Abstract: The spirit of the present invention is to vary the step size parameter in accordance with the error signal and the output acoustic signal, wherein the filter is easy to implement, nonparametric VSS-NLMS algorithm which employs the mean-square error and the estimated system noise power to control the step-size update. The new nonparametric VSS-NLMS algorithm has been shown to perform with fast convergence rate, good tracking, and low mis-adjustment. In comparison with existing VSS-NLMS algorithms, the proposed algorithm has demonstrated consistently superior performance both in convergence and for final error level relative to published algorithms in application on both simulated data and real speech data.

Abstract: The spirit of the present invention is to vary the step size parameter in accordance with the error signal and the output acoustic signal, wherein the filter is easy to implement, nonparametric VSS-NLMS algorithm which employs the mean-square error and the estimated system noise power to control the step-size update. The new nonparametric VSS-NLMS algorithm has been shown to perform with fast convergence rate, good tracking, and low mis-adjustment. In comparison with existing VSS-NLMS algorithms, the proposed algorithm has demonstrated consistently superior performance both in convergence and for final error level relative to published algorithms in application on both simulated data and real speech data.

Abstract: The present invention comprises an adaptive bilinear decision feedback equalizer to be implemented in a receiver to cancel the intersymbol interference. Basically, the decision feedback equalizer has a feed-forward filter, a feedback filter and a bilinear filter. The equalizer coefficients are determined recursively using an RLS or LMS algorithm.