Trust Computation in VANETs

In this work, we study the application of Perron-Frobenius theorem for computing trust in the VANET environment. Safety critical and safety related messages in a VANET can lead to major changes in the behavior of vehicles moving on the road which can prevent unpleasant traffic situations. False messages can result in serious conditions like collisions. Trust management in VANETs is necessary to deter broadcast of selfish or malicious messages and also enable other vehicles to filter out such messages. A decentralized dynamic trust management system must be scalable with an ability to cope with sparsity of direct interactions. In this work, it is shown that messaging behavior of vehicles can be modeled as a primitive graph. This allows the application of Perron-Frobenius theorem. It is found that the eigenvalues of the matrix corresponding to the interaction graph can be used to compute trust values in the VANET setting.