Adaptive Long-Term Prediction Filter for Tonal Feedback Cancellation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Adaptive feedback cancellation systems in hearing aids fail to effectively reduce feedback for tonal input signals, leading to signal distortions and potential howls due to correlation between input and output signals, which existing methods like notch filters do not adequately address.
Innovation Solution
A hearing instrument with a feedback cancellation system that employs long-term prediction filters and spectral shaping filters to estimate and reduce tonal components in the feedback path, using adaptive algorithms to update filter parameters based on input and output signals, thereby minimizing the impact of tonal components on feedback estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If adaptive feedback cancellation algorithms (e.g., NLMS, RLS) are used to reduce feedback, then feedback reduction is achieved for general signals, but the algorithms produce biased estimates for tonal signals due to autocorrelation between input and output signals
Solution Approach 1:
The patent extracts and removes tonal components from the input and output signals before they are fed to the adaptive feedback cancellation algorithm. This is achieved through spectral analysis to identify tonal components and their removal via spectral subtraction or filtering, preventing the autocorrelation problem that causes biased estimates in conventional systems.
Solution Approach 2:
The patent applies preliminary signal processing to remove tonal components from the signals before they enter the adaptive algorithm. By pre-processing the signals to eliminate problematic tonal elements, the system ensures that the subsequent adaptive feedback cancellation operates on cleaned signals, avoiding the correlation issues that would otherwise degrade performance.
2Reliability
If notch filters are used to remove tonal components, then tonal impact is reduced, but the system complexity increases due to cascade of independent notch filters
Solution Approach 1:
The patent combines multiple independent notch filters into a single unified filter structure that achieves the same tonal component removal functionality. Instead of cascading separate notch filters for each frequency, the system uses a consolidated approach that maintains effectiveness while reducing structural complexity and computational overhead.
Solution Approach 2:
The patent employs a universal filter structure that can handle multiple tonal frequencies simultaneously through a single processing block. This multi-functional approach replaces the need for multiple specialized notch filters, each targeting a specific frequency, with one adaptable filter that can address various tonal components in a unified manner.
3Reliability
If feedback cancellation is applied to tonal signals, then feedback reduction is attempted, but signal distortions and loss of intelligibility occur due to removal of target signal components
Solution Approach 1:
The patent uses feedback mechanisms where the output of the filter is monitored and used to adjust the filtering process. By continuously analyzing the filtered signal and comparing it with the original, the system can detect and preserve important signal components that might otherwise be removed, ensuring that feedback cancellation does not inadvertently eliminate meaningful speech or audio information.
Solution Approach 2:
The patent dynamically adjusts filtering parameters based on the characteristics of the input signal. By adapting the filter's frequency response, depth, and bandwidth according to the detected signal properties, the system can selectively remove harmful tonal feedback components while preserving essential speech frequencies and maintaining signal intelligibility and quality.
Data Source
AI summary
A method of estimating acoustic feedback in a hearing instrument in order to reduce the impact of tonal components of acoustic feedback. The hearing instrument comprises an input transducer, an output transducer, a forward path being defined between the input transducer and the output transducer, a signal processing unit defining an input side and an output side of the forward path, and a feedback loop from the output side to the input side. The feedback loop comprises a feedback path estimation unit receiving first and second estimation input signals from the input and output side of the forward path, respectively, wherein the first and second estimation input signal paths comprise first and second long term prediction filters P(z), the feedback cancellation system being adapted to provide that the variable parameters of at least one of the long term prediction filters are estimated based on the filter input signal.