ACTIVE NOISE CONTROL OVER A LARGE REGION WITH MULTIPLE SPHERICAL MICROPHONE ARRAYS IN WAVE DOMAIN

Abstract

Active Noise Control (ANC) over large regions of interest (ROI) traditionally requires numerous evenly distributed error microphones, which is often impractical and obstructive for human occupants. In this paper, we proposed a wave domain adaptive ANC algorithm using the joint information from multiple error spherical microphone arrays (SMAs) on the boundary of the ROI. By mapping the SMA recordings to the virtual sound field and introducing ℓ1 norm on the virtual plane wave weights, the proposed method can achieve good noise reduction over the entire larger ROI, especially when the number of noise sources is finite and sparse. Our finding presents a significant advancement in spatial ANC, paving the way for more effective and unobtrusive solutions in a range of environments.