Hearing Aid Feedback Path Change Detection for Fast Filter Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional adaptive filters in hearing aids face a trade-off between convergence/tracking and steady-state errors, limiting their ability to quickly cancel feedback when situations change, such as when a user gets close to a hard surface.
Innovation Solution
A method to rapidly detect feedback/echo path changes by estimating an instant feedback path transfer function, allowing the adaptive filter to quickly adapt and override the current estimate, thereby improving convergence without compromising steady-state error performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional adaptive filters are used for feedback cancellation, then steady-state errors are reduced, but convergence/tracking speed is compromised
Solution Approach 1:
The patent implements dynamic adaptation of the feedback cancellation system by detecting sudden changes in the feedback path and adjusting the adaptive filter's transfer function accordingly. The system transitions between different operational states (normal adaptation vs. sudden change response) to optimize both convergence speed and steady-state error performance in different conditions.
Solution Approach 2:
The patent changes the parameters of the adaptive filter (specifically the transfer function) in response to detected feedback path changes. By modifying the filter parameters dynamically based on gradient detection and open-loop gain estimation, the system achieves faster tracking of feedback path variations while maintaining acceptable steady-state error levels.
2Speed
If the adaptive filter converges quickly upon feedback changes, then tracking speed is improved, but steady-state errors increase
Solution Approach 1:
The system dynamically adjusts its behavior based on the operational phase: during sudden changes, it prioritizes fast tracking by updating the transfer function based on gradient detection; during steady-state operation, it prioritizes precision by using the fully adapted filter. This dynamic switching resolves the contradiction between speed and precision.
Solution Approach 2:
The patent performs preliminary detection of feedback path changes by monitoring the gradient of the adaptive filter coefficients. When a sudden change is detected before full convergence occurs, the system takes preliminary action to adjust the transfer function, enabling faster tracking without sacrificing the eventual steady-state error performance.
3Measurement precision
If gradient smoothing is applied to detect feedback changes, then detection accuracy is improved, but response time increases
Solution Approach 1:
The patent applies partial smoothing to the gradient by using a first-order filter with a small positive coefficient alpha. This partial smoothing reduces noise in the gradient detection while maintaining relatively fast response time. The smoothing is applied selectively only to the gradient calculation, not to the entire feedback cancellation process, balancing detection accuracy and response speed.
Data Source
AI summary
A method of detecting a sudden change in a feedback/echo path of a hearing aid is based on a gradient of an adaptive filter comprising an adaptive algorithm configured to estimate the feedback/echo path. The method may comprise comparing a smoothed and processed version of the gradient values over time to a threshold value. A hearing aid configured to detect a sudden change in a feedback/echo path is furthermore provided.