Privacy-preserving Online Ride-hailing Service System Based on Taking the Intersection of Private sets of Points of Interest

Rui Li,Juyuan Zhang,Xiaoya Hu,Shihui Zheng,Licheng Wang

DOI: https://doi.org/10.1109/MICCIS58901.2023.00023

2023-04-01

Abstract:Online ride-hailing has gained recognition and become increasingly popular as a way to ease traffic congestion. However, it also poses a serious threat to people's data privacy. While existing schemes have made good progress in protecting the location privacy of drivers and passengers, there are also some problems. When ride-hailing platforms are faced with a large number of requests from passengers and drivers, the time complexity of conducting driver-passenger matching is relatively high; in some schemes based on hidden region division, drivers and passengers located at the boundary of adjacent regions with relatively similar locations cannot be matched successfully. In the paper, we propose a privacy-preserving online ride-hailing service system based on taking the intersection of private sets of points of interest (PIHS). The core idea of the scheme is to use a set of location points to represent the geographic location of a user in an online ride-hailing service. It converts the explicit set into a privacy set that is unrecognizable to an attacker and converts the distance between two location points into a problem of finding the intersection of the privacy sets, so that drivers and passengers with a significant number of intersections between the two privacy sets can be matched successfully. Under the traditional online ride-hailing architecture and the Cloud-Edge-End architecture for the Internet of Vehicles, we use existing cryptographic techniques to design a privacy set intersection solution that can be operated by a third party without frequent interactions, and let the online platform act as a third party to perform operations such as privacy set intersection to provide driver-passenger matching services. In contrast to the latest privacy protection schemes, this scheme performs better when the ride-hailing platform matches a large number of drivers and passengers. Under the premise of protecting the privacy and security of users, the larger the number of users, the more obvious the advantages of this scheme.

Computer Science