An LMS Based Blind Source Separation Algorithm Using a Fast Nonlinear Autocorrelation Method

Blind source separation (BSS) is the technique that anyone can separate the latent data from their mixtures without any knowledge about the mixing process, but using some statistical properties of original source signals. In this paper we will use the nonlinear autocorrelation function as an object function to separate the source signals from the mixing signals. Maximization of the object function using the LMS algorithm will be obtained the coefficients of a basic linear filter which separate the source signals. To calculating the performance of proposed algorithm two parameters such as Performance Index (PI) and Signal to Interference Ratio (SIR) will be used. It will be shown that the proposed algorithm gives better results than other method such as Newton method which has been proposed by Shi.