计算机集成制造系统 ›› 2019, Vol. 25 ›› Issue (第4): 954-961.DOI: 10.13196/j.cims.2019.04.018

• 当期目次 • 上一篇    下一篇

移动边缘计算中能耗优化的多重资源计算卸载策略

徐佳1,李学俊1,丁瑞苗1,刘晓2   

  1. 1.安徽大学计算机科学与技术学院
    2.迪肯大学信息技术学院
  • 出版日期:2019-04-30 发布日期:2019-04-30
  • 基金资助:
    国家自然科学基金资助项目(61672034,61300042);安徽省自然科学基金资助项目(1708085MF160)。

Energy efficient multi-resource computation offloading strategy in mobile edge computing

  • Online:2019-04-30 Published:2019-04-30
  • Supported by:
    Project supported by the National Natural Science Foundation,China(No.61672034,61300042),and the Natural Science Foundation of Anhui Province,China(No.1708085MF160).

摘要: 移动边缘计算环境中边缘设备的能耗优化主要采用计算卸载策略。然而目前常用的计算卸载策略大多只考虑单一的计算资源,没有对移动边缘计算环境中不同种类的计算资源进行综合考虑,无法在保证响应时间约束的情况下充分降低边缘设备能耗。为了解决这一问题,在移动边缘计算环境中提出一种多重资源计算卸载能耗模型,设计了一种新的评价边缘设备能耗的适应度计算方法,并结合工作流管理系统提出了移动边缘计算中能耗优化的多重资源计算卸载粒子群任务调度算法,该算法能够在考虑响应时间约束的情况下,充分降低移动终端能耗。实验表明,与已有4种计算卸载策略相比,新策略所对应的任务调度算法收敛稳定、适应度最优,在用户响应时间约束下,任务调度方案的边缘设备能耗值优于其他4种卸载策略。

关键词: 移动边缘计算, 工作流调度, 能耗优化, 计算卸载, 多重资源

Abstract: In the research on energy efficiency optimization of mobile edge computing,the computation offloading strategy of edge device is emphasis.However,the existing computation offloading strategy can only consider single computing resource and do not take into account the different type of computing resources in mobile edge computing,which cannot reduce the energy consumption of edge device with response time constraint.Therefore,a energy model of multi-resources computation offloading was proposed,and the fitness computation method of task scheduling plan was designed to evaluate the energy consumption of edge device.An energy efficient multi-resource computation offloading strategy task scheduling algorithm was presented to solve the energy consumption optimization problem of edge device.Experimental results showed that the propose algorithm could always achieve stable convergence speed,the optimal fitness and low energy consumption of edge device with the constraint of response time.

Key words: mobile edge computing, workflow scheduling, energy efficient, computation offloading, multi-resource

中图分类号: