Echo Cancellation Using Composite Double-End Talk Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional double-end talk detection methods in communication systems face challenges in accurately identifying double-end talk scenarios due to noise and non-linear echo influences, leading to suboptimal echo elimination performance.
Innovation Solution
A method involving noise energy estimation, echo estimation, and error signal variance calculation to determine a determinant (ξ) that compares with a preset threshold, disabling adaptive filter adaptation when double-end talk is detected, thereby improving echo elimination efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If conventional correlation coefficient methods are used for double-end talk detection, then the detection process is simple, but the detection accuracy deteriorates due to noise and non-linear echo influences
Solution Approach 1:
The patent changes the detection parameters from simple correlation coefficients to a composite metric involving signal power ratios and energy differences. Specifically, it calculates the ratio of error signal power to echo signal power, and the difference between local voice signal energy and noise energy, combining these parameters to improve detection accuracy while maintaining reasonable complexity
Solution Approach 2:
The patent creates a composite detection metric by combining multiple signal characteristics (error signal power, echo signal power, local voice energy, noise energy) into a unified double-end talk detection criterion. This composite approach allows the system to leverage multiple aspects of the signal to achieve more accurate detection under noisy and non-linear conditions
2Adaptability or versatility
If adaptive filter adaptation is enabled during double-end talk, then the echo route estimation can be continuously updated, but the echo elimination performance deteriorates due to incorrect coefficient adaptation
Solution Approach 1:
The patent implements a feedback mechanism where the double-end talk detection result directly controls the adaptation state of the echo canceller. When double-end talk is detected, the adaptive filter coefficients are frozen to prevent incorrect updates, and when single-end talk is detected, adaptation is enabled. This feedback loop ensures reliable echo elimination by adapting the filter only under appropriate conditions
Solution Approach 2:
The patent dynamically adjusts the adaptation state of the echo canceller based on real-time double-end talk detection. The system transitions between two operational modes: adaptation enabled during single-end talk and adaptation disabled during double-end talk. This dynamic control allows the system to optimize echo elimination performance by adapting to changing speech conditions
Data Source
AI summary
A method and system for eliminating echo in a speaker-microphone communication system are provided. The method includes the steps of: performing a noise energy estimating process on a local microphone signal in order to obtain an estimated noise signal, wherein the local microphone signal includes a local voice signal, possible background noise signal, and possible remote voice signal output from a speaker and received by a microphone; performing an echo estimating process on a remote voice signal to obtain an estimated echo signal, wherein the remote voice signal is output from the speaker; determining an error signal from the local microphone signal and the estimated echo signal; calculating a variance (σe2) of the error signal and a variance ({circumflex over (σ)}n2) of the estimated noise signal; calculating a determinant (ξ), wherein the determinant (ξ) corresponds to the variance (σe2) of the error signal and variance ({circumflex over (σ)}n2) of the estimated noise signal; and comparing the determinant (ξ) and a preset threshold, wherein when the determinant (ξ) is lower than the preset threshold, it is determined that double-end talk has not occurred, otherwise, it is determined that double-end talk has occurred.