Noise Cancellation Using Adaptive Filtering Through LMS, NLMS and RLS Algorithms

In this paper, a simulation scheme to simulate adaptive filters using LMS (Least Mean Square), NLMS (Normalized Least Mean Square) and RLS (Recursive Least Square) for noise cancellation is presented. The main idea of noise cancellation is to obtain a noise-free signal, by estimating the noise signal and removing it from input signal. The adaptive noise cancellation estimates the speech signal corrupted by noise or interference. In this method, two different input signals are used. One contains the speech signal (primary input) and the other contains noise signal (reference input). In order to obtain the estimated signal, the reference input is subtracted from the primary input signal after it is adaptively filtered. By using LMS, NLMS and RLS noise canceller algorithms the desired signal, which is corrupted by additive noise, can be recovered. These three adaptive noise cancellers are being used to effectively improve the signal to noise ratio. In this paper, these three algorithms are implemented in MATLAB. Three adaptive filters with white noise and noisy tone signal are simulated. Results indicate that RLS algorithm has the best performance in compare to LMS and NLMS algorithms, but in cost of large computational complexity

کلیدواژه ها:

Noise Cancellation ، LMS Algorithm ، NLMS Algorithm ، RLS Algorithm ، Adaptive Filter