Precision Measuring Matrices for Waveform Source Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately measure and separate sound sources within complex compound waveforms, particularly when signals from different sources overlap in time and frequency, or when there are rapid changes in frequency and amplitude.
Innovation Solution
A machine-implemented method using Precision Measuring Matrices (PMMs) to identify and separate sound sources by relating cells based on domain relationships such as frequency, time, amplitude, harmonic, and patterns, both within defined ranges, enabling the visualization of sound sources within complex compound waveforms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large dimension DFT is used, then frequency resolution is improved, but time resolution deteriorates
Solution Approach 1:
The patent segments the compound waveform into multiple individual waveforms using domain separation techniques. By dividing the complex signal into separate components in different domains (time, frequency, amplitude), the system can analyze each segment with appropriate DFT dimensions without the trade-off affecting the entire signal. This allows simultaneous high frequency and time resolution for different waveform components.
Solution Approach 2:
The patent introduces domain separation as an additional dimension for analyzing waveforms. Instead of relying solely on time-frequency analysis with fixed DFT dimensions, the system separates waveforms across multiple domains (time domain, frequency domain, amplitude domain). This dimensional approach allows optimization of DFT size for specific analysis needs without compromising overall measurement accuracy.
2Loss of time
If a small dimension DFT is used, then time resolution is improved, but frequency resolution deteriorates
Solution Approach 1:
The patent segments the compound waveform into multiple individual waveforms using domain separation techniques. By dividing the complex signal into separate components in different domains (time, frequency, amplitude), the system can analyze each segment with appropriate DFT dimensions without the trade-off affecting the entire signal. This allows simultaneous high frequency and time resolution for different waveform components.
Solution Approach 2:
The patent introduces domain separation as an additional dimension for analyzing waveforms. Instead of relying solely on time-frequency analysis with fixed DFT dimensions, the system separates waveforms across multiple domains (time domain, frequency domain, amplitude domain). This dimensional approach allows optimization of DFT size for specific analysis needs without compromising overall measurement accuracy.
3Device complexity
If traditional time-frequency analysis is used, then measurement process is simplified, but accuracy of separating overlapping signals deteriorates
Solution Approach 1:
The patent segments the compound waveform into multiple individual waveforms using domain separation techniques. By dividing the complex signal into separate components in different domains (time, frequency, amplitude), the system can analyze each segment with appropriate DFT dimensions without the trade-off affecting the entire signal. This allows simultaneous high frequency and time resolution for different waveform components.
Solution Approach 2:
The patent introduces domain separation as an additional dimension for analyzing waveforms. Instead of relying solely on time-frequency analysis with fixed DFT dimensions, the system separates waveforms across multiple domains (time domain, frequency domain, amplitude domain). This dimensional approach allows optimization of DFT size for specific analysis needs without compromising overall measurement accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A machine-implemented method for computerized digital signal processing obtains a digital signal from data storage or from conversion of an analog signal and determines, from the digital signal, Measuring Matrices (MM). Each measuring matrix has a plurality of cells, each cell having an amplitude corresponding to the signal energy in a frequency bin for a time slice. Cells in each measuring matrix having maximum amplitudes within a time slice are identified as maximum cells. Maxima that coincide in time and frequency are identified and a correlated maxima matrix, called a "Precision Measuring Matrix" is constructed showing the coinciding maxima and the adjacent marked maxima are linked into partial chains. If only one MM is constructed, multiple types of maxima are identified to generate the Precision Measuring Matrix. The partial chains are then separated into defined domain relationships for identification and separation.