Im­pro­ving Lo­ca­ti­on Pri­va­cy for the Elec­tric Ve­hi­cle Mas­ses

Tilman Frosch, Sven Schäge, Martin Goll, Thorsten Holz

TR-HGI-2013-001, Ruhr-Uni­ver­si­tät Bo­chum, Horst Görtz In­sti­tut für IT-Si­cher­heit (HGI), June 2013


Electric vehicles (EVs) are becoming increasingly popular, especially since we need alternatives to cars powered by fuel. One main characteristic of EVs is that conventional gas stations become superfluous: since even a quick charging cycle of an EV takes about 30 minutes for a full charge today, we need a more flexible way to charge EVs. As a result, networks with many thousands of so called charging stations (CS) are being built, where a car owner can plug in her car and charge it. The worrying side-effect of this change in how we charge cars is that suddenly this process becomes observable: while today everyone can buy fuel at a gas station in an anonymous way, e-mobility (and especially the billing process) changes the rules significantly and enables an observer to track where a user charges her car.

Simply replacing cash with e-cash would solve most privacy problems in this context. Correctly applied, e-cash can offer a strong protection for customers’ privacy, but lack comparable incentives for the vendor to use it.If vendors should endorse a certain solution, it needs to be beneficial (or at least acceptable) to both sides.

In this paper, we tackle this challenge and propose a system that balances the customer’s legitimate interest to preserve her location privacy with the vendor’s legitimate interest to prevent abuse and the legal requirement to be able to resolve disputes in front of a court of law. The system also supports to authenticate a user in a non-repudiable way in compliance with pre- and post-payed billing such that billing can be handled correctly. Our approach is based on a group signature scheme that we adapt to the setting of next-generation cars. To study the practical feasibility of the proposed system, we implemented a prototype and evaluate it both on a CS for EVs and also on a (simulated) backend. The evaluation results suggest that our system can process more than one million charging processes per hour using off-the-shelf hardware while enabling location privacy.