Efficient Blind Signal Separation In Dynamic Environment Employing Novel Ica Technique

In this paper, a new efficient Independent Component Analysis (ICA) algorithm is proposed for Blind Source Separation (BSS) in dynamic environments. The algorithm is named ICA with Time-Varying Convergence Factor (ICA/TVCF). The gradient-based technique optimizes the convergence factor for each iteration during the adaptation process. By eliminating the matrix inversion operation in the weight update equation, the new algorithm is computationally highly efficient compared with the previously proposed ICA with Optimum Block Adaptation (OBA/ ICA), especially for high dimensional systems. Moreover, the new technique is applicable in the cases where the dimensionality of the system is larger than the processing block size. Computer simulations are performed for separating two BPSK signals in a diversity wireless receiver adopting ICA/TVCF. The results confirm the effectiveness and the advantages of ICA/TVCF over the well-known fast-ICA and online gradient ICA algorithms.