摘要
为满足新兴网络业务日益增长的低时延服务需求,同时考虑到网络服务建设过程中愈来愈高的成本问题,提出一种资源高效的时延敏感型服务功能链的部署算法。首先,基于资源优化模型的路径发现模块获取当前虚拟网络中满足服务请求的全部备选路径。其次,竞争拍卖模块基于路径的时延表现和资源配置对备选路径进行估价排序,通过竞拍流程平衡服务请求的时延要求与底层资源占用间的关系,实现了时延敏感型服务对底层资源利用率的提升。仿真结果表明,所提算法在保证服务低时延的前提下,用户接受率稳定在90%以上,对底层资源的利用率达75%以上。
In order to meet the growing requirements for low latency in emerging network services,while considering the increasing cost of network services,a resource-efficient deployment algorithm for latency-sensitive service function chains is proposed.First,the path discovery module based on the resource optimization model obtains all the alternative paths that satisfy the service requests in the current virtual network.Then,the competitive auction module evaluates and sorts the alternative paths based on the delay performance and resource allocation of the paths,and balances the relationship between the delay requirements of service requests and the underlying resource occupancy through the auction process.Simulation results show that the user acceptance rate of the proposed algorithm is more than 90%,and the utilization rate of the underlying resources is more than 75%under the premise of low service delay.
作者
赵季红
季文君
曲桦
吴豆豆
ZHAO Jihong;JI Wenjun;QU Hua;WU Doudou(School of Communications and Information Engineering,Xi'an University of Posts and Telecommunications,Xi'an 710121,China;School of Electronic and Information Engineering,Xi'an Jiaotong University,Xi'an 710054,China)
出处
《西安邮电大学学报》
2021年第2期1-7,共7页
Journal of Xi’an University of Posts and Telecommunications
基金
国家自然科学基金项目(61531013)
国家重大专项项目(2018ZX03001016)。
关键词
网络功能虚拟化
软件定义网络
服务功能链
资源优化
低时延
拍卖模型
network function virtualization
software-defined network
service function chain
resource optimization
low latency
auction model