基于Laplace机制的加密流量特征集隐私保护方法
A Privacy-preserving Method for Encrypted Traffic Feature Set Based on Laplace Mechanism
靳玮琨 1郭晓军 2杨明芬3
作者信息
- 1. 西藏民族大学信息工程学院,咸阳 712082
- 2. 西藏民族大学信息工程学院,咸阳 712082;西藏网络空间治理研究基地,咸阳 712082
- 3. 西藏自治区科技信息研究所,拉萨 850000
- 折叠
摘要
随着网络安全和隐私问题被广泛关注,越来越多的网络流量采用加密技术进行传输,加密流量分类对于网络监管起到了至关重要的作用.针对在加密流量分类过程中容易出现的用户隐私泄露等问题,提出一种基于Laplace机制的加密流量特征集隐私保护方法.该方法通过生成随机扰动间隔区间的方式,按照生成的区间多次变换扰动间隔对加密流量特征集标签栏进行一定程度的扰动,达到保护用户隐私信息的目的.最后在IS-CX VPN-NonVPN数据集上进行验证,实验结果表明,在保证隐私的前提下,仍能较好地保证加密流量分类精确率,证明了提出方法的有效性和可用性.
Abstract
As network security and privacy issues are widely concerned,more and more network traffic is transmitted by using encryption technology.The classification of encrypted traffic plays a vital role in network supervision.In or-der to solve the problems of user privacy leakage that are easy to occur in the process of encrypted traffic classifica-tion,a privacy protection method of encrypted traffic feature set based on Laplace mechanism was proposed.In this method,the label bar of the encrypted traffic feature set is disturbed to a certain extent according to the generated in-terval by generating random disturbance interval intervals,so as to protect user privacy information.Finally,the pro-posed method is verified on the ISCX VPN-NonVPN dataset.The experimental results show that the accuracy of en-crypted traffic classification can be well guaranteed under the premise of ensuring privacy,which proves the effective-ness and usability of the proposed method.
关键词
加密流量分类/特征集/Laplace机制/隐私保护Key words
Encrypted traffic classification/Feature set/Laplace mechanism/Privacy protection引用本文复制引用
基金项目
新疆维吾尔自治区自然科学基金(XZ2019ZRG-36Z)
西藏民族大学"藏秦喜马拉雅人才发展支持计划-杰出青年学者"项目(324011810216)
西藏民族大学"涉藏网络信息内容与数据安全团队"项目(324042000709)
出版年
2024
NETL
NSTL
万方数据