Abstract:

In order to improve the performance of variable step-size LMS-type algorithm based on Sigmoid function (SVS-LMS) for case of Gaussian noise and impulsive noise interference,firstly, the method of updating step-size controlled by instantaneous error’s power is replaced by adjusting step-size under the control of time-averaged estimate of error autocorrelation. Then, HB weighted method is introduced into the algorithm to restrain fake peaks of adaptive filter’s coefficient vector and normalized process is introduced to obtain a greater dynamic range of input signal. Simulation results of tracking time-varying delay show that, comparing to SVS-LMS and another refrained SVS-LMS algorithm being of fixed parameters, the algorithm and its HB weighted method can achieve superior MSE performance.