Location proximity attacks against mobile targets: Analytical bounds and attacker strategies

Paper in proceeding, 2018

Location privacy has mostly focused on scenarios where users remain static. However, investigating scenarios where the victims present a particular mobility pattern is more realistic. In this paper, we consider abstract attacks on services that provide location information on other users in the proximity. In that setting, we quantify the required effort of the attacker to localize a particular mobile victim. We prove upper and lower bounds for the effort of an optimal attacker. We experimentally show that a Linear Jump Strategy (LJS) practically achieves the upper bounds for almost uniform initial distributions of victims. To improve performance for less uniform distributions known to the attacker, we propose a Greedy Updating Attack Strategy (GUAS). Finally, we derive a realistic mobility model from a real-world dataset and discuss the performance of our strategies in that setting.