物联网学报 ›› 2021, Vol. 5 ›› Issue (2): 60-70.doi: 10.11959/j.issn.2096-3750.2021.00225

所属专题: 边缘计算

• 专题:物联网边缘智能与雾计算技术 • 上一篇    下一篇

基于最优运输理论的物联网边缘计算资源优化机制

张琪1, 蒋宇娜1, 葛晓虎1, 李永会2   

  1. 1 华中科技大学,湖北 武汉 430074
    2 澳大利亚悉尼大学,澳大利亚 悉尼 NSW
  • 修回日期:2021-02-03 出版日期:2021-06-30 发布日期:2021-06-01
  • 作者简介:张琪(1997- ),女,华中科技大学硕士生,主要研究方向为无线通信、环境智适应网络柔性传输理论和边缘计算
    蒋宇娜(1994- ),女,华中科技大学博士生,主要研究方向为无线通信、区块链和物联网
    葛晓虎(1972- ),男,博士,华中科技大学教授,主要研究方向为移动通信、无线网络中的流量建模、绿色通信等
    李永会(1975- ),男,博士,澳大利亚悉尼大学教授,主要研究方向为无线通信、物联网、无线AI等
  • 基金资助:
    国家自然科学基金资助项目(U2001210)

Resource allocation based on optimal transport theory in IoT edge computing

Qi ZHANG1, Yuna JIANG1, Xiaohu GE1, Yonghui LI2   

  1. 1 Huazhong University of Science and Technology, Wuhan 430074, China
    2 The University of Sydney, Sydney NSW, Australia
  • Revised:2021-02-03 Online:2021-06-30 Published:2021-06-01
  • Supported by:
    The National Natural Science Foundation of China(U2001210)

摘要:

随着物联网和边缘计算的发展,物联网设备可以将计算密集型任务卸载到边缘计算服务器上进行处理。由于物联网设备分布以及计算需求的变化,需要对边缘计算资源进行动态管理。利用最优运输理论对物联网中计算资源分配进行优化,提出一种基于物联网设备分布和边缘计算服务器位置的区域优化划分机制,在边缘计算服务器计算能力的约束下对物联网设备的能耗以及时延性能进行优化。仿真结果表明,与传统泰森多边形划分机制相比,该优化机制有更好的均衡性,并且物联网设备的平均能耗最多降低21%,平均时延最多降低45%。

关键词: 物联网, 边缘计算, 资源分配, 最优运输理论, 能耗, 时延

Abstract:

With the development of the Internet of things (IoT) and edge computing, the computation-intensive tasks of IoT devices can be offloaded to edge devices and processed at the edge of networks.Due to the variation of the distribution and computation requirements of IoT devices, the computation resources of edge networks need to be managed dynamically.The optimal transport theory was adopted to optimize the computation resources allocation in IoT networks.An optimized regional partition mechanism was proposed based on the distribution of IoT devices and locations of edge computing devices.Under constraints on the computing capabilities of edge computing devices, the energy consumption and delay of IoT devices were optimized.The simulation results show that, compared with the traditional Voronoi partition scheme, the proposed optimization mechanism shows better balance.The average transmitting power can be reduced by 21% and the average delay can be reduced by 45%.

Key words: Internet of things, edge computing, resource allocation, optimal transport theory, energy consumption, delay

中图分类号: 

No Suggested Reading articles found!