基于LDP的众包位置数据保护方案
Crowdsourcing position data protection scheme based on LDP
吴朝阳 1李晓会1
作者信息
- 1. 辽宁工业大学电子与信息工程学院,辽宁锦州 121000
- 折叠
摘要
针对众包应用采集用户位置数据导致用户隐私信息泄露的问题,提出一个面向移动端设备的众包位置数据保护方法.通过 LAS 算法(location-secure anonymous sets)构造安全匿名集,使用细粒度 GHC 曲线(granularity hilbert curve)对匿名集进行区域划分;采用本地化差分隐私k-RR随机响应机制对每个区域中的位置数据进行扰动;服务器端对扰动数据聚合优化,得到区域用户人数的真实值.实验结果表明,该方法增强了用户隐私信息的保护程度,提高了位置数据的可用性,在隐私保护和数据可用性之间提供了更好的平衡.
Abstract
To solve the problem of user privacy leakage caused by user location data collection by crowdsourcing applications,a crowdsourced location data protection method was proposed for mobile devices.Granularity(GHC)Hilbert curve was partitioned for secure anonymous sets based on the location-secure anonymous sets(LAS)algorithm.The localization differential privacy k-RR random response mechanism was used to perturb the location data in each region.The server-side aggregated and optimized the perturbed data to get the true values of the regional users.Experimental results show that the proposed method enhances the protection degree of user privacy information,improves the availability of location data,and provides a better balance between privacy protection and data availability.
关键词
本地差分隐私/位置数据/Hilbert曲线/数据采集/位置语义/区域划分/随机响应Key words
local differential privacy/location data/Hilbert curve/data collection/location semantic/regional division/random response引用本文复制引用
基金项目
国家自然科学基金青年科学基金(61802161)
辽宁省应用基础研究计划(2022JH2/101300278)
出版年
2024
NETL
NSTL
万方数据