Gain Scheduling of Auxiliary Noise and Variable Step-Size for Online Acoustic Feedback Cancellation in Narrow-Band Active Noise Control Systems

摘要

In this paper, we investigate active noise control systems, which use auxiliary white Gaussian noise for online feedback path modeling and neutralization. Such systems comprise three adaptive filters: 1) active noise control filter, 2) linear prediction filter, and 3) feedback path modeling and neutralization filter. The injection of white Gaussian noise with large/small variance improves/degrades the convergence of feedback path modeling and neutralization filter but on the other hand degrades/improves the noise-reduction performance of active noise control systems. The focus of this paper is to have an adequate compromise between the convergence of feedback path modeling and neutralization filter and the noise-reduction-performance of active noise control systems. In this paper: 1) A tuning free gain scheduling of white Gaussian noise is used to improve the noise-reduction performance of active noise control systems, and 2) a variable step-size is used to compensate for the decrease in the convergence of feedback path modeling and neutralization filter due to gain scheduling. The proposed idea is explained for a single channel as well as for multichannel active noise control systems. Computer simulations have been carried out to demonstrate the performance of the proposed method for a single channel as well as for multichannel active noise control systems.