摘要
文章提出了一种基于分布式约束优化问题(Distributed Constraint Optimization Problem,DCOP)框架的地理分布数据中心(Geographically Distributed Data Center,GDDC)分布式优化方法,开发了一个新的GDDC模型DCOP。其中,每个中数中心(Data Centers,DC)算子由一个代理表示。其次,由于传统的DCOP方法不适用于这类大规模问题,每个代理赋予多个变量和全局约束。经试验,结果表明了新方法的优点。
The goal of this article is to optimize workload allocation between a set of DCs to minimize energy costs while respecting limitations such as data center capacity and virtual machine migration.This article proposes a GDDC distributed optimization method based on the distributed constraint optimization(DCOP)framework.A new GDDC model DCOP has been developed,where each DC operator is represented by a proxy.Secondly,due to the traditional DCOP method not being suitable for such large-scale problems,each agent is assigned multiple variables and global constraints.The experimental results demonstrate the advantages of the new method.
作者
陈一航
李卓晖
李逸谦
杨子江
裴欢
Chen Yihang;Li Zhuohui;Li Yiqian;Yang Zijiang;Pei Huan(China Energy Construction Group Guangdong Electric Power Design and Research Institute Co.,Ltd.,Guangzhou 510663,China)
出处
《无线互联科技》
2023年第22期128-130,164,共4页
Wireless Internet Technology
关键词
跨地理分布
数据中心
优化分析
全局约束
cross geographical distribution
data center
optimization analysis
global constraints
