摘要
随着多媒体新兴应用的不断涌现,网络规模日益复杂.为提高不同优先级多媒体业务的传输能力,保障业务的服务质量需求,结合软件定义网络技术,设计一种基于SDN的队列调度模型.同时,为了有效提高新型队列调度模型下数据的传输质量,避免产生网络拥塞,将复杂的网络抽象为M/M/1和M/D/1排队模型,并提出一种基于SDN的排队论时延模型,分析了新模型下MLFQ分组调度算法并对不同分组调度算法性能进行对比分析.仿真结果表明,基于SDN的多媒体流QoS队列调度机制在满足网络不同多媒体业务优先级要求的基础上,降低了数据传输时延和丢包率,增加了链路带宽利用率.
ed into M/M/1 and M/D/1 queuing models,and a queueing theory delay model based on SDN was proposed.The Multi-Level Feedback Queue(MLFQ)group scheduling algorithm under the new model was introduced and analysed,and the performance of different group scheduling algorithms was compared.The simulation results show that the SDN-based QoS queue scheduling mechanism for multimedia streams reduces data transmission delay and packet loss and increases link bandwidth utilisation while meets the priority requirements of different multimedia services in the network.
作者
崔峻玮
翟亚红
CUI Junwei;ZHAI Yahong(School of Electrical&Information Engineering,Hubei University of Automotive Technology,Shiyan 442002,China)
出处
《湖北大学学报(自然科学版)》
CAS
2023年第3期404-413,共10页
Journal of Hubei University:Natural Science
基金
湖北省教育厅科研计划重点项目(D20211802)
湖北省科技厅计划项目(2022BEC008)资助。
关键词
软件定义网络
服务质量
排队论
队列调度
software-defined networking
quality of service
queueing theory
queue scheduling