Legendre Neural Network for nonlinear Active Noise Cancellation with nonlinear secondary path

In this paper, we propose a computationally efficient Legendre Neural Network (LNN) for nonlinear Active Noise Cancellation(NANC). Update algorithms for NANC with linear secondary path (LSP) based on Filtered-x Least Mean Square (FXLMS), Filtered-e Least Mean Square (FELMS) and Recursive Least Square(RLS) are developed. Update algorithm for NANC with nonlinear secondary path (NSP) is also developed which rests upon virtual secondary path concept. Performance of the proposed network and algorithms are validated through extensive computer simulations.