Verification of a VANET protocol IC-29 miniproject

This report summarizes a miniproject carried out in the scope or the IC-29 course (Self-Organized Wireless and Sensor Networks). The goal of the project was to verify a VANET protocol, namely the Distributed Revocation Protocol (DRP, [1]). This protocol was chosen because of its ad-hoc nature, which makes the verification both interesting and challenging, while being relatively simple, this protocol works. The initial idea was to apply (semi)-automatic verification techniques and tools, like the security model-checker OFMC ([3]). However, in the end manual verification was performed, because OFMC and similar security verification techniques proved not to be suitable for the purpose of DRP verification; additionally, it was possible to detect attacks against the protocol remaining on the non-automatic level. The report develops as follows. First, the VANET environment and the general revocation scheme is described, to give the reader an understanding where DRP fits in. Next, the definition of DRP that will be used for verification is introduced; this section is based on [1] and on interaction with the authors of this paper. The latter was necessary, because [1] does not describe in all detail how the protocol works. The following sections presents the adversary model and specify the properties that the protocol should fulfil (these were partly inspired by [4]). Finally, I give a list of discovered attacks, and conclude. This order of presentation, quite common in verification practice, is usual in cases where verification comes into the picture after the protocol is defined. This differs from the ideal verification approach, where the formal properties are specified before the protocol is defined.