Vulnerability windows in vehicular communications

Vehicular networking (VN) has reached a high level of maturity, but security mechanisms have only recently been explored. The nature of VNs requires that nodes not behaving as expected should be detected without delay and eliminated to avoid possibly fatal traffic accidents. This paper redefines and explores the problem of estimating the vulnerability window as the time it takes to detect and eliminate bad nodes. Furthermore, we analyse the complexities of estimating the maximum or tolerable vulnerability window. Theoretically, with knowledge of the actual vulnerability window and the tolerable vulnerability window, nodes will be able to make a trade-off between security and the performance of the VN protocol. We argue that estimating these values is not easily achievable as they are subject to the operational characteristics and complexities of the most prominent intrusion detection and eviction protocols for VNs. The outcome of this paper is a deeper understanding of how to respond to the limitations of trust establishment mechanism, misbehaviour detections systems and node eviction strategies in vehicular communications (VC).