循环累积量兼有高阶统计量和循环不稳两者的优点,可抑制任何平稳高斯或非高斯噪声以及非平稳的高斯噪声,文中利用循环累计量的这一性质,提出的基于循环累积量矩阵联合对角化盲源分离算法,对接收的每路数据分别运用矩阵联合对角化法处理,实现对具有相同循环频率源信号有效的分离的目的。仿真试验表明该算法能够有效的提取有用信号。

Abstrct: Cyclic cumulate has both the adavatages of the high order statistics and the cyclostationaryproces, which can inhibt any stationary Gaussian noise and non-Gaussian noise and non-stationary Gaussian noise. According to the characteristics of it, this paper proposed a new blind source separation algorithm based on cyclic cumulate. Deal with by the way of the joint approximate diagonalization matrix to arrive the object of separating the sou source signal with the same cyclic frequency. The simulation results shown that the new algorithm is effectively to recover the desired source signa from the received mixed signals.