A Study of the Virtual Microphone Algorithm for ANC System Working in Audio Interference Environment

Abstract

Two kinds of anti-audio interference active noise control (ANC) algorithms are proposed based on a conventional filtered-X least mean squares virtual microphone (FXLMS-VM) algorithm. The proposed filtered-X least mean squares virtual microphone and anti-audio interference (FXLMS-VM-AI) algorithm employs an extra audio path to compensate for the influence of the audio interference phenomenon on the ANC adaptive filter. The proposed time frequency mixed filtered-X least mean squares virtual microphone and anti-audio interference (TFM-FXLMS-VM-AI) algorithm can overcomes this audio interference effect by calculating the coefficients of the ANC adaptive filter in the frequency domain. The time domain acoustic wave field simulation method is adopted to test the performance of the proposed ANC algorithms in an audio interference environment. By analyzing the simulation results, both proposed algorithms show better convergence accuracy in terms of the adaptive filter coefficients and better audio signal performance than the conventional algorithm. The proposed FXLMS-VM-AI algorithm has a faster convergence speed for the ANC adaptive filter than the proposed TFM-FXLMS-VM-AI algorithm. The proposed TFM-FXLMS-VM-AI algorithm has a lower computational complexity than the conventional algorithm and the proposed TFM-FXLMS-VM-AI algorithm.