基于蚁群算法的SDN数据中心负载均衡流调度
Load balancing flow scheduling for SDN data center based on ant colony algorithm
陈永聪 1陈秋莲 1王成栋1
作者信息
- 1. 广西大学计算机与电子信息学院,广西南宁 530004
- 折叠
摘要
软件定义网络(SDN)中大小流共存,难以保障不同数据流对网络服务质量的不同约束要求.为此提出一种基于蚁群算法(ACO)的SDN数据中心网络动态流量调度机制,根据SDN网络的状态信息,建立链路综合评价机制,改进蚁群算法的信息素更新方式,实现大象流和老鼠流差异化调度.实验结果表明,算法增强了对突发大流量的应对能力,保障了不同类型流量的公平传输.应用于银行业,有效保障了数据中心网络的平稳运行.
Abstract
For software defined network(SDN),the coexistence of large and small flows in data center network makes it difficult to guarantee different constraints on network service quality required by different data flows.A dynamic traffic scheduling mecha-nism for SDN data center network based on ant colony algorithm(ACO)was proposed.According to the state information of SDN network,a comprehensive link evaluation mechanism was established,and the pheromone update method of ACO was improved to realize the differentiated scheduling of elephant flow and mouse flows.Experimental results show that,the algo-rithm enhances the ability to cope with sudden heavy traffic and guarantees the fair transmission of different types of traffic.In the practical application of banking,the smooth operation of data center network is effectively ensured.
关键词
软件定义网络/数据中心/蚁群算法/网络状态/链路综合评价/路由算法/差异化调度Key words
software defined network/data center network/ant colony algorithm/network state/link comprehensive evalua-tion/routing algorithm/differential scheduling引用本文复制引用
基金项目
国家自然科学基金项目(71371058)
广西自然科学基金项目(2020GXNSFAA159090)
广西大学基金项目(XBZ200371)
出版年
2024
NETL
NSTL
万方数据