Adaptive blind source separation of colored signals

A new adaptive blind source separation (BSS) algorithm is presented in this paper. The algorithm uses a special cost function mid only the second order statistical information of the colored signals to perform the pre-whitening transform and orthogonal transform, which avoids the singular value decomposition (SVD) as commonly used in off-line BSS operation and makes two-steps adaptive BSS operation possible. Simulations showed the good separation performance of the algorithm.