支持重叠网络多径传输的评价模型和方法研究

发布时间：2018-04-29 07:06

  本文选题：重叠网络 + 多径传输　； 参考：《东北大学》2014年硕士论文

【摘要】：基于传统路由技术构建的单径传输模式,难以有效支持资源调配,限制了高带宽业务需求通信应用的发展。基于应用层中继构建的重叠网络多径传输技术,是实现媒体传输技术升级的一种重要方式。为了实现有QoE保证的多径传输,无论是路径选择,还是负载分配,都依赖于路径评价。一方面,以往针对网络传输层面的评价标准不能直接反映用户对服务的认可程度,网络性能评价指标已从符合网络QoS条件转变为满足用户QoE需求。另一方面,目前成熟的评价方法并不适用于多径传输网络。因此亟待研究适用于重叠网络多径传输的有效评价模型。本文对重叠网络多径传输的评价模型和方法进行了研究,归纳总结目前承载网最常用的Internet QoS体系结构及QoE评价方法,分析了它们在使用过程中可能存在的问题。综合考虑端到端网络QoS及用户QoE,使用模糊数学、随机过程、马尔科夫预测模型、矢量分析法等进行统计评价,提出了一种适用于重叠网络多径传输的评价模型。主要内容及成果如下：第一,设计工作在终端的QoE/QoS报告组件,和工作在网络及源的QoE管理组件,构建端到端QoE保证系统。第二,设计QoE综合评价模型。从RTCP中采集网络评价参数,并对数据矢量化处理,讨论了两种方法将测量值和估计值进行等级量化,分别得到网络可提供QoE等级及用户QoE需求等级。若当前网络的QoE可提供等级高于用户QoE需求等级,则传输有QoE保证,否则没有保证。第三,设计实现QoE动态评价机制。为避免单一路径突变影响多径传输总体评价效果,讨论了两种动态触发情况,实现了端到端多径传输的实时评价。第四,扩展RTCP的自定义分组类型,设计多径传输质量反馈报告,及复合QoE评价报告。此外,本文在设计模型时,还考虑了与3GPP策略管理框架的兼容性,介绍了模型应用及信令流程。使用OMNeT++分别对普通视频通话及高清视频通话两种不同业务进行了仿真测试,并在仿真结果输出后用Matlab进行数据分析。验证了本文提出的解决方案具有合理性与可行性,评价模型能够实时评价网络性能。使用评价控制方案可进一步提高端到端服务质量,使多径传输有QoE保证。
[Abstract]:The single path transmission mode based on traditional routing technology is difficult to support resource allocation effectively, which limits the development of communication applications for high bandwidth services. Overlapping network multipath transmission technology based on application layer relay is an important way to upgrade media transmission technology. In order to realize multipath transmission guaranteed by QoE, both path selection and load distribution depend on path evaluation. On the one hand, the previous evaluation standards for network transmission level can not directly reflect the degree of user approval of the service, and the network performance evaluation index has changed from meeting the network QoS conditions to meeting the QoE requirements of users. On the other hand, the current mature evaluation methods are not suitable for multipath transmission networks. Therefore, it is urgent to study an effective evaluation model for overlapped multipath transmission. In this paper, the evaluation models and methods of multipath transmission in overlapped networks are studied. The most commonly used Internet QoS architecture and QoE evaluation methods are summarized, and the possible problems in the process of using them are analyzed. Considering the end-to-end network (QoS) and user QoS (user QoS), a statistical evaluation model is proposed by using fuzzy mathematics, stochastic process, Markov prediction model, vector analysis method and so on, and an evaluation model for overlapping network multipath transmission is proposed. The main contents and results are as follows: first, design the QoE/QoS report component working in the terminal, and the QoE management component working in the network and the source to build the end-to-end QoE guarantee system. Second, design the comprehensive evaluation model of QoE. The network evaluation parameters are collected from RTCP and the data vectorization is processed. Two methods are discussed to quantify the measured value and the estimated value respectively. The network can provide the QoE level and the user QoE requirement level respectively. If the current network's QoE can provide a higher level than the user's QoE requirement level, then the transmission is guaranteed by QoE, otherwise, there is no guarantee. Thirdly, the dynamic evaluation mechanism of QoE is designed and implemented. In order to avoid the single path catastrophe affecting the overall evaluation effect of multipath transmission, two kinds of dynamic triggers are discussed, and the real-time evaluation of end-to-end multipath transmission is realized. Fourthly, we extend the custom grouping type of RTCP, design multipath transmission quality feedback report and compound QoE evaluation report. In addition, the compatibility with 3GPP policy management framework is considered in the design of the model, and the application of the model and the signaling flow are introduced. OMNeT is used to simulate two kinds of different services, ordinary video call and high-definition video call. After the simulation results are outputted, the data are analyzed by Matlab. It is verified that the proposed solution is reasonable and feasible, and the evaluation model can evaluate the network performance in real time. Using evaluation control scheme can further improve the quality of end-to-end service, so that multipath transmission can be guaranteed by QoE.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09

【相似文献】

相关期刊论文 前8条

1 张巍;张麟兮;郭淑霞;刘宁;;数据链路多径传输模型的物理映射方法研究[J];计算机仿真;2010年04期

2 袁林锋;杜旭;杨宗凯;;基于决定性能量比例的多径传输技术[J];计算机科学;2006年04期

3 孙慧锋;刘文芬;张建辉;;Round-Robin分配方式下多径传输的重排序性能分析[J];电子与信息学报;2011年08期

4 阳旺;李贺武;吴茜;吴建平;;基于最小反馈时延的多径应答路径选择算法[J];清华大学学报(自然科学版);2011年07期

5 樊邦奎;丁冠军;兰海滨;龙腾;王晶;陈志勇;;智能电网应用中的PLC最短路径及多径传输统计模型[J];电工技术学报;2013年S2期

6 孙慧锋;刘文芬;张建辉;;多径传输下的数据包重排序性能分析[J];信息工程大学学报;2010年06期

7 徐祺瑞;尹福昌;;激光在水下多径传输的蒙特卡罗模拟[J];长春理工大学学报(自然科学版);2008年01期

8 ;[J];;年期

相关会议论文 前1条

1 李志伟;杨益新;;基于GCC的水声信号多径传输TOA估计仿真研究[A];中国声学学会水声学分会2013年全国水声学学术会议论文集[C];2013年

相关硕士学位论文 前6条

1 杨樝;支持重叠网络多径传输的评价模型和方法研究[D];东北大学;2014年

2 尚小东;支持实时媒体交换的多径传输控制协议的研究[D];东北大学;2012年

3 程浩;面向多径传输的应用层QoS评价模型和决策方法[D];东北大学;2013年

4 冯雪;支持应用层QoS决策的多径传输路径选择方法研究[D];东北大学;2013年

5 赵奇;向量网多径传输算法研究和仿真实现[D];北京交通大学;2012年

6 杨军;Ad Hoc网络视频多描述编码与多径传输技术研究[D];苏州大学;2009年

，

本文编号：1818906