Adaptive Masking Filter for In-Car Audio Howling Attenuation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In-car communication systems often experience audio howling due to microphone/loudspeaker feedback, which creates an unpleasant noise for listeners, especially in high noise conditions.
Innovation Solution
The system identifies susceptible audio frequencies and uses an adaptive masking filter to suppress only those frequencies, applying it to the loudspeaker output to attenuate howling, leveraging existing components like noise suppression modules and acoustic echo cancellers to minimize additional computations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If audio feedback loop is allowed in in-car communication system, then hands-free communication functionality is achieved, but audio howling occurs due to microphone/loudspeaker feedback
Solution Approach 1:
The patent uses the feedback signal that causes howling as a useful resource to identify problematic frequencies. By analyzing the feedback loop and detecting frequencies where the loop gain exceeds unity, the system converts the harmful feedback into beneficial information for generating an attenuation mask that selectively suppresses howling frequencies while preserving speech quality
Solution Approach 2:
Instead of applying uniform attenuation across all frequencies, the patent applies frequency-selective attenuation only at specific frequencies where howling occurs. The attenuation mask is generated to have frequency-dependent characteristics, attenuating only the problematic frequencies identified from the feedback analysis while leaving other frequencies unchanged, thus maintaining speech quality while eliminating howling
2Object-affected harmful factors
If frequency-selective attenuation is applied to suppress howling, then howling noise is reduced, but computational load increases
Solution Approach 1:
The patent performs preliminary analysis of the feedback loop to pre-identify frequencies where howling is likely to occur. By detecting frequencies with loop gain exceeding unity before howling actually occurs, the system can pre-generate the attenuation mask, avoiding the need for complex real-time frequency analysis during active howling suppression
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously monitors the audio feedback loop and dynamically adjusts the attenuation mask based on detected feedback characteristics. This feedback-driven approach allows the system to adapt to changing acoustic conditions while maintaining computational efficiency by only processing frequencies that actually exhibit feedback behavior
Data Source
AI summary
Regions of the audio frequency spectrum that are most susceptible to howling are identified or determined. When detected, these approaches shape the frequency spectrum of the audio going to loudspeakers so that only those howling frequencies are suppressed.