Enhancement of breathing signal using delayless subband adaptive filter with HPF

Abstract

The paper involves implementation of adaptive noise cancellation using a closed loop delayless subband adaptive filter (SAF) with a high pass filter in the primary microphone path. Adaptive algorithms such as Least Mean Squares (LMS) and Normalized Least Mean Squares (NLMS) have many practical applications since they are simple and less complex. These gradient algorithms such as LMS and NLMS suffer from slow convergence. Recursive least squares (RLS) algorithm is computationally very complex algorithm. Therefore, there is a trade off between complexity and convergence speed. This paper utilizes an algorithm named closed loop delayless SAF with which the computational complexity is reduced and convergence performance is improved. Experimental results show that the utilized algorithm works really well in cancelling the low frequency band noise signal from a wideband signal corrupted with the low frequency noise.