麦克风阵列中基于零陷波束形成的声反馈抑制

在教室和会场环境中,通常使用自适应控制方法进行声反馈抑制,但由于扬声器输出信号和麦克风输入信号之间存在强相关性,利用自适应控制方法进行声反馈抑制容易产生偏心估计问题。针对这一问题,提出在麦克风阵列中利用零陷波束形成实现声反馈抑制,该方法通过在扬声器方向上形成零陷,避免麦克风阵列采集扬声器的输出信号,达到声反馈抑制的目的。考虑到实际环境中噪声对声反馈抑制的效果有影响,利用加权最小二乘法迭代更新波束形成器系数。仿真结果显示,在有噪声的情况下,与最小二乘法、最小-最大法相比,加权最小二乘法对波束形成器反馈抑制性能的提升更有益。

Null-steering Beamforming for Acoustic Feedback Cancellation for Microphone Array

In classroom and conference environments, adaptive feedback control(AFC) method was commonly used for acoustic feedback cancellation. However, AFC method was typically biased due to the strong correlation between the loud-speaker signal and the incoming signal. In order to avoid the bias, this paper proposed to use the null-steering beamforming to canceled the acoustic feedback. This method canceled the acoustic feedback by steer a spatial null in the direction of the loudspeaker, avoided the acquisition of the loudspeaker''s output signal by the microphone array. To decreased the noise interference in the real environment, this paper proposed to updated the beamformer coefficients by the weighted least squares (WLS). In the presence of noise, the simulation results show that the weighted least squares(WLS) is more beneficial to the improvement of the beamformer feedback suppression performance than the least square(LS) and the min-max(MM).