移动边缘计算的移动性管理研究

发布时间:2018-02-11 作者:王秋宁,谢人超,黄韬

[摘要] 提出了移动边缘计算(MEC)中基于马尔可夫决策过程(MDP)的虚拟机迁移策略。该策略可以在保证用户体验的前提下,能使得系统成本达到最优。每次用户进入新的MEC区域后,都通过该策略决定是否进行迁移。通过与常规迁移策略对比,仿真结果表明本策略总能达到最优情况,有效地降低了系统成本。

[关键词] MEC;MDP;移动性管理;成本最优

[Abstract] In this paper, a virtual machine migration strategy based on Markov decision process (MDP) in mobile edge computing (MEC) is proposed, which can minimize the system cost while ensuring the user experience. After entering a new MEC area, the users can be determined whether or not to migrate by this policy. Compared with the conventional migration strategy, the simulation results show that this strategy can always achieve the optimal situation and effectively reduce the system cost.

[Keywords] MEC; MDP; mobility management; cost optimization

下载阅览: PDF