基于多维感知的智能路由器大数据插件技术研究与实现

发布时间：2018-06-22 18:07

本文选题：智能路由器 + 深度包检测技术　；参考：《南京邮电大学》2017年硕士论文

【摘要】：随着互联网的发展,传统路由器已成为用户上网的必备入口,受制于功能、性能等方面的问题,传统路由器不能很好满足用户的上网体验。智能路由器以其高性能、开放性等特点深入人们的视野。然而,当前的智能路由器具备开放平台可供用户自行安装应用插件,但此类插件均无用户感知能力,自然无法获悉并优化用户体验。同时,作为用户上网的入口,智能路由无法提供很好的大数据资产积累和流量变现的方式。据此,本文提出了一种基于智能路由器的大数据插件系统框架。大数据插件具备用户数据采集的能力,并且采用DPI深度包检测技术,对用户上网流量进行多维感知。插件同时具备精确信息推送的能力,可提供基于时间、内容、形式等多策略的信息推送。此外,插件还具备上传用户行为数据的能力,为运营商或商业Wifi运营商的提供了大数据资产积累的能力。针对此研究目标,本文的主要贡献与创新如下:1)提出一种基于智能路由器大数据插件的系统框架。该框架可实现流量采集和数据上传、多维感知、精确信息推送等能力,并提供插件管理和大数据分析平台,形成一整套完整的插件管理平台体系。2)大数据插件采用多维感知技术,实现了对用户身份、业务应用、终端、位置等多维度感知和提取。3)本文采用Nginx、Node.js和ActiveMQ设计出了支持高并发、有消息解析的能力的技术方案,以解决数据上传的高并发问题。4)本文提出了一种URL快速匹配算法,以提高精确推送中的URL和多重策略的匹配效率问题。5)通过对插件采集的数据进行大数据平台之上的挖掘分析,实现了对用户身份信息、终端信息的准确识别。
[Abstract]:With the development of the Internet, the traditional router has become a necessary entrance for users to access the Internet, which is restricted by the problems of function and performance, so the traditional router can not satisfy the user's Internet experience very well. Intelligent router with its high performance, openness and other characteristics to go deep into the field of vision. However, the current intelligent routers have an open platform for users to install their own application plug-ins, but none of these plug-ins have the ability to perceive users, so it is naturally impossible to learn and optimize the user experience. At the same time, as a user's gateway to the Internet, intelligent routing can not provide a good way to accumulate big data assets and realize traffic. Therefore, this paper presents a big data plug-in system framework based on intelligent router. The big data plug-in has the ability of user data acquisition, and adopts the technology of depth packet detection to realize the multidimensional perception of the user's Internet traffic. The plug-in also has the ability of accurate information push, which can provide information push based on time, content, form and so on. In addition, the plug-in has the ability to upload user behavior data, providing operators or commercial WiFi operators with the ability to accumulate big data assets. Aiming at this research goal, the main contributions and innovations of this paper are as follows: 1) A system framework based on intelligent router big data plug-in is proposed. The framework can achieve traffic collection and data upload, multidimensional perception, accurate information push and other capabilities, and provide plug-in management and big data analysis platform. A complete set of plug-in management platform system .2) big data plug-in using multidimensional awareness technology, to achieve the user identity, business applications, terminals, location and other multi-dimensional awareness and extraction 3) this paper uses NginxFN Node.js and ActiveMQ to design the support of high concurrency. In order to solve the high concurrency problem of data upload, this paper proposes a fast URL matching algorithm. In order to improve the efficiency of URL and multi-strategy matching in accurate push, the user identity information and terminal information can be accurately identified by mining and analyzing the data collected by plug-in on big data platform.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.05;TP311.13

【参考文献】