Time-Series Data Transmission Using PCA and DTW Change Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large volumes of complex data in industrial and healthcare applications lead to high transmission and storage costs, with existing data compression methods compromising data quality.
Innovation Solution
A method that uses principal component analysis (PCA) and dynamic time warping (DTW) to detect data changes, generating and transmitting data packages only when significant changes are detected, thereby reducing the data volume without compromising quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data compression is performed before transmission, then transmission volume is reduced, but data quality deteriorates
Solution Approach 1:
The patent transforms the data representation parameters by projecting multivariate time series onto principal components, changing the coordinate system rather than compressing the data values. This parameter transformation reduces dimensionality while preserving the essential variance and relationships in the data, thereby reducing transmission volume without compromising data quality.
Solution Approach 2:
The patent applies dimensionality reduction by projecting data from multiple dimensions (multivariate) onto a fewer number of principal component dimensions. This dimensional transformation allows the same information to be represented in a lower-dimensional space, reducing transmission requirements while maintaining data integrity through the mathematical properties of PCA.
2Quantity of substance
If data manipulation such as averaging is performed, then data volume is reduced, but data quality deteriorates significantly
Solution Approach 1:
Instead of manipulating data values through averaging or aggregation, the patent changes the parameter representation by projecting onto principal components. This preserves the original data information while transforming it into a more compact form that captures the essential variance, avoiding the quality loss inherent in averaging operations.
3Loss of information
If all data packages are transmitted, then data completeness is maintained, but transmission costs increase
Solution Approach 1:
The patent implements a feedback mechanism that compares consecutive data packages using DTW alignment and calculates distance metrics. Based on this comparison, the system selectively transmits only those packages that contain significant changes, using the threshold-based decision rule to balance data completeness with transmission cost reduction.
Solution Approach 2:
The patent changes the transmission decision parameter from transmitting all packages to transmitting only packages exceeding a distance threshold. This parameter change in the transmission strategy, combined with PCA dimensionality reduction, significantly reduces transmission costs while maintaining data completeness for meaningful changes.
Data Source
AI summary
A computer-implemented method for data transmission at reduced data volumes includes sampling a data stream in consecutive time intervals and generating consecutive data packages, calculating, via principal component analysis, projections of its data points onto a set of principal components, aligning, via dynamic time warping, the projected data points to respective projected data points of an earlier generated data package and calculating distances between the aligned data points, determining and comparing the maximum of the calculated distances to a threshold value, and transmitting the newly generated data package to a data sink and setting the newly generated data package as the earlier generated data package for the next aligning step if the maximum distance exceeds the threshold value or otherwise inhibiting the transmission and retaining the earlier generated data package for the next aligning step.