Design and Evaluation of a Privacy-Preserving Architecture for Vehicle-to-Grid Interaction

摘要

Charging battery-electric vehicles can pose a significant load to the power grid. Letting a central instance control vehicle charging processes can reduce the grid load and allows for vehicles to be used as distributed grid resources. It is commonly assumed that vehicle owners are willing to reveal their driving patterns to the control instance. As we show, current privacy-preserving technologies can be used to construct an architecture that reduces the need to reveal such sensitive information. Yet, we identify limitations to such an approach and demonstrate how an adversary can use information inherent to the context to decrease vehicle owner privacy. As a concrete case, we discuss an adversary algorithm based on travel times and show how to obtain anonymity sets for individual vehicles. This allows us to make an important step towards understanding and quantifying privacy achievable in practice.