Echo Path Change Detector Avoiding Double Talk
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Acoustic echo cancellers face challenges in effectively adapting to echo path changes during double talk situations, leading to reduced quality of echo cancellation due to conflicting adaptation strategies.
Innovation Solution
Implementing an echo path change detector that avoids detection during double talk by calculating and comparing four specific signal statistics with predefined thresholds, ensuring that echo path changes are only detected in the absence of double talk, thereby preventing instability in the adaptive filter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the echo path change detector increases the rate of adaptation when echo path changes are detected, then the AEC can quickly adapt to new echo conditions, but the estimated echo may diverge when double talk occurs simultaneously
Solution Approach 1:
The system dynamically adjusts the adaptation rate based on detected conditions. During echo path changes, the adaptation rate is increased to quickly track new conditions. During double talk, the adaptation rate is reduced or stopped to prevent divergence. The system transitions between different adaptation states based on real-time detection of acoustic conditions.
Solution Approach 2:
An intermediary detection mechanism is introduced to distinguish between echo path changes and double talk conditions before triggering adaptation. The system uses statistical analysis of acoustic signals to detect the type of condition occurring, and only triggers increased adaptation when echo path changes are detected in the absence of double talk, thereby preventing premature or inappropriate adaptation.
2Reliability
If the AEC slows adaptation during double talk to prevent divergence, then the estimated echo remains stable, but the AEC cannot quickly adapt to echo path changes when they occur simultaneously
Solution Approach 1:
The adaptation rate is dynamically controlled based on the detected acoustic condition. The system can operate in different adaptation modes: normal adaptation during single talk, reduced adaptation during double talk, and accelerated adaptation during echo path changes. This dynamic control allows the system to optimize between stability and adaptability based on real-time conditions.
Solution Approach 2:
The system changes the adaptation parameter (step size or convergence rate) based on detected conditions. During double talk, the adaptation parameter is reduced to maintain stability. When echo path changes are detected without double talk, the adaptation parameter is increased to accelerate tracking of new echo conditions, thereby optimizing performance across different operational scenarios.
3Measurement precision
If the system uses multiple statistical comparisons to avoid false echo path change detection during double talk, then the detection accuracy improves, but the computational complexity increases
Solution Approach 1:
The detection process is segmented into multiple independent statistical comparisons. Instead of using a single complex detection metric, the system divides the detection into separate analyses of different signal statistics (such as correlation coefficients, energy levels, and spectral characteristics). Each statistic is compared independently against thresholds, and only when multiple conditions are satisfied is an echo path change declared, thereby reducing false detections during double talk.
Data Source
AI summary
An echo path change detector may be used to control the rate of adaptation in an acoustic echo canceller. When echo path change is declared, the rate of adaptation may be increased. However, echo path change should not be declared in the presence of double talk, because rapid adaptation during double talk is undesirable. Accordingly, various features are disclosed for detecting echo path changes while avoiding the declaration of such changes in the presence of double talk.