基于用户代理及互联网知识的应用识别  

Identifying Apps Based on User-Agent and Web-Knowledge

作　　者：桂小林[1] 刘军[1] 乔媛媛[1] 池沐聪 雷振明[1] 

机构地区：[1]北京邮电大学网络体系构建与融合北京市重点实验室,北京100876

出　　处：《北京邮电大学学报》2017年第2期102-105,共4页Journal of Beijing University of Posts and Telecommunications

基　　金：高等学校学科创新引智计划(111计划)项目(B08004);国家自然科学基金项目(61601414)

摘　　要：为了在网络端识别安装在移动设备中的应用名称,提出了一种新的应用识别方法.该方法能根据模糊或不完整的用户代理(UA)特征字符串识别应用名称.首先,从超文本传输协议请求报文中获得UA特征;然后,从互联网中获取该UA对应的特征文本;最后,对该特征文本进行分析进而识别应用名称.基于大规模网络端采集的网络流量数据进行实验,结果表明所提方法的识别效果优于其他已知方法.A new method was proposed to identify apps via analyzing network traffic.This method can identify apps according to incomplete and ambiguous features of user-agent.Firstly, the user-agents from hyper text transfer protocol headers was extracted;then, the related text feature was obtained from Internet by using the corresponding user-agent;finally, the apps by analyzing the obtained text features was identified.Experiments are conducted by using hyper text transfer protocol traffic traces collected from Gn interface.It is shown that the proposed techniques are superior to the known methods.

关 键 词：应用识别 互联网知识 用户代理 超文本传输协议 

分 类 号：TN911.22[电子电信—通信与信息系统]