MEMS Microphone Array Signal Source Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multiple-microphone systems face limitations in effectively separating audio sources due to inadequate microphone spacing and noise interference, particularly in environments with closely spaced microphones where phase differences and direction of arrival estimates are noisy.
Innovation Solution
A micro-electromechanical system (MEMS) microphone unit with closely spaced elements, integrated circuitry, and probabilistic inference techniques, such as Belief Propagation, are used to process direction of arrival information and sound structure, enabling effective source separation by constructing time-frequency masks and reducing noise interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If microphones are spaced closely together to reduce device size, then device compactness is improved, but direction of arrival estimation accuracy deteriorates due to noisy phase differences
Solution Approach 1:
The patent introduces an intermediary processing stage that computes direction of arrival histograms from multiple closely-spaced microphones. Instead of directly using noisy phase differences, the system aggregates directional information across multiple time-frequency bins to form histograms, which serve as a statistical intermediary that filters out noise and provides more reliable direction estimates for source separation.
2Measurement precision
If microphones are spaced far apart to improve direction of arrival estimation, then measurement precision is improved, but device complexity and size increase
Solution Approach 1:
The patent replaces the mechanical solution of widely-spacing microphones with a signal processing substitution. Instead of relying on physical distance to obtain adequate phase differences, the system uses computational methods including histogram aggregation and probabilistic source separation to achieve accurate direction of arrival estimation from closely-spaced microphones, thereby substituting mechanical arrangement with information processing.
3Volume of moving object
If conventional beamforming is used with closely spaced microphones, then device compactness is improved, but source separation performance deteriorates due to inadequate microphone separation
Solution Approach 1:
The patent fundamentally changes the processing parameters and methods beyond conventional beamforming. It transforms the approach by computing direction of arrival histograms across multiple time-frequency bins and using probabilistic source separation that exploits W-disjoint orthogonality. This parameter and method transformation enables effective source separation from closely-spaced microphones where traditional beamforming fails.
Data Source
AI summary
In one aspect, a microphone with closely spaced elements is used to acquire multiple signals from which a signal from a desired source is separated. The signal separation approach uses a combination of direction-of-arrival information or other information determined from variation such as phase, delay, and amplitude among the acquired signals, as well as structural information for the signal from the source of interest and/or for the interfering signals. Through this combination of information, the elements may be spaced more closely than may be effective for conventional beamforming approaches. In some examples, all the microphone elements are integrated into a single a micro-electrical-mechanical system (MEMS).