Comparative Assessment of Ica Methods for Blind Source Separation of Instantaneous Mixtures

The paper presents a comparative assessment of Blind Source Separation (BSS) methods for instantaneous mixtures based on namely generalized eigen-value decomposition, geometrical concepts, differential of mutual information and Kalman filtering applied to Nonlinear Principal Component Analysis (Nonlinear PCA). The methods highlight the independence concept underlying Independent Component Analysis (ICA). The methods have been tested on instantaneous mixtures of synthetic periodic signals, monotonous noise from electromechanical systems and speech signals. A comparison among the methods has been made on the basis of separation ability, processing time and accuracy. The quality of output, complexity of algorithms and simplicity (implementation) of the methods are some of the performance measures which have been highlighted with respect to the above signals.