Angular Discrimination via Phase-Shifted Comb Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The 'cocktail party problem' poses a challenge in discriminating a specific auditory source from background noise, particularly difficult for the hearing impaired, as existing methods fail to effectively isolate and reduce interference from multiple sound sources.
Innovation Solution
The technique involves using an array of sensors, such as microphones, to sample multiple signals, determine transform signals, and apply a comb filter function based on phase differences to suppress off-axis noise, allowing for angular discrimination and inverse transformation to enhance directional sound reception.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple signals are sampled with multiple spatially separated sensors to enable angular discrimination, then the ability to isolate specific sound sources is improved, but the complexity of the signal processing system increases
Solution Approach 1:
The patent segments the frequency spectrum into multiple discrete frequency bins and processes each bin independently through the comb filter function. This segmentation allows the system to handle complex multi-source signals by breaking them down into manageable frequency components, each of which can be selectively filtered based on its phase characteristics.
Solution Approach 2:
The patent changes the phase parameter of signal components based on their frequency and phase angle relative to a reference direction. By applying phase shifts that are functions of frequency and angle, the system creates constructive interference for on-axis signals and destructive interference for off-axis signals, enabling directional discrimination without requiring complex spatial arrangements.
2Object-affected harmful factors
If a comb filter function is applied to suppress off-axis noise based on phase differences, then off-axis noise reduction is improved, but the complexity of the filtering operation increases
Solution Approach 1:
The patent applies different filtering operations to different frequency bins based on their local phase characteristics. Each frequency bin is processed independently with a comb filter function that is tailored to that specific frequency's phase angle, allowing precise local suppression of off-axis noise while preserving on-axis signals.
Solution Approach 2:
The patent creates a comb filter function that copies and processes each frequency component separately, applying the appropriate phase shift and attenuation based on its angular characteristics. This copying approach allows the system to handle complex noise suppression by treating each frequency component as an independent copy that can be filtered according to its specific properties.
3Measurement precision
If transform signals are modified by altering kernel bins based on difference thresholds, then signal discrimination capability is improved, but the computational complexity increases
Solution Approach 1:
The patent applies partial action by only modifying those kernel bins where the difference between transform signals exceeds a predetermined threshold. This selective modification approach processes only the frequency components that require discrimination, avoiding unnecessary computational operations on bins that already meet the discrimination criteria.
Solution Approach 2:
The patent changes parameters of transform signals by selectively altering kernel bins based on threshold comparisons. When the phase or magnitude difference between signals at a given frequency bin exceeds the threshold, the system modifies that bin's parameters (such as applying attenuation or phase correction) to enhance discrimination between competing sound sources.
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 method effectively reduces off-axis noise and enhances the reception of communication signals, improving the ability to isolate specific sound sources amidst background noise, particularly beneficial for hearing aids and other applications requiring directional sound sensitivity.
Implementation Method 1
a criterion applied to the phase angle associated with that frequency value
Implementation Method 2
A comb of weights is created for each value of frequency, with the weights being determined based upon a criterion applied to the phase angle associated with that frequency value
Data Source
AI summary
Angular discrimination between signals is described. Multiple signals are sampled with multiple signal sensors spatially separated from each other. Respective transform signals are determined that are representative of each sampled signal. If a difference between the transform signals in a given transform kernel bin is greater than a difference threshold, a first transform signal is modified by altering that kernel bin. Then the resulting modified transform signal is inverse transformed.