Noisy blind source separation by nonlinear autocorrelation

When source signals have nonlinear autocorrelation temporal structure, nonlinear autocorrelation has been used as a statistical property for solving blind source separation (BSS) problem (Z. Shi, Z. Jiang, F. Zhou, A fixed-point algorithm for blind source separation with nonlinear autocorrelation, Journal of Computational and Applied Mathematics (2009)). The application of this method is, however, limited to noise-free mixtures, which does not consider the noisy case. Therefore in this paper, we consider the blind separation of the noisy model using the temporal characteristics of sources. An objective function, which combining Gaussian moments to nonlinear autocorrelation is proposed. Maximizing this objective function, we present a blind source separation algorithm for noisy mixtures. Simulations show the better performance of the proposed algorithm.