Discrete Wavelet Transform for Sensor Data Redundancy Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data storage systems face challenges in managing large amounts of sensor data, particularly in edge computing architectures, where bandwidth limitations and storage costs are concerns, and existing solutions like cloud storage may not be ideal due to bandwidth constraints and data security risks.
Innovation Solution
The proposed solution involves using discrete wavelet transformations to identify and eliminate redundancies in video signals from multiple sensors, reducing data storage size by storing only non-redundant information, which is achieved through the determination of temporally stationary background data and non-stationary data, and implementing this in a distributed data storage and computation system with edge nodes that perform local computation and storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If all sensor data is stored to ensure complete information retention, then data completeness is improved, but storage capacity and bandwidth usage are worsened
Solution Approach 1:
The patent extracts and removes redundant information from sensor data streams by identifying temporally stationary background components that are common across multiple sensors. Only the unique, non-redundant portions of data are retained for storage, thereby maintaining data completeness while significantly reducing storage requirements.
Solution Approach 2:
The patent transforms sensor data from the time domain to the frequency domain using discrete wavelet transforms. This parameter change in representation allows for efficient identification and separation of redundant stationary components from unique dynamic components, enabling optimized storage decisions.
2Device complexity
If redundant data is stored to simplify processing, then processing complexity is reduced, but storage costs and bandwidth requirements are increased
Solution Approach 1:
The patent performs preliminary analysis of sensor data to identify and mark redundant portions before the actual storage process. By pre-identifying temporally stationary background data that can be reconstructed from other sensors, the system avoids storing unnecessary information, reducing both storage costs and bandwidth requirements while maintaining processing efficiency.
3Quantity of substance
If discrete wavelet transformations are applied to identify redundancies, then storage efficiency is improved, but computational complexity and power consumption are worsened
Solution Approach 1:
The patent segments sensor data into distinct frequency components using discrete wavelet transforms, separating temporally stationary background components from dynamic foreground components. This segmentation allows the system to efficiently identify and eliminate redundancies without requiring excessive computational resources, as the transform provides a natural decomposition that facilitates targeted processing.
Data Source
AI summary
Two or more video signals from two or more sensors are received and discrete wavelet transformations are performed on the two or more video signals. Temporally stationary and non-stationary background data is determined from the discrete wavelet transformations. Redundancies are determined using the temporally stationary background data, the redundancies indicating an overlap between the two or more video signals Data of the two or more video signals is stored without the redundancies to reduce a storage size.