Location Privacy in Public Access Points Positioning: an Optimization and Geometry Approach.

Yunlong Mao,Yuan Zhang,Xiaoyan Zhang,Fengyuan Xu,Sheng Zhong
DOI: https://doi.org/10.1016/j.cose.2017.11.005
IF: 5.105
2017-01-01
Computers & Security
Abstract:This study proposes a method to tackle a new threat in current location-based services (LBS) wherein access points (APs) use shared public IP addresses. In this scenario, if an LBS request originates from one user, the location of other users who are using the same public IP address will also be revealed, which is a significant threat to user privacy. In this paper, we propose a novel approach for the protection of location privacy when APs have shared public IP addresses. The basic idea behind ensuring location privacy is to divide APs into two exclusive groups, namely, one for LBS users, and the other for non-LBS users. Achieving this bipartition while keeping every user in the service region connecting to one AP is difficult. To resolve this challenge, we first propose two optimal algorithms to perform the bipartition based on different interests. We then use Loopy Belief Propagation to propose a min-sum belief propagation algorithm on a loopy graph model to guarantee the quality of service by complete coverage. Experimental results demonstrate the accuracy and feasibility of our algorithms.
What problem does this paper attempt to address?