Privacy-preserving Partner Selection for Ride-sharing Services

Abstract

Ride-sharing services (RSSs) assist drivers to find proper riders for vacant seats on the road, providing appealing benefits of shared travel cost and improved vehicle occupancy, which have revolutionized transportation business, as witnessed by the success of Lyft Line, UberPool, and Waze Carpool. Selecting proper ride-share partners for drivers based on riders’ trip data is essential for RSSs, but it also leads to the exposure of drivers’ and riders’ future locations and trajectories. To preserve the individual privacy during partner selection, in this paper, we propose a privacy-preserving ride-matching scheme for selecting feasible ride-share partners in RSSs. First, we design a spatial region-based selection mechanism, which allows the Ride-Sharing server (RS-server) to prechoose riders in the matched regions with drivers, without exposing their accurate sources and destinations. Second, with the encrypted itineraries of drivers and riders, the RS-server further selects potential ride-share partners according to the travel time saving (TTS) and the feasibility of time schedules. Third, the RS server determines proper ride-share partners with the objective of maximizing the system-wide TTS. With the three-step partner selection, suitable riders can be discovered for the drivers to share vacant seats, resulting in the saving of total travel time and expenditure for riders and drivers. Finally, we demonstrate that the proposed scheme offers strong privacy guarantees to both riders and drivers, while maintaining the efficiency and practicality of RSSs.