基于软件定义的车联网QoS路由框架研究

发布时间：2018-05-14 12:27

本文选题：车联网 + SDN(软件定义网络)　；参考：《华东师范大学》2017年硕士论文

【摘要】：软件定义网络(SDN,Software Defined Networking)是最近几年提出来的一种新型网络架构。SDN将网络的控制平面分离到远程的控制器上,使得网络管理人员可以集中地对网络进行管理,并力图解决现有网络的僵化问题。鉴于SDN的这些特性,本论文将其运用于车联网中,设计了一种基于SDN的车联网架构 SDVANET(Software Defined Vehicular Ad-hoc NETwork),使得车联网能够有效地被管理。同时,本文还分析了传统Ad Hoc网络确定性路由协议在车联网环境下的不足之处,并为SDVANET设计了一种新的机会路由协议。该协议将车辆到基站的线路质量、车辆的移动速度等信息作为候选转发节点的选择依据。另一方面,近几年无人驾驶技术的发展,也大大增加了车联网的应用场景。通过云端海量数据的计算,车辆可以在不依靠人的情况下智能地做出驾驶决策。在这类场景中,因为车辆数据极大影响行车安全和道路安全,所以保证车辆的数据传输质量非常重要。传统的网络架构由于其本身设计的局限性导致无法充分利用整个网络的资源来保障数据传输的服务质量;虽然SDN架构集中控制的特点使得其可以从全局的视角来对网络资源进行分配,但是现有SDN架构下的QoS(Quality of Service)框架并没有考虑到拓扑节点的频繁移动问题,因而不适用于车联网环境。基于上述背景,本文为SDVANET设计了一套新的QoS框架,包括OpenFlow网络和无线网络两个部分:在OpenFlow网络部分,本文从资源监控,数据流预测,路由调度等几个方面对其进行了详细设计,并通过数据层面的计量表和端口队列来实现;而在无线网络部分,本文从介质访问控制层,网络层和应用层设计了一套自下而上的QoS方案。最后,我们在NS3网络模拟器上实现了 SDVANET以及相应的QoS框架,并基于不同的场景进行了仿真实验,验证了 SDVANET架构的可行性和有效性。同时实验结果也表明,即使在网络拥堵的情况下,高优先级的数据包也能够保持较低的平均时延和丢包率。
[Abstract]:SDN (Software Defined networking) is a new network architecture proposed in recent years. SDN separates the control plane of the network from the remote controller, which enables network managers to manage the network centrally. And try to solve the existing network ossification problem. In view of these characteristics of SDN, this paper applies it to vehicle networking, and designs a kind of SDN based vehicle networking architecture, SDVANET(Software Defined Vehicular Ad-hoc net workstation, so that the vehicle networking can be managed effectively. At the same time, this paper also analyzes the shortcomings of the traditional Ad Hoc network deterministic routing protocol under the environment of vehicle networking, and designs a new opportunistic routing protocol for SDVANET. In this protocol, the route quality and the moving speed of the vehicle to the base station are taken as the basis for the selection of the candidate forwarding nodes. On the other hand, the development of driverless technology in recent years has greatly increased the application scenarios of vehicle networking. By computing massive cloud data, vehicles can make driving decisions intelligently without relying on human beings. In this kind of scenario, it is very important to ensure the quality of vehicle data transmission because vehicle data greatly affect driving safety and road safety. Because of the limitation of its own design, the traditional network architecture can not make full use of the resources of the whole network to ensure the quality of service of the data transmission. Although the centralized control of SDN architecture enables it to allocate network resources from a global perspective, the existing QoS(Quality of Service framework under the SDN architecture does not take into account the frequent movement of topology nodes. Therefore, it is not suitable for vehicle networking environment. Based on the above background, this paper designs a new QoS framework for SDVANET, including two parts of OpenFlow network and wireless network: in the part of OpenFlow network, this paper designs it in detail from several aspects, such as resource monitoring, data flow prediction, routing scheduling and so on. In the wireless network part, we design a bottom-up QoS scheme from the media access control layer, network layer and application layer. Finally, we implement the SDVANET and the corresponding QoS framework on the NS3 network simulator, and carry out simulation experiments based on different scenarios to verify the feasibility and effectiveness of the SDVANET architecture. At the same time, the experimental results show that even in the case of network congestion, the high priority packets can maintain low average delay and packet loss rate.
【学位授予单位】：华东师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U495;TN929.5

【相似文献】