On the Impact of Location Errors on Localization Attacks in Location-Based Social Network Services

Abstract

Location-based Social Network (LBSN) services, such as People Nearby in WeChat, enable users to discover users within the geographic proximity. Though contemporary LBSN services have adopted various obfuscation techniques to blur the location information, recent research has shown that based on the number theory, one can still accurately pinpoint user locations by strategically placing multiple virtual probes. In this paper, we conducted a comprehensive simulation study to examine the impact of location errors on localization attacks to track target users based on the number theory by using the LBSN services provided by WeChat. Our simulation experiments include four location error models including the exponential model, the Gaussian model, the uniform model, and the Rayleigh model. We improve the one-dimensional and two-dimensional localization algorithms where the location errors exit. Our simulation results demonstrate that the number theory based localization attacks remain effective and efficient in that target users can still be pinpointed with high accuracy.