Statistical analysis of a signal separation method based on second-order statistics

This correspondence explores a method for separation of dynamically mixed s ources, which is based on second-order statistics. Here, a statistical anal ysis is given of a generalized version of the original algorithm, The gener alized method includes a weighting matrix, and a result of the statistical analysis is that the best possible weighting is found. In eases where the s ources have similar color, the weighted algorithm significantly improves th e estimates of the mixing parameters. The problem of model validation Is di scussed as well.