Improved NLMS algorithm with fixed step size and filter length using adaptive weight updation for Acoustic Noise Cancellation

Abstract

An acoustic noise cancellation is an approach used for reduction of additive noise in the speech signal. Normalized Least Mean Square (NLMS) algorithm is the most popular adaptive filter algorithm for noise cancellation. But in NLMS algorithm, selection of step size and filter length of adaptive filter for different type of noise with different noise level (dB) that gives maximum SNR is difficult. This needs various trials of step size and filter length to get optimum solution. This paper proposes a solution to this problem. Proposed algorithm uses constant step size, constant filter length and a ratio of energy spectral density (ESD) of speech and noise for updating filter weights. Depending on the ratio, weights are adjusted automatically. Compared to LMS and NLMS algorithm, proposed NLMS algorithm exhibits better performance in terms of signal to noise ratio (SNR), Mean Square Error (MSE) and convergence time. The proposed algorithm is validated by extensive experimental analysis and simulation. Results show that proposed algorithm outperforms existing LMS and NLMS algorithms.