Blind deconvolution based on temporal predictability and PSO.

This paper proposed a novel blind deconvolution algorithm based on PSO to remove the convolution effect for sound signals in communication channel. It used the temporal predictability as the measure of deconvolution and used PSO for optimizing the cost function based on it. Then, the filter coefficient for deconvolution could be got and the signals could be deconvoluted successfully. Simulation results show that the deconvolution property of the algorithm for sound signals is good. The values of correlation coefficient and signal-to-noise ratio between recovered signals and source signals are all high. [ABSTRACT FROM AUTHOR]