Research And Application Of Mobile Application And Key Information Recognition In Network Traffics

In recent years,China's mobile Internet industry is developing rapidly.More and more people access the Internet by their mobile phones.Apps on mobile platforms are becoming richer and richer and their functions are more and more powerful.Against such a backdrop,it's more and more essential to dig out meaningful information from mobile traffic.In this paper I study the existing network traffic identification technologys,analyze the L7-filter packet identification system and the Snort intrusion-detection system,capture and analyze100 APP's traffic.On these foundations,I design the method of mobile phone application recognition and key information extraction.In the aspect of mobile APP recognition,I choose the DIP technology.DPI technology has good real-time performance and high recognition accuracy,therefore it is widely used in industrial field.Because some applications do not use the HTTP protocol mainly but almost all the applications use DNS protocol,I design the recognition method for DNS message and HTTP message.HTTP message based recognition is achieved by "four tuples for locating" and DNS message based recognition is achieved by domain name recognition.In the aspect of HTTP message classification,I also choose the DPI technology.I has tried 5 kinds of characteristics and finally choose the Host header and a character string from User-Agent header as the basis of classification.In the extraction of key information,this paper only aims at the HTTP message.The rules used to extract information includes 4 part: APP name and the type of extracted information,”four tuples for locating”,extraction method and extracting string,the method of information encapsulation and transmission.In this paper,100 common domestic mobile phone applications are identified and information extracted.I complete three rule libraries,including the rule library of domain recognition,information extraction rule library and the classification rule library of HTTP message.Based on the previous jobs a system is developed.The information extracted includes three aspects,the first aspect is the information related to mobile phone,such as phone number and IMEI code.Second is the information related to the applications.Third is the information related to the users,Such as QQ id,comment information,shopping information,publishing information,etc.