基于内容感知和智能决策的移动网络加速系统的研究与实现

发布时间：2018-01-28 19:36

本文关键词： 内容感知网络架构智能决策本地代理　出处：《北京邮电大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着移动互联网的迅速发展,使用手机等移动终端进行移动上网的网民越来越多。与此同时,移动网络带宽资源变得日益紧张,而且对网络的优化也变得更加困难和复杂。如何在现有网络带宽的基础上有效地对移动网络进行优化是移动互联网发展的关键问题,具有很重要的现实意义。目前针对移动网络的优化和加速问题,理论研究的重点主要侧重于对各种移动网络加速技术以及新型的网络架构的研究,但是移动平台上相关的应用产品还处于起步阶段。本文通过创新性地融合多种网络加速技术以及合理的网络架构设计进而实现一个移动网络加速系统。首先,在研究和分析了现有的内容网络架构和面向未来的网络架构的基础上,提炼出了一个具备内容感知能力的移动网络部署架构。这个架构通过在现网基础上增加用户层、边缘代理服务层和控制中心服务层,并采用了基于最近访问和有限多径探测的资源定位算法,在应用层打造一个基于内容的感知层来对用户的网络请求进行全方位的感知。其次,在上述内容感知网络架构的基础上,针对不同的网络请求融合了多种网络加速方法。并且针对这些不同特点和适用情况的加速技术,在用户层移动端本地代理和边缘代理服务层上分别设计了配套的智能决策流程。这样达到对每一个网络请求灵活地使用相应的优化方案,提升系统的智能性。最后,.基于内容感知和智能决策的功能,主要在Android移动平台上设计和实现了高性能且稳定的移动端本地代理应用,和后台部署的服务器一起完成了整体的加速功能。这个Android应用具有灵活配置功能以及良好的用户交互接口,带来了良好的用户体验。
[Abstract]:With the rapid development of mobile Internet, more and more Internet users use mobile terminals such as mobile phones to access the Internet. At the same time, the bandwidth resources of mobile network become increasingly tight. How to optimize the mobile network effectively on the basis of the existing network bandwidth is a key issue in the development of the mobile Internet, which has a very important practical significance. At present, for the optimization and acceleration of mobile networks, the focus of theoretical research is mainly on the various mobile network acceleration technologies and new network architecture. However, the related application products on mobile platform are still in their infancy. In this paper, a mobile network acceleration system is implemented through the innovative integration of various network acceleration technologies and reasonable network architecture design. First of all, based on the research and analysis of the existing content network architecture and future-oriented network architecture. A content-aware mobile network deployment architecture is developed, which adds user layer, edge agent service layer and control center service layer to the current network. The resource location algorithm based on recent access and finite multipath detection is used to create a content-based awareness layer in the application layer to fully perceive the users' network requests. Secondly, on the basis of the above content aware network architecture, a variety of network acceleration methods are integrated for different network requests, and according to these different characteristics and applicable acceleration technology. The corresponding intelligent decision flow is designed on the mobile local agent and edge agent service layer of the user layer, which can flexibly use the corresponding optimization scheme for each network request and enhance the intelligence of the system. Finally, based on the function of content awareness and intelligent decision, the high performance and stable local agent application of mobile side is designed and implemented on Android mobile platform. The Android application has a flexible configuration function and a good user interaction interface, which brings a good user experience.
【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】：TP393.01;TN929.5

【参考文献】