Edge computing service migration method based on multi-objective optimization strategy

Abstract

An edge computing service migration method based on a multi-objective optimization strategy is characterized in that a challenge for deploying mobile edge computing service in a cellular network is to support mobility of a user, and particularly, when the moving speed is relatively high, an unloaded task can be transferred to the mobile edge computing service under the condition that the resource utilization efficiency and the link reliability are not influenced. Seamless migration is carried out between the base stations. In a service migration scene, three problems need to be solved: when user equipment passes through an adjacent edge computing server, whether a virtual machine corresponding to the user equipment needs to be migrated is solved; if the virtual machine decides to migrate, the virtual machine should migrate to which edge computing server; and how to find the optimal communication path between the user equipment and the corresponding virtual machine. The problems of virtualization, I / O interference between virtual machines, multi-user access interference and the like are considered, a novel method based on relaxation and rounding is provided, the overall service quality is improved to the maximum extent, and the migration cost is minimized. Experimental results show that the method can make an optimal decision in a real scene.

Description

technical field [0001] The invention belongs to the field of the Internet of Things, and in particular relates to an edge computing service migration method based on a multi-objective optimization strategy. Background technique [0002] Edge computing can be called a promising computing paradigm because it can accelerate almost all mainstream mobile applications (such as facial recognition, etc.). In practical applications, edge computing is generally integrated into the Internet of Things platform. In an edge computing environment, a base station is equipped with a certain amount of computing resources and can provide computing services for mobile users within the service range. When the amount of data that needs to be processed in the IoT device is too large, in order to ensure low latency, the large amount of data generated can be transferred to the "edge" of the network for computing and processing. The edge computing paradigm is not a remote central cloud processing d...

AI Technical Summary

Problems solved by technology

Due to the constraints of limited resources, on the one hand, judging from the perspective of users, when many users request allocation, the edge server at a short distance cannot provide services for all users, and many tasks will be queued in the server, which will reduce the quality of service (Quality On the other hand, judging from the perspective of edge servers, there is no superior resource allocation strategy (such as too many or too few resources requested by users, and some servers are idle for a long time) will waste a lot of energy on the edge computing platform

Images