计算机工程与应用 ›› 2023, Vol. 59 ›› Issue (15): 55-67.DOI: 10.3778/j.issn.1002-8331.2210-0369

• 热点与综述 • 上一篇    下一篇

云边端和D2D边缘架构下的计算卸载策略综述

陈琳,董甲东,潘凯,王兆祥   

  1. 安庆师范大学 电子工程与智能制造学院,安徽 安庆 246011
  • 出版日期:2023-08-01 发布日期:2023-08-01

Survey of Computation Offloading Strategies Under Cloud-Edge-End and D2D Edge Architectures

CHEN Lin, DONG Jiadong, PAN Kai, WANG Zhaoxiang   

  1. School of Electronic Engineering and Intelligent Manufacturing, Anqing Normal University, Anqing, Anhui 246011, China
  • Online:2023-08-01 Published:2023-08-01

摘要: 随着工业互联网的兴起,边缘计算(edge computing)成为当前最热门的研究课题之一。计算卸载作为其关键技术,将用户的计算任务卸载到边缘服务器上进行处理,有效解决终端设备计算能力弱、时延长和能耗高等问题。从计算卸载步骤、计算卸载策略以及计算卸载方式三个方面介绍边缘卸载技术,简单介绍了云-边-端、边-端、多接入边缘和分布式D2D(device-to-device)边缘四种边缘架构,并分析总结计算卸载在中心式的云-边-端架构和分布式的D2D边缘架构两种不同边缘架构中的卸载策略和研究成果,对所研究的两种架构的优缺点进行总结。对计算卸载领域的未来研究方向进行思考,总结归纳了目前计算卸载策略在用户的移动性管理和资源的协同调度两个方面所面临的挑战。

关键词: 工业互联网, 边缘计算架构, 卸载策略, D2D

Abstract: With the rise of industrial internet, edge computing has become one of the hottest research topics. Computing offloading is a key technology of edge computing, which offloads user’s computing tasks to edge servers for processing, effectively solving the problems of weak computing power, prolonged computing time and high energy consumption of terminal devices. Firstly, the technology of edge offloading is introduced from three aspects:the procedure of computing offloading, the strategy of computing offloading and the mode of computing offloading. Secondly, the four edge architectures of cloud-edge-end, edge-end, multi-access edge and distributed device-to-device(D2D) edge are briefly introduced. The offloading strategies and research results of two different edge architectures, the central-type cloud-edge-end architecture and the distributed D2D edge architecture, are analyzed and summarized, and the advantages and disadvantages of the two architectures are summarized. Finally, the future research directions in the field of computing offloading are considered, and the challenges faced by current computing offloading strategies in user mobility management and resource co-scheduling are summarized.

Key words: industrial internet, edge computing architecture, offloading strategy, device-to-device(D2D)