软件定义网络资源发现及组网应用研究

发布时间：2018-06-22 19:35

本文选题：软件定义网络 + 资源发现　；参考：《北京邮电大学》2014年硕士论文

【摘要】：软件定义网络(Software Defined Network, SDN)是一种新型网络创新架构。该架构的核心技术OpenFlow通过将网络设备控制层面和数据转发层面分开来实现了网络流量的灵活控制,并为核心网络及不同应用的创新提供了良好的网络架构。网络资源发现是网络技术的重要组成部分。针对软件定义网络的网络资源发现研究还为数不多。网络资源发现核心就是根据资源请求者的资源请求描述,在网络上为请求者找到满足请求描述要求的合适资源的过程。本文提出了一种新的集中式网络资源发现方式。新的网络资源发现方式能够很好的适应软件定义网络集中控制的理念。仿真实验的结果表现新的资源发现方式在控制器上工作正常。针对采用OpenFlow协议栈的网络设备进行组网应用时流表下发速度较慢的情况,对TCAM匹配项更新效率问题进行研究。结合前缀长度顺序限制算法和简单更新算法进行对比分析,并提出改进的更新算法：动态前缀长度顺序限制算法。通过理论分析和实验仿真结论：动态前缀长度顺序限制算法在时间复杂度上明显优于简单更新算法,略低于前缀长度顺序限制算法,动态前缀长度顺序限制算法在空间复杂度上要明显优于前两者。新算法降低了TCAM的功耗,同时提高了TCAM的存储空间的利用率。
[Abstract]:Software defined Network (SDN) is a new network innovation architecture. OpenFlow, the core technology of this architecture, realizes the flexible control of network traffic by separating the control layer of network devices from the layer of data forwarding, and provides a good network architecture for the innovation of core networks and different applications. Network resource discovery is an important part of network technology. There are few researches on network resource discovery for software defined network. The core of network resource discovery is the process of finding the appropriate resource for the requester on the network according to the resource request description of the resource requester. In this paper, a new centralized network resource discovery method is proposed. The new network resource discovery method can well adapt to the concept of software definition of network centralized control. The simulation results show that the new resource discovery method works well on the controller. The efficiency of updating TCAM matches is studied in view of the slow sending speed of flow table in the application of network devices using OpenFlow protocol stack. Combining prefix length sequence restriction algorithm with simple updating algorithm, an improved updating algorithm, dynamic prefix length sequence limitation algorithm, is proposed. Through theoretical analysis and experimental simulation, it is concluded that the dynamic prefix length sequence restriction algorithm is obviously superior to the simple updating algorithm in time complexity, and is slightly lower than the prefix length sequence restriction algorithm. Dynamic prefix length sequence constraint algorithm is superior to the first two algorithms in space complexity. The new algorithm reduces the power consumption of TCAM and improves the utilization of storage space of TCAM.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.02

【参考文献】