Practical Privacy-Preserving Ride Sharing Protocol with Symmetric Key

The advancement of mobile technologies and their ability to utilize the Global Positioning System (GPS) to accurately locate their substantial number of users, prompt Location-Based Services (LBS) significantly. Ride-sharing is a popular means of transportation that utilizes LBS. With the rapid development of smart cities and their impact on addressing the critical issues of urban life such as transportation, we can safely assume that the autonomous vehicles (AVs) will be a desired way of transportation in the near future. Therefore, the ride sharing service (RSS) providers will need to arrange their services via AVs. However, a user who wants to use a RSS has to submit their trip data (which contains location data) to the service provider. On one hand, the popularity of RSSs makes them an attractive target for cyber-attacks, and on the other hand, multiple studies show that a user’s location data can reveal sensitive information about that user. In this paper, we present a practical ride-sharing protocol for AVs that preserves both anonymity and location privacy of the users. Most of the previous works on the topic, does not provide security against malicious server and/or clients. Moreover, the previously proposed protocols rely on additional entities (e.g., a trusted third party) to satisfy the objectives of their protocols. In the presence of the malicious entities, our proposed protocol guarantees the security and privacy of the server and the clients, without relying on any additional parties. To the best of our knowledge, our protocol is the first scheme that satisfies perfect location privacy. We evaluate the performance of our protocol in a realistic setting and demonstrate its feasibility in the real life application areas, i.e., the protocol only requires 20 milliseconds to process and respond to 1000 simultaneous ride-sharing requests. Moreover, we propose a novel private sum aggregation (PSA) scheme that is designed for the use-cases where the private elements are chosen from a limited set. We believe that our novel PSA scheme may be of independent interest.