Cross-Microphone Echo Cancellation Without Synchronized References
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional linear acoustic echo cancellation methods struggle to effectively remove echoes in conference sessions due to varying time delays between endpoints, as they rely on locally generated reference signals that are not synchronized.
Innovation Solution
Implementing reference-less cross-microphone echo cancellation (No-Ref AEC) techniques that distinguish between wanted and unwanted audio signals based on loudspeaker status and reverberation detection, suppressing echoes without relying on synchronized reference signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional linear acoustic echo cancellation is used with locally generated reference signals, then echo cancellation can be implemented at each endpoint, but echo cancellation effectiveness deteriorates due to substantially differing time delays between endpoints
Solution Approach 1:
The patent introduces a central server as an intermediary that collects audio data from all endpoints and performs unified echo cancellation processing. This mediator receives audio streams from multiple endpoints, aligns them in a common time reference frame, and applies AEC algorithms centrally, eliminating the time delay synchronization issues that plague distributed endpoint-based AEC systems.
Solution Approach 2:
The patent merges the echo cancellation functionality from distributed endpoints into a centralized processing architecture. Multiple audio streams from different endpoints are combined and processed together at the server, allowing unified time alignment and coordinated echo suppression across all participants, thereby resolving the time delay differential problem.
2Device complexity
If reference-less cross-microphone echo cancellation is implemented, then synchronization requirements are eliminated, but system complexity increases due to audio distortion detection and classification mechanisms
Solution Approach 1:
The patent replaces complex mechanical/time-synchronization mechanisms with audio signal processing-based detection. Instead of relying on precise time synchronization hardware and protocols, the system uses acoustic analysis methods to detect echo, reverberation, and loudspeaker distortion through audio signal characteristics, substituting temporal coordination with spectral and temporal audio feature analysis.
Solution Approach 2:
The patent changes the detection parameters from time-based synchronization metrics to audio signal characteristics such as spectral content, temporal envelope, and acoustic impulse response features. By analyzing these audio parameters, the system can distinguish between desired speech and unwanted echoes/reverberation without requiring synchronized reference signals.
Data Source
AI summary
A method is performed by an endpoint device that includes a microphone and a loudspeaker. The method comprises: muting the loudspeaker; participating in a conference session with a neighbor endpoint device that shares a space with the endpoint device; detecting audio in the space using the microphone to produce detected audio; determining whether audio distortion, originating at a neighbor loudspeaker of the neighbor endpoint device, is present or absent in the detected audio; and taking action to transmit the detected audio to the conference session, or not transmit the detected audio to the conference session to prevent echo, based on a result of the determining.