Signal Decomposition Using Reference Signal Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Blind Source Separation (BSS) techniques face challenges in accurately identifying and isolating individual source signals from mixed sensor responses, as they are hindered by the ambiguity of source power, permutation, time-shifting, and spectral differences, leading to undefined 'hidden' source signals.
Innovation Solution
The system processes data signals by subtracting identified signal terms and using reference signals organized into mutually independent partitioning support sets to compute additional independent signal terms, employing methods that go beyond traditional BSS to achieve more precise signal decomposition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional Blind Source Separation (BSS) techniques are used to separate source signals from mixed sensor responses, then source separation can be achieved, but accuracy is reduced due to ambiguity in source power, permutation, time-shifting, and spectral differences
Solution Approach 1:
The patent segments the mixed sensor response into multiple candidate source signals by identifying different time-shifted and spectrally-modulated versions of each source. Instead of treating the separation problem as a single ambiguous task, the system divides it into multiple discrete candidate signals that can be systematically evaluated and compared, thereby resolving the permutation and time-shifting ambiguities that plague traditional BSS methods.
Solution Approach 2:
The system generates multiple candidate source signals by applying parameter transformations to reference signals, specifically varying time-shifting and spectral modulation parameters. By systematically exploring different parameter combinations, the system creates a comprehensive set of candidates that covers the ambiguity space, allowing for more accurate source identification through comparison with actual sensor responses.
2Measurement precision
If multiple candidate source signals are generated and compared to identify the best match, then source identification accuracy improves, but computational complexity increases
Solution Approach 1:
The system performs preliminary generation of candidate source signals before the actual matching process. By pre-computing multiple candidate signals with different time-shifting and spectral characteristics, the system prepares a ready-to-use set of candidates that can be quickly compared against sensor responses during operation, avoiding the need for complex real-time optimization and reducing overall computational burden.
Solution Approach 2:
The patent employs computationally inexpensive candidate signal generation methods that create multiple disposable candidate signals for comparison. These candidate signals are generated using simple time-shifting and spectral modulation operations rather than complex transformations, allowing for rapid generation and disposal of numerous candidates without excessive computational cost, thereby improving accuracy without proportionally increasing system complexity.
Data Source
AI summary
A system processes data signals consisting of sums of independent signal terms, zero or more of which signal terms may already have been identified, in order to generate one or more additional terms. Deflated versions of the data signals are created by subtracting from the data signals any previously identified signal terms. Additional independent signal terms are computed using a set of reference signals organized into mutually independent partioning support sets. The images of each support set are computed on the data signals. Computed images on a data signal that are non-zero are identified as additional independent signal terms of that data signal.