Research Of Acoustic Echo Cancellation Based On Microphone Array

In hands-free voice calls,the appearance of acoustic echo can affect voice quality.Traditional acoustic echo cancellation systems use time-domain adaptive filters to track simulated echo paths for echo cancellation.When the microphone receives signals including echo and near-end human voice signals,the detection may be missed or wrong due to inaccurate dual-talk detection,resulting in the inability to eliminate echoes or incomplete voice signal transmission.Traditional single-microphone speech acquisition has a small range and has been gradually replaced by multi-microphone speech acquisition methods,but acoustic echo cancellation has also become difficult.For the above question,Based on the research of acoustic echo cancellation and double-talk detection,this paper proposes a highly accurate double-talk state detection method to ensure a complete voice call process.Then based on the adaptive beamforming technology of the microphone array,an acoustic echo cancellation method is proposed to improve the acoustic echo cancellation effect.The detailed work is as follows:(1)For the problem of adaptive divergence in the acoustic echo cancellation system in the case of double-talk,This paper proposes a dual-talk state detection method based on a combination of signal envelope detection and cross-correlation estimation.This method first uses the energy envelope of the far-end signal and the microphone signal to determine the double-talk state.When the near-end speech signal energy is low,the envelope detection method cannot accurately determine the double-talk state,and then use the cross-correlation estimation of the far-end signal and the microphone signal to determine the final judgment result of the double-talk state.The simulation results show that compared with the traditional energy comparison method and cross-correlation detection method,the method proposed in this paper accurately judges the occurrence of the double-talk process,avoids the problem of adaptive filter divergence,and is more beneficial to the improvement of acoustic echo cancellation performance.(2)For the problem of poor echo cancellation using a time-domain adaptive filter in a multi-microphone acoustic echo cancellation system,This paper proposes an acoustic echo cancellation method combining frequency-domain adaptive filtering and adaptive beamforming.The voice signal collected by the microphone is first detected by the voice state,and the frequency domain adaptive filter is controlled for echo cancellation based on the detection result.Then use adaptive beamforming to direct the nulling direction to the direction of the interference source to eliminate the residual echo,and the main lobe to the direction of the speaker for speech enhancement.Take voice enhancement.Simulation results show that the solution can improve the effect of echo cancellation and increase the range of speech collection.(3)Design a four-element microphone array test system to verify the acoustic echo cancellation method proposed in this paper.The test system is designed with the Xcore processor as the core and is based on the USB audio system framework design to achieve acoustic echo cancellation.Finally,the test was performed in an actual conference room.The results show that the complete voice call process is guaranteed while eliminating the acoustic echo,and it has a 5m voice acquisition range.