Delay Estimation Using Artificial Near-End Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing delay estimation methods in communication systems face challenges when there are disturbances, such as double-talk or noise, which affect the accuracy of echo cancellation by misrepresenting the correlation values and making it difficult to determine the true delay between far-end and near-end signals.
Innovation Solution
The method generates artificial near-end signals when disturbances are present, maintaining a constant correlation value with the far-end signals to robustly estimate the delay, thereby reducing the impact of disturbances on the delay estimation process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If correlation values are calculated using received near-end signals during disturbance conditions (double-talk or noise), then the delay estimation may be computed continuously, but the accuracy of delay estimation deteriorates due to misrepresentation of correlation values
Solution Approach 1:
The system performs preliminary action by detecting disturbance conditions (double-talk or noise) before they corrupt the correlation calculation. When disturbance is detected, the system proactively switches to using previously stored clean near-end signals instead of continuing to use corrupted received signals, thereby preventing accuracy deterioration while maintaining continuous operation
Solution Approach 2:
The system creates a copy of clean near-end signals stored in memory when disturbance conditions are detected. This copied clean signal replaces the corrupted received signal in the correlation calculation, allowing continuous delay estimation to proceed with accurate data rather than disturbed data
2Measurement precision
If artificial near-end signals are generated to maintain constant correlation values during disturbance, then delay estimation accuracy is improved, but device complexity increases
Solution Approach 1:
The system extracts and isolates the disturbance components (double-talk or noise) from the near-end signal. By identifying and separating the problematic elements, the system can exclude them from correlation calculations and use only the clean signal portions, improving accuracy without requiring complex artificial signal generation
Solution Approach 2:
The system introduces an intermediary mechanism (disturbance detection and signal selection logic) that mediates between the received corrupted signal and the correlation calculation process. This intermediary selectively routes clean stored signals to the correlation calculator when disturbance is detected, providing a simple control-based solution rather than complex signal processing
Data Source
AI summary
A delay between a first signal and a second signal is estimated. The first signal and second signals are received and for each of a plurality of candidate delays between the signals, a correlation value is determined. Based on the correlation values, one of the candidate delays is selected to be used as an estimate of the delay between the first and second signals.