基于位置服务中半可信用户的位置隐私保护方法

Location Privacy Preserving for Semi-honest Users in LBS

随着移动计算和无限设备的蓬勃发展,基于位置服务中的位置隐私保护研究受到了学术界广泛关注。为保护位置隐私,学术界提出了很多匿名算法。但现有这些算法均基于一个假设,即网络中的所有用户均可信。实际上,网络中并不存在任何可信实体,现有基于真实位置生成匿名集的方法应用领域有限。针对此问题,根据用户发出信号强弱判断位置邻近性,提出了一种在k*NNG上的匿名算法,并通过"协作"的方式在不暴露用户任何隐私的前提下生成匿名区域。在不同系统设置下对算法进行实验,实验结果表明,算法在效率和服务质量上取得了较好的平衡。

Location privacy preservation has recently received considerable attention in LBS. A large num- ber of location cloaking algorithms have been proposed for protecting the location privacy of mobile users. How- ever, most of existing cloaking approaches assume that mobile users are trusted. And exact locations are required to protect location privacy, which are just the information mobile users want to hide. /12 this paper, we propose a p-anti-conspiring privacy model to anonymize over semi-honest users. Furthermore, a k ＊ NNG -based cloaking algorithm k ＊ NNCA is proposed to protect the location privacy without exact locations. The preliminary experi- mental results show the effectiveness of the proposed algorithm.