Hearing Aid Feedback Suppression via Signal Decorrelation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing hearing aids face challenges in reducing internal feedback due to biased estimations of the feedback path, particularly when dealing with correlated input signals like music, which can lead to feedback oscillations and affect sound quality.

Innovation Solution

A hearing aid design that incorporates a synthesizer with a noise generator to create a synthesized signal based on a sound model, which is uncorrelated with the input signal, reducing bias by including this signal in the output, thereby decorrelating the input and output signals and simplifying the circuitry, using white noise or colored noise for efficient decorrelation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If adaptive feedback cancellation is used to suppress feedback, then feedback suppression is improved, but biased estimation occurs when input signals are correlated (e.g., music)

Engineering Contradiction:

Improvefeedback suppressionVSAvoidfeedback path estimation

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The patent introduces a synthesized signal as an intermediary element that mediates between the correlated input signal and the feedback path estimation. This synthesized signal, generated by a sound model excited with noise, serves as a proxy that breaks the correlation while preserving the essential characteristics needed for accurate feedback path estimation, thereby resolving the bias problem without compromising feedback suppression.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent changes the parameter of signal correlation by transforming the correlated input signal into a synthesized signal with different statistical properties. By using a noise-excited sound model, the synthesized signal exhibits uncorrelated characteristics that eliminate the bias in feedback path estimation, while still maintaining the ability to represent the acoustic environment for effective feedback cancellation.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If signal de-correlating operations are introduced to reduce bias, then estimation accuracy is improved, but device complexity increases

Engineering Contradiction:

Improvefeedback path estimationVSAvoidcircuitry complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent creates a synthesized copy of the input signal through a sound model that captures the essential acoustic characteristics. This synthesized copy, rather than the original correlated signal, is used for feedback path estimation. The copying approach achieves de-correlation while maintaining computational efficiency, as the sound model can be implemented with relatively simple algorithms compared to other de-correlation techniques.

Inventive Principle:

Principle #26Copying

3Manufacturing precision

If voiced activity detection and pitch estimation are implemented for vowel synthesis, then speech quality is improved, but computational load increases

Engineering Contradiction:

Improvespeech synthesis qualityVSAvoidcomputational load

Core Design Contradiction:

Manufacturing precisionVSUse of energy by moving object

Solution Approach 1:

The patent replaces the computationally expensive voiced activity detection and pitch estimation mechanisms with a simpler noise-based excitation approach. The noise-excited sound model produces synthesized speech that is sufficient for feedback cancellation purposes, sacrificing some speech quality details in exchange for dramatically reduced computational requirements. This approach uses inexpensive, simple-to-compute noise signals instead of complex analysis algorithms.

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach effectively reduces feedback bias, improving dynamic performance and maintaining speech intelligibility and sound quality, with increased stable gain and robustness against feedback whistling, especially in high-frequency regions.

Implementation Method 1

a noise generator configured for excitation of the sound model for generation of the synthesized signal including synthesized vowels

Methodology Applied

Scientific EffectNoise generation:

Implementation Method 2

a microphone for converting sound into an audio input signal

Methodology Applied

Scientific EffectAcoustic transduction:

Implementation Method 3

a receiver for converting an audio output signal into an output sound signal

Methodology Applied

Scientific EffectElectrical transduction:

Data Source

PatentUS8345902B2Hearing aid with means for decorrelating input and output signals

Publication Date: 2013.01.01 GN HEARING AS

AI summary

A hearing aid includes a microphone for converting sound into an audio input signal, a hearing loss processor configured for processing the audio input signal or a signal derived from the audio input signal in accordance with a hearing loss of a user of the hearing aid, a synthesizer configured for generation of a synthesized signal based at least on a sound model and the audio input signal, the synthesizer comprising a noise generator configured for excitation of the sound model for generation of the synthesized signal including synthesized vowels, and a receiver for generating an output sound signal based at least on the synthesized signal.