Security and privacy in VANET-based intelligent transport systems (ITS).

Abstract

The hostile environment in which Vehicular Ad hoc Networks (VANETs) operate make them vulnerable to a plethora of security and privacy issues. VANET technology is applicable in a variety of domains such as Intelligent Transport Systems (ITS) and often exhibit characteristics similar to Delay Tolerant Networks (DTNs) such as inadequate infrastructure and unreliable connectivity especially at the early stages of deployment. Hence, security and privacy techniques proposed for conventional VANETs may not scale well in such early deployment scenarios. In this thesis, novel security and location privacy schemes based on pseudonym communication, mix-zone, group communication, trust and certificate revocation are explored. First, the location privacy issue in DTN-based VANETs is analysed. Secondly, the use of an On-the-Fly Mix-zone Scheme (OFMS) together with a Privacy-preSerVing (PriSerV) scheme where vehicles can form mix-zones without the need for a Road Side Unit (RSU) is proposed. An extension to the PriSerV scheme considers full-scale deployment using a distributed architecture which represents later stages of deployment. Lastly, a Delay Tolerant Revocation Scheme (DTRvS) is proposed. In the scheme, malicious vehicles that drop packets are excluded from network communication. The opinion of other vehicles is used to identify misbehaving vehicles. The proposed OFMS shows an 8% increase in pseudonym change for a vehicle density of 250 vehicles. By varying the vehicle density from 50 to 250, there is about 6 to 10% increase in the pseudonym change compared to the basic scheme. When compared to the baseline scheme, the proposed OFMS-x show a 50% increase in the delivery probability and has a considerable decrease in the delivery overhead even with the increase in the number of vehicles in the network. In terms of the achievable location privacy, the proposed scheme performed better than AVATAR and the traditional mix-zone schemes. The security and efficiency analysis of the proposed DTRvS show that it is more robust compared to the baseline schemes. The delivery probability improved by 2 to 11% with DTRvS when 0 to 50% malicious vehicles are in the network. The chances of CRL delivery increases with the increase in the number of RSUs. Again, there is about 5% improvement when DTRvS is used compared to the use of native Spray & Wait-based routing scheme.