Privacy-Aware Dynamic Ride Sharing

Preeti Goel,Lars Kulik,Kotagiri Ramamohanarao

DOI: https://doi.org/10.1145/2845080

2016-01-01

ACM Transactions on Spatial Algorithms and Systems

Abstract:Dynamic ride sharing is a service that enables shared vehicle rides in real time and on short notice. It can be an effective solution to counter the problem of increasing traffic jams at peak hours in cities. The growing use and popularity of smart phones and GPS-enabled devices provides us with tools required to efficiently implement ride sharing and significantly enhance carpooling. However, privacy and safety concerns are the main obstacles faced when encouraging people to use such a service. In this work, we present “Match Maker,” a negotiation-based model that hides exact location information data for system participants while implementing privacy preserving ride sharing. We use the concept of imprecision (not being precise about location of the user out of set of n locations) and follow the idea of obfuscation, which equates a higher degree of imprecision with a higher degree of privacy. We identify two attack types that could circumvent privacy preserving ride sharing. We compare the Match Maker model with the standard central trusted server model collecting precise location data, which we term eBay model. We provide the first comprehensive approach that integrates privacy, safety and trust in a single model. We present a recursive ellipse-based algorithm to compute an optimal driver path as well as three negotiation strategies for drivers and passengers. We conduct extensive experiments on real road networks and compare the strategies for privacy and effectiveness of ride sharing in terms of traffic load and vehicle km reduction. We show that ride sharing saves between 9% and 21% (on average 12%) of vehicle km if drivers are only prepared to accept slight detours of their usual trips. In the city of Melbourne, with 11.6 million trips a weekday and an average trip length of 10.2 km, this would save 14.2 million km per weekday.