Rapid Establishment of Transient Trust for NDN-Based Vehicular Networks
by Sanjeev Kaushik Ramani and Alex Afanasyev
Recent advances of vehicular networking technologies are giving birth to the emerging class of road safety and collaborative driving applications, where cars need to communicate and assist each other with various tasks, such as sharing views from onboard cameras, performing distributed predictions of car positions, etc. The proposed Named Data Networking (NDN) architecture promises a straightforward way to implement such applications with enhanced data delivery and security capabilities. However, one of the critical elements of vehicular applications is the need of reliable trust: the communicating cars must ensure that parties will do the job they are asked to do. Such trust, while can be mechanically implemented through the use of cryptographic signatures, requires the establishment of trust relations between the cars. This is especially a challenge, given the individual cars in a typical urban environment can very infrequently meet each other and each encounter could be very short-term. Although one can rely on a manufacturer-centric endorsement of car actions (e.g., by installing manufacturer certificates as global trust anchors, creating manufacturer-signed certs for individual cars, and trusting data that is signed by such certificates), it is not sufficient for the collaborative applications. Such certificate can only tell that data is coming from the specific car model, but does not guarantee the quality of the data/actions. This paper considers ideas and concepts developed based on the Swift Trust model and explores their use in vehicular environments with NDN-based communication. With Swift Trust, vehicles in the communication range can quickly make shortterm trust decisions for secure publishing, consumption, and processing of data (e.g., to cooperatively analyze the nearby environment for potential safety issues). As an initial step, the paper explores a task-oriented method of establishing trust based on request-response communication. The paper also highlights several potential threats and attacks and discusses possible directions to mitigate them.