Cloud-Edge Feature Processing With Uncertainty-Guided Data Offloading
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cloud-edge collaborative processing methods suffer from high communication overhead due to the lower computing capability of the edge, which often performs simple preprocessing and data compression, leading to excessive data transmission between the cloud and the edge.
Innovation Solution
A cloud-edge collaborative data processing method and system that segments processing between the edge and cloud, using lightweight decoders and compression models to transform data into smaller dimensions, selectively compressing data based on feature uncertainty, and dynamically determining processing locations for reduced communication overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the edge performs simple preprocessing and compresses data before sending to the cloud, then the computing capability constraint is satisfied, but the communication overhead increases due to large amount of data transmission
Solution Approach 1:
The patent segments the data processing workflow into multiple stages: edge-side feature extraction and initial processing, uncertainty assessment, selective compression, and cloud-side final processing. This segmentation allows different operations to be performed at appropriate locations, reducing unnecessary data transmission while maintaining processing efficiency.
Solution Approach 2:
The patent introduces dynamic decision-making at the edge based on feature uncertainty assessment. The system dynamically determines whether to compress and transmit data or process locally, adapting to varying data characteristics and processing requirements. This dynamic approach optimizes the balance between edge computing capabilities and cloud communication overhead.
2Measurement precision
If more data is transmitted from edge to cloud for processing, then the processing accuracy improves, but the communication time and bandwidth consumption increase
Solution Approach 1:
The patent changes the parameter being transmitted from raw data to processed features with associated uncertainty metrics. This parameter transformation allows the cloud to make informed decisions about further processing based on uncertainty levels, achieving accurate results with minimal data transmission.
Solution Approach 2:
The patent introduces an uncertainty assessment mechanism as an intermediary between edge processing and cloud processing. This intermediary evaluates feature quality and determines the necessary level of data transmission, acting as a gatekeeper that prevents unnecessary data transfer while ensuring accurate processing when needed.
3Manufacturing precision
If the edge performs comprehensive processing, then the processing quality improves, but the edge computing resource constraints are violated
Solution Approach 1:
The patent implements partial processing at the edge by performing feature extraction and uncertainty assessment, but not complete processing. This partial action approach allows the edge to contribute meaningfully to processing quality while avoiding the computational burden of comprehensive processing, with the cloud completing the remaining work.
Data Source
AI summary
A cloud-edge collaborative data processing method and system, a device, and a storage medium are provided, relating to the field of telecommunication technology. The method includes extracting, in an edge encoder, intermediate features of data to be processed, inputting the intermediate features into a lightweight edge decoder for decoding to obtain an edge decoding result; and in response to a feature uncertainty of the edge decoding result being less than or equal to a preset threshold, taking the edge decoding result as a target processing result, or in response to the feature uncertainty being greater than the preset threshold, compressing the intermediate features using an edge compression model and then sending the compressed intermediate features to a cloud server in the cloud-edge collaborative data processing system.