摘要
考虑边缘计算网络中的负载均衡需求,提出一种基于集中控制的调度机制。首先决定在哪些网络节点部署边缘计算功能,再针对用户的数据和请求,在满足相关负载均衡约束的前提下通过调度尽量降低流量的平均端到端延迟。评估结果表明,边缘计算节点的数量、计算资源和网络资源的负载均衡程度均可能影响流量的平均端到端延迟。只需选择少量合适的节点作为边缘计算节点,再将计算资源和网络资源的负载均衡调配到合适程度即可有效降低平均端到端延迟。
Considering the load balancing needs in edge computing network,this paper proposed a scheduling scheme based on centralized control.This scheme decided which nodes to deploy edge computing module firstly,then considered users’data and requests,minimized the average end-to-end delay of traffic through scheduling while meeting the load balancing constraints.The evaluation results indicate that the number of edge computing nodes and the load balancing level of computing and network resources may affect the average end-to-end delay of traffic.Selecting a small number of suitable nodes as edge computing nodes and making the load balancing level of computing and network resources appropriate can effectively reduce the average end-to-end delay.
作者
董谦
马宇翔
李俊
Dong Qian;Ma Yuxiang;Li Jun(Computer Network Information Center,Chinese Academy of Sciences,Beijing 100190,China;University of Chinese Academy of Sciences,Beijing 100049,China;Foshan University,Foshan Guangdong 528000,China)
出处
《计算机应用研究》
CSCD
北大核心
2020年第3期856-859,867,共5页
Application Research of Computers
基金
国家重点研发计划资助项目(2017YFB1401500)
国家自然科学基金资助项目(61672490)
中国科技云建设工程(Y72923)。
关键词
边缘计算
集中控制
负载均衡
分段路由
edge computing
centralized control
load balancing
segment routing
