Performance analysis of ICA algorithm for Blind Source Separation

Abstract

Blind source separation is the task of isolating original source signal from mixed environment with or without knowing the source signal information. This paper focus on Blind source separation problem with two speech signal. each separation process requires specific pre-processing steps. In this proposed work different algorithms were analysed and from those algorithms, Independent component analysis(ICA), Non-negative matrix factorization(NMF), and principal component analysis (PCA) algorithms are implemented with taking two speech signals as input signals the result is compared with the hardware resources utilised for FPGA is explicated in this paper. The result shows that ICA has good scope in separation of mixed signals.