基于VNF间性能干扰的服务请求调度策略  

Service Request Scheduling Strategy Based on VNF Performance Interference

作　　者：郭胜 史久根[1] 孙立 谢熠君 GUO Sheng;SHI Jiu-gen;SUN Li;XIE Yi-jun(School of Computer and Information,Hefei University of Technology,Hefei 230601,China)

机构地区：[1]合肥工业大学计算机与信息学院,安徽合肥230601

出　　处：《计算机技术与发展》2021年第1期142-148,共7页Computer Technology and Development

基　　金：国家重大科学仪器设备开发专项(2013YQ030595)

摘　　要：网络功能虚拟化(NFV)技术是当今研究的热点技术之一。目前的虚拟网络功能(VNF)放置方法大都忽视了同位间VNF即处在同一个物理机中的虚拟网络功能,对硬件资源的竞争所形成的干扰问题,会导致网络吞吐量下降。针对此问题,该文建立了以最大化网络吞吐量为目标的混合整数线性规划模型,设计了一个两步调整策略。第一步设计了组合放置算法(CAPA),将资源需求互补的虚拟网络功能进行组合放置;第二步根据不同虚拟网络功能对数据流量的处理特性,设计了流量感知算法(TAA)进行服务请求调度,进一步缓解了在大的数据流量场景下,底层硬件资源竞争更加激烈使得干扰增强的问题。实验结果表明,该文提出的两步调整策略,与忽视干扰因素的一般策略相比提高了网络吞吐量。The network function virtualization(NFV)is one of the hot technologies of today’s research.The current virtual network function(VNF)placement method mostly ignores the interference caused by the competition of hardware resources of the VNF between peers in the same physical machine,which leads to the decrease of network throughput.To solve this problem,we establish a mixed integer linear programming model aiming at maximizing network throughput and design a two-step adjustment strategy.In the first step,we design the combination and placement algorithm(CAPA)to combine the VNFs with complementary resource requirements combined placement.In the second step,according to the processing characteristics of different VNFs on data traffic,the traffic aware algorithm(TAA)is designed to perform service request scheduling,which further alleviates the problem of interference enhancement caused by the scenario of large data traffic that the competition of underlying hardware resources is more fierce.The experiment shows that the proposed strategy makes the network performance increase compared with general strategy for ignoring interference factors.

关 键 词：网络功能虚拟化 性能干扰 网络吞吐量 组合放置 服务请求调度 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]