A multi-channel feedforward ANC system using a novel FXLMS algorithm in solving classroom aviation-noise problems

Abstract

Adaptive or Active Noise Control (ANC) system is now widely used in many applications. In this paper, a novel multi-channel feedforward ANC system is presented to lower the level of the aviation noises which occurred in the KMITL University's classrooms. A system is simulated in MATLAB/Simulink environment. The proposed multichannel feedforward ANC system configurations including related equations are described in this paper. The modified Filtered-X Least-Mean-Square (FXLMS) method is used for updating the weight vector in ANC control block diagram in order to achieve the optimal weight vector resulting in minimal errors. How to find bounds on the step size and how to estimate the rate of convergence as well as the steady-state errors of the proposed system with FXLMS algorithm are also elaborated. The secondary-path effects are considered and compensated in this research. The simulation results are presented in terms of noise attenuation capabilities and the Mean-Square-Error (MSE) convergence time. Simulation results and conclusions presented in this paper will be used for further analysis to improve the system performance of the proposed ANC methodology before implementing in the actual classroom.