Tor(the onion router)目前存在用户访问网站指纹信息泄露,真正问题不是它的惟一性,而是它的可链接性,即每次访问网站都能链接到同一个指纹.基于此发现,利用随机化策略,设计并实现了一种新颖的网站指纹防御方案:随机化隐私改变器。随机化隐私改变器以随机频率发送随机时间、随机长度数据包,同时随机请求一个虚假网站作为噪音流量,全方位对网站指纹信息进行了模糊.实验结果表明,网站指纹攻击的准确度在随机化隐私改变器防御下大大地降低了,极大地提高了Tor的安全性,而且它的防御效果比同类的Tamaraw好.
Website Fingerprint can disclose users' information even users utilize the Tor. The real problem with fingerprint is not the uniqueness but the linkability, i. e. , the ability to connect the same fingerprint across multiple visits. Based on this foundation, we design and implement a novel website fingerprint defense scheme called randomized privacy modifier. The randomized privacy modifier can send the random length packet of random length for random time, and randomly requests a fake website as the noise flow, thus the full range of fingerprint information on the website was vague. Experimental results show that the accuracy of website fingerprint attack is greatly reduced under the randomized privacy modifier defense, and the effect is better than similar defense Tamaraw. Therefore our scheme improves the security of Tor.