针对HTTPS协议网站精细化指纹攻击方法

Refined Fingerprint Attack Method for HTTPS Protocol Website

当前网站采用HTTPS协议加密,对其实施监管与审查仅能识别站点,而不能进一步精细化识别子页面,针对这一问题,提出了一种针对HTTPS协议加密站点在使用内容分发网络(Content Distribute Network,CDN)分发资源情况下的精细化指纹攻击方法。首先利用CDN分发过程中将用户重定向到就近镜像服务器产生的域名系统(Domain Name System,DNS)查询序列作为页面指纹,然后使用支持向量机(Support Vector Machine,SVM)模型进行页面识别,最后采用在Internet中收集的数据集进行验证。结果表明:该方法获得了93%的站点子页面识别率,能有效精细化识别HTTPS加密站点的子页面。

Aiming at the problem that the current website uses HTTPS protocol encryption,its supervision and review can only identify the site,and cannot further identify subpages,a refined fingerprint attack method for HTTPS protocol encrypted site using Content Distribute Network（CDN） distribution is proposed. Firstly,the method uses the Domain Name System（DNS） query sequence generated during CDN distribution that users are redirected to the nearest CDN mirror server as page fingerprint. Then the Support Vector Machine（SVM） model is used for pages identification. Finally,dataset collected on the internet is used for verification. The result shows that the method gets a 93% subpage identification rate and can finely and effectively identify the subpages of the HTTPS encrypted site.