Adaptive edge server placement and dynamic adjustment optimization method
By using reinforcement learning algorithms in the edge computing system, intelligently adjusting the server location, the problem of high computing cost of server placement strategies in large-scale dense base stations is solved, and the effect of load balancing and delay reduction is achieved.
CN119653386AActive Publication Date: 2025-03-18ZHENGZHOU UNIV
8 Cites 0 Cited by
Patent Information
- Application Number
- CN202411546433.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-01
- Publication Date
- 2025-03-18
- Estimated Expiration
- 2044-11-01
AI Technical Summary
Technical Problem
The existing server placement strategy is costly and difficult to achieve real-time dynamic adjustment when dealing with large-scale dense base stations, resulting in latency and load balancing problems.
Method used
The reinforcement learning algorithm is used to model the locations of each server in the system as a state. The agent makes real-time decisions based on the system state, merges server selection and movement into one action, and selects the optimal server movement location through the distance matrix to achieve adaptive dynamic adjustment.
Benefits of technology
It effectively reduces system latency and load, provides a load balancing strategy, and reduces computing costs, and realizes adaptive server placement and dynamic adjustment.
✦ Generated by Eureka AI based on patent content.
Abstract
The invention provides a self-adaptive edge server placement and dynamic adjustment optimization method, which comprises the following steps of: firstly, acquiring the positions of all base stations in a region, and randomly generating the positions of all servers; secondly, the intelligent agent generates state features according to the dynamic position information of the server; then the intelligent agent determines a server to be selected to move and a moving direction according to the state characteristics; and calculating a distance matrix of all base stations in the direction, and obtaining a position to which the server needs to move according to the distance matrix. And finally, according to the moved server information and the dynamic base station information, calculating the delay and load in the current system as rewards, and feeding back and updating the decision network of the intelligent agent. The method is suitable for the problem of dynamic server placement in a mobile edge computing scene, the position of the server is dynamically adjusted by sensing the real-time base station characteristics, and the delay of the system is reduced while an effective load balancing strategy is provided.
Need to check novelty before this filing date? Find Prior Art