An improved subband sigmoid function based feedback Active Noise Cancellation

Active Noise Cancellation (ANC) is an effective way to reduce noise signals that are present along with desired signals needed for signal processing. The noise can occur due to many reasons and at any point of signal processing including signal generation stage, the processing stage or the detection and storage stage. Passive techniques were used traditionally but as they were inefficient at low frequency ranges, they were replaced by active noise cancelling techniques. ANC requires adaptive algorithms to attain an optimized performance. The ANC system also requires very little space and is cost efficient. Since the 1980s, as DSP chips grew in applications a number of adaptive algorithms have been developed. Different algorithms can be used for different applications. Least Mean Square (LMS) algorithm, Recursive Least Square (RLS) algorithm and Filtered-x-LMS (FXLMS) algorithm are some of the developed adaptive algorithms. In this paper, a new enhanced adaptive algorithm for feedback ANC systems has been developed. Simulations taking MATLAB as a platform has been carried out to show that the proposed algorithm gives a better performance than the conventional feed-forward FXLMS algorithm.