Hierarchical Device Communication Using Edge Data Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In modern industrial processes, the centralized data backbone approach wastes bandwidth by transferring high-frequency data to master devices and back, as most of this data is not used by non-experts, leading to inefficient communication and data processing in hierarchical networks of distributed devices.
Innovation Solution
A method that involves client devices collecting and storing data, generating meta-data-based classification information, and sharing it with other devices to optimize data handling, allowing each edge device to be aware of data location and availability, thereby reducing unnecessary data transmission and improving bandwidth usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If data is transferred through centralized master device, then data can be stored and accessed centrally, but bandwidth is wasted due to unnecessary data transmission
Solution Approach 1:
The patent segments the centralized data management into distributed data handling at edge devices. Each edge device maintains local data and generates classification information independently, eliminating the need to transfer all data through a centralized master device. This segmentation reduces bandwidth waste while distributing management complexity across multiple devices rather than concentrating it.
Solution Approach 2:
The patent extracts only the essential classification information from raw data at the edge device level, rather than transferring complete datasets. By taking out only the necessary metadata and classification details, the system reduces data transmission volume significantly while maintaining the ability to manage and access data efficiently.
2Productivity
If all data is transferred to master device, then centralized storage is achieved, but communication efficiency decreases
Solution Approach 1:
The patent performs preliminary classification and metadata generation at the edge device before data transfer. By preparing classification information in advance, the system enables faster data retrieval and reduces the need for subsequent data transfers, thereby improving communication efficiency while maintaining data availability through pre-processed information.
Solution Approach 2:
The patent introduces classification information as an intermediary between raw data and data retrieval operations. This intermediary layer enables efficient data access by providing structured metadata that guides data retrieval without requiring complete data transfer, thus improving communication efficiency while preserving data availability.
3Loss of energy
If data is stored locally at edge devices, then bandwidth usage is reduced, but data sharing between devices becomes challenging
Solution Approach 1:
The patent implements feedback mechanisms where edge devices exchange classification information about their locally stored data. This feedback loop enables devices to understand what data is available at other edge devices without transferring the actual data, thus maintaining low bandwidth consumption while enabling intelligent data sharing through coordinated access requests.
Solution Approach 2:
The patent creates and shares copies of classification information and metadata rather than copying actual data. These lightweight copies enable edge devices to understand data availability and characteristics at remote devices, facilitating data sharing capabilities while maintaining minimal bandwidth consumption since only metadata copies are transmitted.
Data Source
Figure 1~2
AI summary
The present invention relates to a method (100) for providing an efficient communication in a hierarchical network (70) of distributed devices (72, 74, 76) comprising at least a master device (72), at least a first client device (74), wherein the at least first client device (74) is connected via a first communication interface (50) with the master device (72), comprising: - collecting (102) first data (10) from the at least first client device (74) sent by at least one sensor device (78) via a second communication interface (52); - storing (104) the received first data (10) in a data storage (80) of the at least one client device (74), - determining (106), by the at least first client device (74), meta data (12) of the first data (10); - receiving (108), by the at least first client device (74), a requirement information (14) of the first data (10); - generating (110) a classification information (20) of the received first data (10), by the at least first client device (74), according to the meta data (12) and the requirement information (14) of the first data (10); - providing (112) the classification information (20) of the first data (10) by the at least first client device (74) to the master device (72) and / or to at least a second client device (76) according to a rule information (30); and - updating (114) the classification information (20), when the first data (12) is changed.