Effective adaptive noise canceller design using normalized LMS

This paper presents the application of Adaptive filters in noise cancellation during various communication processes, where non-stationary signals are transmitted. Adaptive filter estimates the noise signal and by applying the appropriate weights the estimated noise signal is eliminated from the information. For noise cancellation applications most efficient Adaptive filter Algorithms LMS and its normalized form NLMS are used and their comparative analysis is done in form of their output power, error power and SNR. In both of the algorithms concept of negative feedback is utilized in the cancellation of noise from the signal and thus both of these are also called negative feedback algorithms. Implementation and analysis is done by applying different step sizes on different order of filter. Order of filter is taken as 4, 8, 12 and then by changing the values of coefficients the graphical and computational analysis is done. Finally an efficient design using NLMS algorithm is implemented where order is taken as 8 and step size 0.2. This results as a low error power (11.7221 db) and a high value of SNR (1.1445) than that of LMS algorithms.