Blind Source Separation of Temporal Correlated Signals and its FPGA Implementation

摘要

In this paper, we present a new framework for blind source separation (BSS). The difference between proposed method and original blind source separation method is that the source separation is performed in the residual level. We discuss two types of BSS problem: instantaneous BSS and convolutive BSS. The cost function is derived by simplifying the mutual information of residual signals for both problems. And then we develop efficient learning algorithms respectively. FPGA implementation of the proposed algorithm based on Signal-Master platform is also discussed. Computer simulations are given to show the separation performance of the proposed algorithm and some comparisons with other algorithms are also provided.