[1]何利文,唐澄澄,周 睿,等.基于遗传算法的 SDN 增强路径装箱问题研究[J].计算机技术与发展,2019,29(07):150-154.[doi:10. 3969 / j. issn. 1673-629X. 2019. 07. 030]
　HE Li-wen,TANG Cheng-cheng,ZHOU Rui,et al.Research on Packing Problem of SDN Path Enhancement Based on Genetic Algorithm[J].,2019,29(07):150-154.[doi:10. 3969 / j. issn. 1673-629X. 2019. 07. 030]
点击复制

基于遗传算法的 SDN 增强路径装箱问题研究()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
29
期数:
2019年07期
页码:
150-154
栏目:
应用开发研究
出版日期:
2019-07-10

文章信息/Info

Title:
Research on Packing Problem of SDN Path Enhancement Based on Genetic Algorithm
文章编号:
1673-629X(2019)07-0150-05
作者:
何利文1; 2 ; 唐澄澄1 ; 周   睿1 ; 侯小宇1
1. 南京邮电大学,江苏 南京 210003; 2. 中兴通讯股份有限公司,江苏 南京 210012
Author(s):
HE Li-wen 1; 2 ; TANG Cheng-cheng 1 ; ZHOU Rui 1 ; HOU Xiao-yu 1
1. Nanjing University of Posts and Telecommunications,Nanjing 210003,China; 2. ZTE Corporation,Nanjing 210012,China
关键词:
SDN 网络; 流量工程; 遗传算法; 装箱问题
Keywords:
software defined network; traffic engineering; genetic algorithm; packing problem
分类号:
TP301
DOI:
10. 3969 / j. issn. 1673-629X. 2019. 07. 030
摘要:
在 SDN 路由器的网络环境中,网络流量负载不均衡时,网络中部分线路带宽利用率过高,其余网络资源利用率过低,导致无法添加新的业务,系统负载下降。 在现有的网络拓扑结构下,为优化路由选择的路径,找到流量分配的最优方案,需要对网络实施流量工程。 文中研究了基于 SDN 网络中的这一流量工程,增强路径装箱问题。 传统的路由器解决方案中,路由的选择是通过以跳数,延迟,所占带宽资源与链路代价为量度的最短路径优先算法实现的,没有解决多约束的问题。 为解决该问题,文中主要对比目前国外瞻博网络公司和诺基亚贝尔实验室对网络装箱问题的研究算法,分析了它们性能的优势,并提出了基于遗传算法的 SDN 增强路径装箱方法。 分析验证表明,该算法不仅可以解决多约束条件下的装箱问题,调整已有的业务所在的通信链路,部署新的业务,而且能够实现调整业务所在的通信链路时对系统的扰动最小,有效优化网络的负载,提高网络资源的利用率。
Abstract:
In the network environment of the SDN router,when the network traffic load is not balanced,the utilization ratio of some lines in the network is too high,and that of other network resources is too low,which leads to the failure to add new services and the decrease of system load. In the existing network topology, in order to optimize the routing path and find the optimal solution for traffic distribution,traffic engineering needs to be implemented on the network. We study the traffic engineering based on the SDN network and the path packing enhancement. In the traditional router solutions,routing is realized by the shortest path priority algorithm measured by the number of hops,delay,bandwidth resource and link cost,without solving the problem of multiple constraints. In order to solve this problem,we mainly compare the current algorithms of network packing problems between Juniper Networks and Nokia Bell Labs,analyze the advantages of their performance,and propose an enhanced path packing method based on genetic algorithms. Analysis and verification show that the proposed algorithm can not only solve the problem of packing under multiple constraints,adjust the existing communication links and deploy new services,but also realize the minimum disturbance to the system when adjusting the communication links,effectively optimize the network load and improve the utilization rate of network resources.

相似文献/References:

[1]李蓬 黄河.基于NS2的MPLS流量工程仿真研究[J].计算机技术与发展,2008,(09):53.
　LI Peng,HUANG He.Simulation Research of MPLS Traffic Engineering Based on NS2[J].,2008,(07):53.
[2]孙自翔 李兴明.基于MPLS TE的多光纤带宽适配算法[J].计算机技术与发展,2011,(08):71.
　SUN Zi-xiang,LI Xing-ming.Bandwidth Adaption Algorithms for Multi-fiber Based on MPLS TE[J].,2011,(07):71.
[3]陈胜 李兴明 陈捷.PTN网络中基于Diff-Serv TE的QoS研究[J].计算机技术与发展,2011,(09):128.
　CHEN Sheng,LI Xing-ming,CHEN Jie.QoS Research Based on Diff-Serv TE in PTN Network[J].,2011,(07):128.

更新日期/Last Update: 2019-07-10