面向SDN用户体验质量的路由模型研究

发布时间：2018-04-21 02:04

本文选题：软件定义网络 + 用户体验质量　；参考：《大连海事大学》2017年硕士论文

【摘要】：随着互联网的蓬勃发展,网络规模扩大,使用人数剧增,业务种类也不再单一,音频、视频等多媒体业务逐渐占据主导地位,这导致网络流量呈现爆炸性增长,但是有限的网络资源已经难以负担起多变的业务类型,终端用户的体验质量更是难以得到保障,如何优化用户的体验质量成为了时下研究热点。传统网络由于本身结构上的限制已经很难在这方面取得进展。而近几年,软件定义网络(Software Defined Networks,SDN),这种具有控制逻辑与转发解耦、网络动态视图可获取、逻辑控制可编程化实现等特点的新生代网络结构的提出,为网络性能的优化开拓了思路。对此,本文决定在SDN环境下,以用户体验质量(Quality of Experience,QoE)为业务传输的衡量指标,通过对路由算法进行优化研究,实现资源分配合理化、提升用户体验质量。针对路由问题的多约束性以及可行解的不唯一性,本文引入了蚁群算法作为路由优化的核心算法,并在SDN网络环境下构建了以QoE蚁群路由为核心的路由模型。该模型主要包括有拓扑感知、蚁群路由、用户终端信息交互等功能模块。在业务的传输过程中,由控制器获取全局的链路信息与用户相关参数,依照MOS效用函数,得出QoE量化指标作为蚁群选路的约束条件,从而为用户选取一条最能保障体验质量的路径。本文在Floodlight控制器与Mininet搭建的仿真环境下进行本文模型的仿真实验,在对不同业务类型传输的情况下,将本文中的QoE蚁群路由模型与基于QoS的蚁群路由模型以及Dijkstra路由的实验结果进行比对,主要比对的参数有传输过程中的时延和丢包率、二者不同业务下的MOS量化值,多业务请求下的MOS量化值的变化。通过实验表明,本文提出的模型在网络传输流量较大的情况下,相比于传统算法在音频与视频业务的传输上更具有用户体验质量的保障。
[Abstract]:With the rapid development of the Internet, the scale of the network is expanding, the number of users is increasing, the types of business are no longer single, and the multimedia services, such as audio and video, have gradually dominated. This leads to the explosive growth of network traffic, but the limited network resources have been difficult to afford a variety of business types. The quality of the experience of the end users is more than that of the Internet. It is difficult to ensure that how to optimize the user's quality of experience has become a hot spot of research. The traditional network has been difficult to make progress in this area because of its structural limitations. In recent years, the software defined network (Software Defined Networks, SDN), this kind of control logic and forwarding decoupling, network dynamic view can be obtained, logic control. The new generation network structure, which has the characteristics of programmable realization, has opened up a train of thought for the optimization of network performance. In this paper, this paper decides to use the user experience quality (Quality of Experience, QoE) as the measure index of the service transmission under the environment of SDN, to optimize the resource allocation by optimizing the routing algorithm, and to improve the user. This paper introduces ant colony algorithm as the core algorithm of routing optimization, and constructs a routing model based on QoE ant colony routing in SDN network environment. This model mainly includes topology perception, ant colony routing, user terminal information interaction and so on. In the transmission process of the business, the controller obtains the global link information and the user related parameters. According to the MOS utility function, the QoE quantized index is used as the constraint condition of the ant colony selection, thus the user can select a path to ensure the best quality of experience. This paper is under the simulation environment built by the Floodlight controller and Mininet. In the simulation experiment of this model, the QoE ant routing model in this paper is compared with the results of the QoS based ant routing model and the experimental results of the Dijkstra routing in the case of different traffic types. The main comparison parameters have the time delay and packet loss rate in the transmission process, the MOS quantization value under the two different services, and the multi service. The change of the MOS quantized value under the request shows that the model proposed in this paper has a better quality of user experience than the traditional algorithm in the transmission of audio and video services in the case of large network traffic flow.

【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.02;TP18

【相似文献】