Convergence Analysis of Local Algorithms for Blind Decorrelation

In this paper we analyze and extend a class of adaptive networks for second order blind decorrelation of instantaneous signal mix tures Firstly we compare the performance of the decorrelation neural network employing global knowledge of the adaptive co e cients in with a similar structure whose coe cients are adapted via local output connections in Through statistical analyses the convergence behaviors and stability bounds for the algorithms step sizes are studied and derived Secondly we ana lyze the behaviors of locally adaptive multilayer decorrelation net works and quantify their performances for poorly conditioned sig nal mixtures Thirdly we derive a robust locally adaptive network structure based on a posteriori output signals that remains sta ble for any step size value Finally we present an extension of the locally adaptive network for linear phase temporal and spatial whitening of multichannel signals Simulations verify the analy ses and indicate the usefulness of the locally adaptive networks for decorrelating signals in space and time