A Privacy-Friendly Framework for Vehicle-to-Grid Interactions

In the next decades, Electric Vehicles (EVs) are expected to gain increasing popularity and huge penetration in the automotive mar- ket, thanks to their potentialities for close interaction with the Smart Grid ecosystem. Firstly, recharging EV's batteries with energy produced by renewables will allow for a consistent reduction of pollution due to the carbon emissions of traditional gasoline combustion; secondly, bat- teries could be exploited to store/inject energy from/to the grid in order to compensate the unpredictable uctuations caused by Renewable En- ergy Sources (RES). To this aim, a load aggregator is envisioned as a scheduling entity to plan the EVs' battery recharge/discharge according to the user's needs and the current power generation of the grid. The main drawback of the introduction of such load aggregator is a potential harm of users' privacy: gathering information about the EVs' recharge requests and plug/unplug events could make the scheduler able to infer the private travelling habits of the customers, thus exposing them to the risk of tracking attacks and to other privacy threats. To address this issue, this paper proposes a security infrastructure for privacy-friendly Vehicle-to-Grid (V2G) interactions, which enables the load aggregator to schedule the EV's battery charge/discharge without learning the current battery level, nor the the amount of charged/discharged energy, nor the time periods in which the EVs are available for recharge. Our proposed scheduling protocol is based on the Shamir Secret Sharing scheme. We provide a security analysis of the privacy guarantees provided by our framework and compare its performance to the optimal schedule that would be obtained if the aggregator had full knowledge of the charging- related information.