Proof-of-relevance: Filtering false data via authentic consensus in Vehicle Ad-hoc Networks

Emerging applications in vehicle ad hoc networks (VANETs) not only open tremendous business opportunities and navigation benefits; they also pose formidable research challenges in security provisioning. A critical security threat to VANETs is false data injection, i.e., an attacker disseminates false information to disrupt the behavior of the other drivers. Information driven operations of vehicular networks make false data injection a very effective attack. As the first line of defense, this paper presents the notion of proof-of-relevance (PoR), which consists in proving that the event reporter is authentically relevant to the event it has reported. The PoR is accomplished by collecting authentic consensus on the event from witness vehicles in a cooperative way. Event reports from attackers who fail to provide this PoR are disregarded, making the network immune to bogus data. Performance evaluation and security analysis demonstrate the efficiency and security of the proposed scheme.