Pseudonym Limitation for Privacy in Cooperative Transport Systems

摘要

C-ITS is one of the most challenging domains in either research or industry. Their social and economic benefits are of importance for fuel consumption, traffic management, and road safety. Consequently, they have been deployed in many countries. Privacy is a critical feature of any shared system. This issue is crucial for deployed C-ITS in order to avoid tracking drivers who send continuous messages containing their locations. Anonymity is ensured by changing the identity of drivers frequently. In the meantime, authentication of messages is highly required. For this reason, adapted public key infrastructure) offers each vehicle a set of pseudonym certificates, which will be used to sign each sent message and will be used as a part of the identity. The identity change is done when the driver changes the pseudonym certificate. This change adds confusion for the tracker and protects the driver's privacy. In this article, we analyze a dataset of messages sent by vehicles, and we show, under some conditions, that it is possible to identify a driver's behavior based only on data correlation between collected data over the air. Then we can track drivers even when they change their pseudonyms frequently. Finally, we give some hints on how to protect drivers' privacy by adding some other confusion mechanisms such as introducing silent periods and exchanging pseudonyms with neighbors.