Simplified Plant Estimation for Multichannel Active Noise Control

This paper describes novel plant identification and adaptive noise control algorithms that exploit a specific spatio-temporal structure of the multichannel secondary propagation path in simple acoustic environments. A gradient-based plant identification method is proposed that significantly reduces the number of parameters within the secondary path model. We then propose a hybrid adjoint/filtered-X LMS adaptive algorithm that reduces the overall error feedback delay of the adjoint LMS algorithm by taking advantage of the separable structure of the proposed plant model. Simulations verify the accurate estimation capabilities of the proposed methods.