Location Privacy Protection in Vehicle-Based Spatial Crowdsourcing via Geo-Indistinguishability
Chenxi Qiu, Anna Squicciarini, Ce Pang, Ning Wang, Ben Wu
Abstract
Nowadays, vehicles have been increasingly adopted in many spatial crowdsourcing (SC) applications. Similar to other SC applications, location privacy is of great concern to vehicle workers as they are required to disclose their own location to servers to facilitate the service utilities. Traditional location privacy protection mechanisms cannot be applied to vehicle-based SC since they assume workers’ mobility on a 2-dimensional plane without considering the network-constrained mobility features of vehicles. Accordingly, in this paper, we aim at addressing issues related to Vehicle-based spatial crowdsourcing Location Privacy (VLP) over road networks. Our objective is to design a location obfuscation strategy to minimize the quality loss due to obfuscation with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">geo-indistinguishability</i> satisfied. Considering the computational complexity of VLP, by resorting to discretization, we first approximate VLP to a linear programming problem that can be solved by well-developed approaches. To further improve the time-efficiency, we conduct constraint reduction for VLP by exploiting key features of geo-indistinguishability in road networks and problem decomposition based on VLP’s constraint structure. Finally, we carry out both trace-driven simulation and real-world experiments, where our experimental results demonstrate the superiority of our approach over a known state-of-the-art location obfuscation strategy in terms of both quality-of-service and privacy.