基于本地差分隐私的Wi-Fi指纹定位隐私保护方法  

Privacy protection method for Wi-Fi fingerprinting location based on local differential privacy

作　　者：刘浪[1] 蔡利平[1] 何文涛 周绪川[2] LIU Lang;CAI Li-ping;HE Wen-tao;ZHOU Xu-chuan(Center for Information and Modern Educational Technology,Southwest Minzu University,Chengdu 610041,China;National Ethnic Affairs Commission Key Laboratory for Computer Systems,Southwest Minzu University,Chengdu 610041,China)

机构地区：[1]西南民族大学信息与现代教育技术中心,四川成都610041 [2]西南民族大学计算机系统国家民委重点实验室,四川成都610041

出　　处：《西南民族大学学报（自然科学版）》2023年第5期537-542,共6页Journal of Southwest Minzu University(Natural Science Edition)

基　　金：2021年四川省高等教育学会教育信息化项目(GJXHXXH21-YB-10)。

摘　　要：针对当前位置隐私保护中遭遇的背景知识攻击,服务器不可靠等原因所造成的问题,在Wi-Fi指纹定位技术的基础上,结合RAPPOR算法,提出了一种满足本地差分隐私的室内位置隐私保护方法.该方法通过参考点的无线电信号特征来划分位室内环境的区域,使得用户数据满足RAPPOR算法的输入,最后将扰动向量作为输出发送到服务器端.在真实数据集上的实验也表明,该方案在保证位置隐私的前提下也能得到不错的数据效用性.Aiming at the problems caused by background knowledge attacks and unreliable servers in current location privacy protection,this paper proposed an indoor location privacy protection method that satisfied local differential privacy based on Wi-Fi fingerprint positioning technology combined with RAPPOR algorithm.This method divided the indoor environment into different areas based on the wireless signal characteristics of the reference points,so that the user data met the input of the RAPPOR algorithm.Then,the disturbance vector was sent as output to the server.Experiments on real datasets showed that this scheme could achieve good data utility while ensuring location privacy.

关 键 词：指纹定位 本地差分隐私 隐私保护 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]