Edge Computing Resource Allocation via User Instance Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Edge Computing (EC) systems face challenges in managing resource allocation effectively, particularly in scaling and handling user mobility, due to a lack of transparency between the EC management entity and EC applications regarding user assignment to application instances.
Innovation Solution
The proposed method involves an EC management entity detecting user requests, providing and managing application instances, and determining mappings between user equipment and application instances based on user assignment information, allowing for proactive scaling and mobility management by maintaining accurate user-instance mappings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If the EC management entity makes independent decisions on scaling and user assignment without application involvement, then the management process is simplified and automated, but the accuracy of resource allocation deteriorates due to lack of application-specific user grouping information
Solution Approach 1:
The patent introduces an intermediary mechanism where the EC application provides user assignment information (indicating which users belong to the same group) to the EC management entity. This intermediary information flow allows the management entity to make automated decisions while being guided by application-specific knowledge, thus maintaining both automation and allocation accuracy.
2Measurement precision
If the EC application manages user distribution independently with full logic control, then the allocation accuracy improves through application-specific knowledge, but the system complexity increases due to duplicated management logic
Solution Approach 1:
The patent extracts only the essential user assignment information from the EC application logic and provides it to the EC management entity. The complex decision-making logic remains in the management entity, while the application only needs to provide user grouping data. This extraction approach maintains allocation accuracy without duplicating complex management logic in the application.
3Speed
If the EC management entity scales resources without accurate user-instance mapping information, then the scaling speed increases, but the reliability of resource allocation deteriorates due to incorrect user assignment
Solution Approach 1:
The patent implements preliminary action by having the EC application provide user assignment information to the EC management entity before scaling operations are executed. This advance provision of user grouping data enables the management entity to perform accurate resource allocation and user assignment during scaling, maintaining both speed and reliability.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A method for managing resource allocation in an Edge Computing, EC, system (30), carried out in an EC management entity (300) connected to at least one EC host (220). The method comprises detecting (501) a request from a user equipment, UE, (303), which request is associated with an EC application; providing (502) at least one application instance of said EC application in a host of the EC system; receiving (503) user assignment information from the EC application; and determining (504) a mapping between the UE and a specific application instance based on the received user assignment information.