摘要
针对超定含噪盲源分离问题,提出了一种基于迭代的白化算法,并将该算法作为白化预处理算法应用于FastICA算法中。该算法可以准确估计每路含噪混合信号中的噪声方差,经过白化后能够完全去除含噪信号中有用信号的相关性。实验仿真结果表明:该算法的抗噪声性能优于基于特征值分解的准白化FastICA算法,尤其是在每路含有不同噪声方差的情况下。
To solve the problem of overdetermined noisy blind sources separation, a whitening algorithm based on iteration is proposed, and it is used as whitening preproeessing in FastICA. The proposed algorithm can accurately estimate the noise variance of each channel noisy mixed signal, and the correlation of useful signals can be completely removed from the noisy signal after quadratic whitening. Simulation results show that the anti-noise performance of the algorithm proposed in this paper is greatly improved over the quasi-whitened FastICA algorithm based on eigen value decomposition, especially when each channel has different noise variance.
出处
《信息工程大学学报》
2016年第6期681-685,共5页
Journal of Information Engineering University
基金
国家自然科学基金资助项目(61401511)
