摘要
基于DPI的流量识别方法,通过匹配应用流量报文独有的特征字符串来识别移动应用程序产生的流量,具有较好的识别效果,但特征字符串需要人为提取。对此,提出一种通过全面观察移动应用程序产生的流量报文以自动学习移动应用指纹的方法。实验结果表明,该方法用于移动网络流量识别时的应用覆盖率可达83.3%,流覆盖率、字节覆盖率均可达较高水平。
The DPI-based traffic identification methods can identify the traffic generated by the mobile application by matching the unique feature strings of application traffic message,which has a better recognition effect,but the feature strings need to be manually extracted.Based on this,this paper proposes a method to automatically learn mobile application fingerprints by comprehensively observing the traffic packets generated by mobile applications.The experimental results show that the application coverage of the proposed method can reach 83.3%,and the flow coverage and byte coverage can reach a high level.
作者
饶亲苗
彭艳兵
Rao Qinmiao;Peng Yanbing(Wuhan Research Institute of Posts and Telecommunications,Wuhan 430000,Hubei,China;Nanjing Fiberhome World Communication Technology Co.,Ltd.,Nanjing 210000,Jiangsu China)
出处
《计算机应用与软件》
北大核心
2021年第4期328-333,共6页
Computer Applications and Software
关键词
深度报文检测
移动网络
流量识别
应用指纹
应用覆盖率
Deep packet inspection
Mobile network
Traffic identification
Application fingerprint
Application coverage
