Adaptive Acoustic Echo Delay Estimation via Weight Spike Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Accurate acoustic echo delay estimation is challenging due to disturbances and noise, leading to inaccuracies in echo cancellation in communication devices like mobile phones.
Innovation Solution
An adaptive filter with a momentum term and time-decaying learning factor is used to improve convergence rates and accuracy of delay estimation, selecting a sampling point with significant weight increases for precise delay calculation, and applying multiplicative gains to render echoes inaudible.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional delay estimation methods are used, then the system is simple to implement, but the measurement precision of delay estimation deteriorates due to disturbances and noise
Solution Approach 1:
The patent applies preliminary action by performing adaptive filtering and delay estimation during a training phase before actual echo cancellation. The system pre-processes the acoustic environment to establish filter coefficients and delay values, which are then reused during communication. This separates the complex estimation process from real-time operation, improving precision without compromising real-time performance.
Solution Approach 2:
The patent introduces an adaptive filtering system as an intermediary between the raw microphone signal and the echo cancellation process. This intermediary component processes the noisy signal to extract accurate delay information, effectively mediating between the noisy environment and the precision requirements of echo cancellation.
2Reliability
If real-time echo cancellation is implemented, then the productivity of communication is improved, but the reliability of echo cancellation deteriorates due to inaccurate delay estimation
Solution Approach 1:
The system performs delay estimation and filter adaptation in advance during a training phase, storing the results for reuse. This preliminary action ensures reliable delay values are established before actual communication begins, eliminating the need for continuous real-time estimation and ensuring consistent reliability throughout the communication session.
Solution Approach 2:
The adaptive filtering system uses the available training data to self-calibrate and establish accurate delay estimates without requiring continuous external intervention or complex real-time computation. The system serves itself by leveraging the training phase results to maintain reliable echo cancellation throughout operation.
Data Source
AI summary
A method for acoustic echo cancellation is disclosed herein. A microphone receives a second acoustic signal from a near-end environment, the second acoustic signal including a delayed version of the first acoustic signal from a far-end environment. A processor models a relationship between the first acoustic signal and the second acoustic signal using an adaptive filter. The adaptive filter uses sampling points of the first acoustic signal and the second acoustic signal along a timeline as inputs. The processor identifies a sampling point among the sampling points, wherein weight values of the adaptive filter associated with the identified sampling point experience a significant increase (e.g., 50% increase). The identified sampling point along the timeline represents an estimated delay between the first acoustic signal and the second acoustic signal. The processor further removes the delayed version of the first acoustic signal from the second acoustic signal based on the estimated delay.