提出了一种基于信号隐含周期性的循环平稳信号盲源分离算法。该方法充分利用了循环平稳信号具有隐含周期性成分的特性,不需要估计信号的循环频率,特别适用于循环平稳信号的分离。本算法在提取循环平稳信号的周期性分量的同时,剔除其中的随机分量,因此,计算机仿真和实验验证了该方法的有效性和循环平稳信号的分离次序。
A novel blind separation algorithm is proposed for cyclostationary signals. The hidden periodicity of cyclostationary signals is used and cyclic frequency in the algorithm is unknown. It is particularly suitable for the separation of cyclostationary signals. In this algorithm,the periodic components of the signal are extracted and the random components is removed,therefore,the effectiveness of the method and separation orders of cyclostationary signals are successfully verified by computer simulations and experiments.