Impact of time-varying population density on location privacy preservation level
Yuya Matsuno, Masaki Ito, Kaoru Sezaki · 2016
Prevalence of mobile devices equipped with Global Positioning System (GPS) receivers has made available various services providing useful information related to user location. However, since users have to send location information which may reveal private information, preserving location privacy has become a significant challenge. We require an environment which enables users to enjoy the benefit of services without privacy concerns. Location privacy preservation methods should be constructed by considering the adversary's knowledge of the user's movement. In this paper, we consider time-varying distribution of population density as the adversary's knowledge to estimate user locations. We evaluated the impact of the time-varying distribution of population density on location privacy preservation level in a simulation experiment. We used a people flow data set of the Tokyo metropolitan area to obtain the time-varying distribution of population density and user location traces. The results demonstrate that the adversary is able to estimate user locations more correctly and more confidently by considering the time-varying distribution of population density. We also found a problem in terms of large dispersion of the location privacy preservation level between regions. In other words, it is hard to satisfy the criteria required by the location privacy preservation level. Given the simulation results, we suggest that considering the time-varying distribution of population density is effective for estimation of user location and degrading the performance of location privacy preservation mechanisms.