Effect of wind noise on active noise control headphones and wind noise suppression in distributed filtered-x least mean squares algorithms

Mingyang Chen,Yijing Chu,Yuezhe Zhao,Feng Niu,Haitao Sun

DOI: https://doi.org/10.1016/j.apacoust.2024.109951

IF: 3.614

2024-03-12

Applied Acoustics

Abstract:Wind noise has detrimental impacts on audio equipment. This paper evaluates the impact of wind noise on the noise control performance of ANC headphones in a variety of windy environments. The findings indicate that the noise reduction performance of headphones degrades as the wind speed increases. To improve the performance of ANC systems in windy environments, the diffusion filtered-x least mean squares (Diff-FxLMS) algorithm has been employed that can effectively suppress the wind noise and disperse the computational burden among nodes across acoustic sensor networks. The theoretical analysis of the Diff-FxLMS algorithm in wind noise environment indicates that an estimation bias has been introduced in the steady-state solution. Computer simulations using recorded wind noise and measured acoustic channels demonstrate that the diffusion algorithm could improve the noise reduction performance in windy environments up to 5 dB approximately.

acoustics