excellent performance and practicability of the proposed algorithm.
出处
《信息安全与通信保密》
2009年第1期111-113,共3页
Information Security and Communications Privacy
关键词
协方差矩阵
实时处理
迭代算法
神经网络
covariance matrix
real-time processing
iterative algorithm
neural network