Noise cancelling device and method, and noise cancelling program
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise cancelling methods using adaptive filters face divergence issues due to rapid changes in reference signal amplitude and phase, leading to mistaken noise cancellation and requiring pre-set threshold values based on device and environmental characteristics.
Innovation Solution
A noise cancelling device and method that classifies adaptive filter coefficients into groups using a basic tap position, detects divergence by comparing indices derived from these groups, and initializes coefficients to prevent divergence without relying on pre-set threshold values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pre-set threshold values are used to detect divergence, then divergence detection is possible, but the system requires complex threshold settings based on device and environmental characteristics
Solution Approach 1:
The system performs self-diagnosis by automatically detecting divergence through monitoring the update amount of filter coefficients, eliminating the need for external threshold settings. The adaptive filter monitors its own state and autonomously determines when divergence occurs, making the system self-sufficient without requiring complex pre-configured thresholds.
Solution Approach 2:
The invention changes the detection parameter from fixed threshold values to dynamic monitoring of coefficient update amounts. Instead of comparing against predetermined thresholds, the system detects divergence by measuring whether the change in filter coefficients exceeds a dynamically determined threshold based on recent update history, adapting to varying operating conditions automatically.
2Productivity
If the adaptive filter coefficient update continues without divergence detection, then noise cancellation processing is continuous, but the filter coefficient may diverge causing mistaken noise cancellation
Solution Approach 1:
The system implements feedback by continuously monitoring the update amount of filter coefficients during the adaptation process. This feedback mechanism allows the system to detect when the coefficient changes become excessively large, indicating impending divergence, and takes corrective action by stopping the update temporarily, preventing mistaken noise cancellation while maintaining continuous operation.
3Reliability
If the filter coefficient update is stopped frequently to prevent divergence, then divergence is avoided, but the noise cancellation response time increases
Solution Approach 1:
The system applies partial action by stopping the coefficient update only when necessary - specifically when the update amount exceeds the dynamically determined threshold indicating divergence risk. Rather than continuously stopping updates to ensure stability, the system intervenes only when actual divergence is detected, minimizing interruptions to the noise cancellation process while maintaining reliability.
Data Source
AI summary
A noise cancelling device for detecting divergence of a coefficient of an adaptive filter without setting a threshold value in advance based on characteristics of a device from which noise is to be cancelled or an environment. The noise cancelling device includes: a divergence coefficient unit for obtaining a filter divergence coefficient from a filter coefficient of a section not including a filter coefficient at a basic tap position decided by the distance between a reference input unit and an audio input unit; a divergence threshold value calculation unit for obtaining a filter divergence threshold value from a filter coefficient of a section including the filter coefficient at the basic tap position; and a divergence detection unit for comparing the filter divergence coefficient to the filter divergence threshold value so as to detect the divergence condition of the adaptive filter.