基于OpenFlow网络的QoS保障子系统研究
发布时间：2018-09-08 17:46
【摘要】：网络中日益涌现的多种应用均对服务质量(Quality of service,QoS)有严格的要求,研究者在过去的二十年里已经探索出许多QoS架构,但由于各种原因,这些QoS架构均未得到真正的推广使用。OpenFlow作为一种管控分离的新型网络交换模型,能够提供灵活的流定义,并且集中式的网络控制能够获取最新的全局网络状态。此外,OpenFlow能对网络设备进行瞬时管理以无缝适应端到端的网络行为,所以在OpenFlow上部署的任意QoS机制/框架都将能支持端到端QoS。目前,国内外基于OpenFlow网络的QoS体系架构研究还相对较少,这些研究主要利用资源预留、优先级队列以及QoS路由等为应用提供QoS保证。但是前两者会对无QoS要求的流产生较大影响,而当前的QoS路由研究并没有真正结合当前网络状态,所以计算的路径缺乏准确性。针对上述问题,本论文利用OpenFlow全局控制、支持统计信息收集的特点提出了一个QoS保障子系统,该QoS保障子系统主要通过收集网络实时的QoS参数信息,利用QoS路由技术为应用提供QoS保证。该QoS保障子系统包含五个重要模块:路径计算、网络测量、接入控制、流管理、拓扑管理。本文在FloodLight控制器开发平台中设计并实现了路径计算和网络测量这两个核心模块,并对拓扑管理模块进行了修改。路径计算用于计算满足用户QoS要求的路径,本文在研究了多种QoS路由算法后,采用NR_DCLC路由算法来为QoS应用计算优化的路径。网络测量用于收集当前的网络状态参数信息,它又分为三个子模块,分别是链路使用测量模块、丢包测量模块和时延测量模块。路径计算模块参照网络测量提供的QoS状态信息计算优化的QoS路由,网络测量也会在资源无法满足应用的QoS要求时通知路径计算模块重新选择路径。本文利用Mininet仿真平台,对QoS保障子系统进行了网络仿真与验证,仿真结果表明网络测量能够正确获取网络状态信息,基于本文的路径计算方法能够提高资源利用率并且在网络拥塞的情况下能实现动态重路由。本文所提出的新型QoS保障子系统虽然在仿真环境中实现,但是设计和实现的功能、代码可以运用于真实的网络环境中,只需要每个交换机支持OpenFlow协议。所以本文的工作对于理论研究和实际部署都有较大的意义。
[Abstract]:In the past two decades, researchers have explored a lot of QoS architecture, but for various reasons, None of these QoS architectures have been popularized. OpenFlow is a new network exchange model with separated control and can provide flexible flow definition and centralized network control can obtain the latest global network state. In addition, OpenFlow can manage network devices instantaneously to adapt end-to-end network behavior seamlessly, so any QoS mechanism / framework deployed on OpenFlow will be able to support end-to-end QoS. At present, there are few researches on QoS architecture based on OpenFlow network at home and abroad. These researches mainly use resource reservation, priority queue and QoS routing to provide QoS guarantee for applications. However, the former two will have a great impact on the flow without QoS requirements, and the current QoS routing research has not really combined the current network state, so the calculation path is not accurate. In order to solve the above problems, this paper proposes a QoS support subsystem based on the global control of OpenFlow and the characteristics of supporting the collection of statistical information. The QoS guarantee subsystem mainly collects the real-time QoS parameter information of the network. Use QoS routing technology to provide QoS guarantee for applications. The QoS support subsystem consists of five important modules: path calculation, network measurement, access control, flow management, and topology management. In this paper, two core modules of path calculation and network measurement are designed and implemented in the FloodLight controller development platform, and the topology management module is modified. Path calculation is used to calculate paths that meet the QoS requirements of users. After studying various QoS routing algorithms, the NR_DCLC routing algorithm is used to compute the optimized paths for QoS applications. Network measurement is used to collect the current network state parameter information, which is divided into three sub-modules: link usage measurement module, packet loss measurement module and delay measurement module. The path calculation module computes the optimized QoS route according to the QoS state information provided by the network measurement, and the network measurement notifies the path calculation module to re-select the path when the resource can not meet the QoS requirements of the application. In this paper, the network simulation and verification of QoS support subsystem are carried out by using Mininet simulation platform. The simulation results show that the network measurement can obtain the network state information correctly. The path calculation method based on this paper can improve resource utilization and realize dynamic rerouting in the case of network congestion. Although the new QoS support subsystem proposed in this paper is implemented in the simulation environment, the design and implementation of the functions and codes can be applied to the real network environment, only need each switch to support the OpenFlow protocol. Therefore, the work of this paper has great significance for theoretical research and practical deployment.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09
[Abstract]:In the past two decades, researchers have explored a lot of QoS architecture, but for various reasons, None of these QoS architectures have been popularized. OpenFlow is a new network exchange model with separated control and can provide flexible flow definition and centralized network control can obtain the latest global network state. In addition, OpenFlow can manage network devices instantaneously to adapt end-to-end network behavior seamlessly, so any QoS mechanism / framework deployed on OpenFlow will be able to support end-to-end QoS. At present, there are few researches on QoS architecture based on OpenFlow network at home and abroad. These researches mainly use resource reservation, priority queue and QoS routing to provide QoS guarantee for applications. However, the former two will have a great impact on the flow without QoS requirements, and the current QoS routing research has not really combined the current network state, so the calculation path is not accurate. In order to solve the above problems, this paper proposes a QoS support subsystem based on the global control of OpenFlow and the characteristics of supporting the collection of statistical information. The QoS guarantee subsystem mainly collects the real-time QoS parameter information of the network. Use QoS routing technology to provide QoS guarantee for applications. The QoS support subsystem consists of five important modules: path calculation, network measurement, access control, flow management, and topology management. In this paper, two core modules of path calculation and network measurement are designed and implemented in the FloodLight controller development platform, and the topology management module is modified. Path calculation is used to calculate paths that meet the QoS requirements of users. After studying various QoS routing algorithms, the NR_DCLC routing algorithm is used to compute the optimized paths for QoS applications. Network measurement is used to collect the current network state parameter information, which is divided into three sub-modules: link usage measurement module, packet loss measurement module and delay measurement module. The path calculation module computes the optimized QoS route according to the QoS state information provided by the network measurement, and the network measurement notifies the path calculation module to re-select the path when the resource can not meet the QoS requirements of the application. In this paper, the network simulation and verification of QoS support subsystem are carried out by using Mininet simulation platform. The simulation results show that the network measurement can obtain the network state information correctly. The path calculation method based on this paper can improve resource utilization and realize dynamic rerouting in the case of network congestion. Although the new QoS support subsystem proposed in this paper is implemented in the simulation environment, the design and implementation of the functions and codes can be applied to the real network environment, only need each switch to support the OpenFlow protocol. Therefore, the work of this paper has great significance for theoretical research and practical deployment.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09
【引证文献】
相关硕士学位论文 前2条
1 叶s，
本文编号：2231239
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2231239.html
