云边协同下的虚拟网络功能部署和流量调度方法

Virtual Network Function Deployment and Traffic Scheduling Method in Cloud-side Collaboration

基于网络功能虚拟化(Network Function Virtualization,NFV)软硬件解耦的网络架构,运营商通过部署多个端到端逻辑网络为垂直行业提供多样化服务.然而,面对时延敏感型服务,需要将部分虚拟网络功能(Virtual Network Function,VNF)下沉到网络边缘.基于此,多接入边缘计算(Multi-Access Edge Computing,MEC)技术应运而生.但是,与核心云相比,边缘云的资源容量和计算能力有限,因此,在部署网络服务时既要考虑云边协同下的资源分配问题,还要考虑如何进行流量调度以提升边缘云的资源利用率并解决底层网络中的负载均衡问题.鉴于此,本文提出了一种基于动态流量拆分的VNF部署和流量调度方法.通过动态流量拆分将流量请求分配到多条路径和多个节点,从而使全局流量分布的更加均衡,根据流量拆分结果进行VNF部署,进而减少发生网络拥塞和流量请求超时的概率.其中,每条流量请求的拆分条数与拆分比例可以根据底层网络的资源状态和流量的特性灵活调整.因此,整个过程包括有效的流量拆分策略、VNF部署策略和流量调度策略.本文将云边协同下的虚拟网络功能部署和流量调度问题描述为混合整数线性规划(Mixed Integer Linear Programming,MILP)问题,并提出了一种基于禁忌搜索算法和遗传算法的联合优化(TSGA)算法.仿真结果表明,TSGA算法相比于TS算法和RB算法可以分别提高7.7%和12.9%的流量请求接受率,并分别减少5.4%和7.8%的流量平均开销.

Based on the network architecture of network function virtualization(NFV)decoupling software and hardware,operators can provide diversified services for vertical industries by deploying multiple end-to-end logical networks.However,for latency-sensitive services,some virtual network functions(VNFs)need to be sunk to the network edge.Based on this,the multi-access edge computing(MEC)technology came into being.However,compared with the core cloud,the resource capacity and computing power of the edge cloud are limited.Therefore,when deploying network services,it is necessary to consider not only the problem of resource allocation under cloud-side collaboration,but also how to perform traffic scheduling to improve the resource utilization of the edge cloud and solve the load balancing problem in the underlying network.In view of this,this paper proposes a VNF deployment and traffic scheduling method based on dynamic traffic spltting.Dynamic traffic spliting refers to distributing traffic requests to multiple paths and nodes,so that the global traffic distribution is more balanced,thereby reducing the probability of network congestion and traffic request timeouts.Among them,the number of splits and the split ratio of each traffic request can be flexibly adjusted according to the resource status of the underlying network and the characteristics of the traffic.Therefore,the whole process includes effective VNF deployment strategy,traffic spliting strategy and traffic scheduling strategy.This paper describes the virtual network function deployment and traffic scheduling problem under cloud-edge collaboration as a Mixed Integer Linear Programming(MILP)problem,and proposes a joint optimization(TSGA)algorithm based on the tabu search algorithm and genetic algorithm.The simulation results show that compared with the TS algorithm and the RB algorithm,the TSGA algorithm can increase the traffic request acceptance rate by 7.7%and 12.9%,and reduce the average traffic overhead by 5.4%and 7.8%,respectively.