Towards Privacy-Preserving Travel-Time-First Task Assignment in Spatial Crowdsourcing

With the ubiquity of mobile devices and wireless networks, spatial crowdsourcing (SC) has gained considerable popularity and importance as a new tool of problem-solving. It enables complex tasks at specific locations to be performed by a crowd of nearby workers. In this paper, we study the privacy-preserving travel-time-first task assignment problem where tasks are assigned to workers who can arrive at the required locations first and no private information are revealed to unauthorized parties. Compared with existing work on privacy-preserving task assignment, this problem is novel as tasks are allocated according to travel time rather than travel distance. Moreover, it is challenging as secure computation of travel time requires secure division which is still an open problem nowadays. Observing that current solutions for secure division do not scale well, we propose an efficient algorithm to securely calculate the least common multiple (LCM) of every workers speed, based on which expensive division operation on ciphertexts can be avoided. We formally prove that our protocol is secure against semi-honest adversaries. Through extensive experiments over real datasets, we demonstrate the efficiency and effectiveness of our proposed protocol.