A variable step-size adaptive noise canceller using signal to noise ratio as the controlling factor

Abstract

This paper introduces an adaptive noise canceller (ANC) using a proposed variable step size least mean-square (LMS) algorithm. The step size varies between two hard limits based on a predetermined nonlinear decreasing function of signal to noise ratio (SNR) estimated at every iteration of the algorithm. The performance of the proposed algorithm is studied for different power levels of both stationary and nonstationary Gaussian noise added to the original speech. Compared with other several variable step size algorithms, computer simulations show performance superiority of the proposed algorithm in decreasing excess mean square error (EMSE) in both stationary and nonstationary noise environments. Simulations of the proposed method also show substantial improvements in decreasing misadjustment and reverberation.