Identifying Web Pages In Encrypted HTTP/2 Traffic

After entering the 21 st century,with the rapid development of the Internet,it has become an indispensable part of human life.Network bringing convenience to us,but also brings us a crisis.How much data is passed on the Internet will imply the same security risks.Network security is already a part of life.How to prevent cyber attacks is also an important part of the Internet.Traffic analysis is an important means of network security monitoring and web page identification is one of the methods of traffic analysis.It can identify abnormal or illegal behavior by web pages and help to improve network security.In the past,although HTTP1.X also protects web page information by encrypting web pages,the encrypted web page header information still has an identifiable part.Therefore,web page identification methods can still be used to monitor HTTP1.X encrypted web pages to ensure web page security.Prior to this,although many methods for identifying HTTP1.X encrypted web pages have been developed,as HTTP/2 protocols are increasingly developed,they are gradually replacing HTTP1.X,since HTTP/2 and HTTP1.X have different characteristics,HTTP1.X's web page recognition method may no longer apply on HTTP/2 encrypted web pages,which brings new security risks.Therefore,it is necessary to design a web page identification method for HTTP/2encrypted web pages.After observing the HTTP/2 network packet structure,we find that there are still features that can be used for identification.Therefore,how to correctly acquire features and identify them is the focus of this paper.This article will build 1000 HTTP/2 web pages,each web page will be accessed 100 times and crawl its network packets for testing.Traffic characteristics will be extracted from each network packet and identified by LL,HWF,DCRS and LCC classifiers.Test method hit rate.This paper uses different traffic characteristics than past research,which provides a new direction for future web page recognition work,and also provides unique ideas for improving network security.