基于MTF启发法的OpenFlow虚拟流表高效查找算法

发布时间：2018-02-20 22:51

本文关键词： OpenFlow 软件交换机流表查找 MTF启发法　出处：《小型微型计算机系统》2017年07期 　论文类型：期刊论文

【摘要】：OpenFlow软件交换机由于成本低、配置灵活、且便于调度部署,因而广泛应用于软件定义网络试验床的构建,但在应用过程中遭遇到严重的性能瓶颈,流表查找是其中的关键一环.为此,本文应用启发式策略,提出一种面向软件交换机的OpenFlow虚拟流表高效查找算法.首先分析指出掩码遍历是影响流表查找性能的关键因素,进而基于网络流量局部性特点,应用M TF(M ove-To-Front)启发法自适应调整掩码顺序,以加快后续数据包的掩码查找命中速度.最后,借助实际网络流量样本,对本文所提算法的查找性能进行了实验评估.实验结果表明:本文所提算法的平均查找长度明显小于目前主流软件交换机中的流表查找方法,加快了OpenFlow流表查找速度,有效提升了OpenFlow软件交换机的数据转发性能.
[Abstract]:Because of its low cost, flexible configuration and convenient scheduling and deployment, OpenFlow software switch is widely used in the construction of software-defined network test bed, but it encounters a serious performance bottleneck in the process of application. Stream table lookup is one of the key links. In this paper, a efficient OpenFlow virtual stream table lookup algorithm for software switch is proposed by using heuristic strategy. Firstly, it is pointed out that mask traversal is the key factor affecting the performance of stream table lookup. Then, based on the local characteristics of network traffic, the M TF(M ove-To-Frontheuristic method is applied to adjust the order of mask codes to speed up the masking search and hit speed of subsequent data packets. Finally, with the help of actual network traffic samples, The experimental results show that the average search length of the proposed algorithm is obviously smaller than that of the current stream table lookup method in the mainstream software switch, and the speed of OpenFlow stream table lookup is accelerated. Effectively improves the OpenFlow software switch data forwarding performance.
【作者单位】：长沙理工大学计算机与通信工程学院;长沙理工大学综合交通运输大数据智能处理湖南省重点实验室;
【基金】：国家自然科学基金项目(61502056,61303043)资助湖南省自然科学基金项目(2015JJ3010,13JJ4052)资助湖南省教育厅科研项目(15B009)资助湖南省研究生科研创新项目(CX2017B487)资助;湖南省研究生科研创新项目(CX2017B487)资助
【分类号】：TP301.6;TP393.02

【相似文献】