An Efficient and Privacy-Preserving Spatial Crowdsourcing Protocol From Hash Functions for IoT

The proliferation of smart devices has propelled the advancement of IoT-based spatial crowdsourcing. The issue of location privacy in task allocation for IoT-based spatial crowdsourcing has attracted significant attention. Therefore, the main goals of privacy-preserving spatial crowdsourcing (PriSC) are: 1) achieving better location privacy for both participants and tasks and 2) achieving better allocation performance, i.e., accuracy and average moving distance. The homomorphic encryption-based approaches can achieve these goals, yet they suffer from heavy computation and large communication overhead. Although the differential privacy (DP)-based approaches are very efficient, these approaches leverage allocation performance to achieve better location privacy. Motivated by the deficiencies of these existing approaches, we propose a lightweight hash-based spatial crowdsourcing protocol, which not only protects both task location and participant location from the server but also reduces service providers' computation and communication overhead. Besides, our design is independent of the concrete hash function and thus can be instantiated by any collision-resistant cryptographic hash function. Experiment results demonstrate that our protocol outperforms related works in terms of accuracy and average moving distance.