VM migration strategy minimizing the task unloading cost in the mobile edge computing environment

Abstract

The invention relates to a VM migration strategy under a mobile edge computing environment in a 5G network, and belongs to the technical field of mobile communication. The service migration strategy in the existing research neglects task unloading cost in QoS requirements of users, and task division in the existing research is mostly a coarse-grained model or a fine-grained chain model and is notsuitable for complex association scenes among subtasks. Problem for migration decisions, According to the method, fine-grained directed acyclic graph division is performed on tasks, a task time delaymodel and an unloading cost model are established, the VM migration decision problem of minimizing task unloading cost under the time delay constraint condition is constructed, and finally, migrationdecision is performed on the basis of a genetic algorithm to obtain a virtual machine migration decision result of each subtask. The method and the device have important significance for ensuring theservice continuity of user task unloading and meeting QoS requirements.

Description

technical field [0001] The invention belongs to the technical field of mobile communication, and relates to a virtual machine (Virtual Machine, VM) migration strategy for minimizing task offloading costs in a mobile edge computing environment. Background technique [0002] With the rapid development of mobile communication technology in recent years, in addition to mobile phones and computers, mobile network service terminal devices have added many new business scenarios, such as augmented reality (Augment Reality, AR), online games, in-vehicle Internet of Things, autonomous driving, etc. In addition, there are IoT business scenarios that are close to life, such as smart agriculture, smart grid, and environmental monitoring. In response to the rapid development of the mobile Internet and the Internet of Things, 5G needs to meet new business requirements such as ultra-low latency, ultra-low power consumption, ultra-high reliability, and ultra-high-density connections. Mobile...

AI Technical Summary

Problems solved by technology

However, due to user movement, it is necessary to migrate the virtual machine containing computing task information from one MEC server to another MEC server to ensure the continuity of user services. At this time, we are faced with a migration decision problem.

[0003] The service migration strategy in existing research focuses on factors such as user energy consumption, MEC server load, VM migration time, and total task execution delay, ignoring the task offloading cost in QoS requirements

Moreover, most of the task division in existing research is a coarse-grained model (the entire mobile terminal task is migrated as a whole) or a fine-grained chain model (a mobile application is composed of a set of subtasks in a linear topological order, each subtask Tasks are executed in sequence), not suitable for scenarios with complex associations with subtasks

Images