基于差分隐私的个人轨迹信息保护机制

Personal Trajectory Information Protection Based on Differential Privacy Mechanism

随着智能手机的发展,基于位置的服务越来越受欢迎,这种服务正在引起严重的隐私问题,因为许多用户不愿看到他们的位置信息泄露给服务提供商。近年来研究人员将广义的差分隐私概念引入了位置信息保护中,提供了一个具有严格数学证明的专用隐私保护框架。直观地说,差分隐私意味着通过扰动,使给定距离内的任何两个可能的发布位置的生成概率相似,因此攻击者无法了解用户的真实位置。然而,在保证隐私的前提下,用户总是希望所访问服务的质量损失是最小的。针对上述问题给出了一种后置映射的方法来实现。后置映射机制可以在满足相同的隐私级别同时改善其平均服务质量,并结合真实数据,对机制进行了仿真分析,结果显示机制的服务质量损失低于平面拉普拉斯机制。

With the development of smartphones, location-based services are becoming more and more popular, and such services are causing serious privacy problems because many users are reluctant to see their location information leaked to service providers. In recent years, researchers have introduced the concept of differential privacy in a broad sense into location information protection, providing a private privacy protection framework with strict mathematical proof. Intuitively speaking, differential privacy means that by perturbation, the probability of generating any two possible release locations within a given distance is similar, so the attacker cannot understand the true location of the user. However, under the premise of ensuring privacy, users always want the quality loss of the accessed service to be minimal. A post mapping method is implemented to solve the above problem. The post-mapping mechanism can improve the average quality of service while satisfying the same privacy level, and combines the real data to simulate the mechanism. The result shows that the service quality loss of the mechanism is lower than the plane Laplacian mechanism.