摘要
针对基于位置的服务带来的用户位置隐私暴露问题,提出了一种基于隐私偏好的二次匿名位置隐私保护方法,融合k-匿名技术和差分隐私技术确保用户位置隐私,设计隐私等级划分策略,支持用户个性化设置隐私保护级别。根据隐私级别确定k匿名集大小,通过基于位置熵的k匿名算法求解k-1个匿名点,使k匿名集的点具有最大概率相似性;在此基础上进一步求解获取位置服务的匿名位置,提出了基于差分隐私的匿名位置生成算法,在保护用户位置隐私的同时确保获取精确的位置服务。实验结果表明在用户隐私等级设置范围内,所提方法能有效兼顾位置隐私保护和LBS服务质量。
Location based service brings the challenging problem of privacy leakage. The method proposes a parameterized location privacy protection method based on two-level anonymity for the problem. The system applies the k-anonymity and differential privacy methods with customized protection level for different users. This method selects the k - 1 anonymous coordinates from the set, which achieves the best probability likelihood of the request, using location entropy based k-anonymity algorithm according to users＇ protection level. Moreover, the system propose a differential privacy based method to generate a dummy position which is indistinguishable and in proper distance with the real position. The experiment results show that our method can protect users＇ privacy as well as preserving the accuracy of location based service.
作者
毕晓迪
梁英
史红周
田辉
BI Xiao-di LIANG Ying SHI Hong-zhou TIAN Hui(Research Center for Ubiquitous Computing Systems, Chinese Academy of Sciences, Beijing 100190, China University of Chinese Academy of Sciences, Beijing 100190, China China Academy of Telecommunication Research of MIIT, Beijing 100142, China)
出处
《山东大学学报（理学版）》
CAS
CSCD
北大核心
2017年第5期75-84,共10页
Journal of Shandong University(Natural Science)
基金
国家重点研发计划项目(2016YFB0800403)
国家高技术研究发展计划(863计划)基金项目(2015AA015803)
北京市科技计划课题项目(Z161100001616009)
关键词
位置隐私保护
k匿名
差分隐私
位置服务
隐私等级
location privacy protection
k-anonymity
differential privacy
location based service
privacy level
