Serverless Edge Computing Mutual Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Multi-access Edge Computing (MEC) systems face long application relocation times due to centralized decision-making, which can lead to service failures when user terminal mobility is high, and have limited resource utilization due to scarce resources at the network edge.
Innovation Solution
A serverless computing method and apparatus that enables communication between network edges without a central control unit, using event handlers, event generating modules, and auto-scalers to monitor and scale applications based on event responses, and resource availability, allowing for efficient application relocation and resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If centralized control (MEO) is used for application relocation decisions, then system management is simplified, but application relocation time increases significantly
Solution Approach 1:
The patent divides the centralized control function into distributed event handlers at each network edge. Each edge independently monitors events and makes relocation decisions locally, eliminating the need for centralized MEO coordination and significantly reducing relocation time.
Solution Approach 2:
The patent combines event monitoring, event filtering, and application scaling functions into an integrated serverless computing framework at the network edge, allowing simultaneous optimization of both management simplicity and response speed.
2Reliability
If applications are continuously running at network edges, then service availability is maintained, but resource scarcity limits system capacity
Solution Approach 1:
The patent implements event-triggered periodic action where applications are scaled or activated only when specific events occur, rather than running continuously. This allows services to be available when needed while conserving limited network edge resources during idle periods.
Solution Approach 2:
The patent dynamically changes application scaling parameters based on event conditions and resource availability, adjusting the number of running instances or activation state to match actual service demands and available resources.
Data Source
AI summary
The present invention relates to a serverless computing method and apparatus, and more particularly, to a serverless computing method and apparatus using mutual monitoring between network edges. The serverless computing method according to one embodiment of the present invention comprises monitoring, by an event handler of the first network edge, an event occurring in the second network edge, generating, by an event generating module of the second network edge, an event response signal through filtering on the event in response to the event occurring, transmitting, by an event delivery module of the second network edge, the event response signal to the event handler of the first network edge, and scaling an application, by an auto-scaler, in the first network edge based on the event response signal.