Physical Layer Security of Vehicular Networks: A Stochastic Geometry Approach

In vehicular communications, the identities, position, actions, and trajectories of users should be secured perfectly for protecting the users' personal safety. Physical layer security (PLS) is an effective way for enhancing the information security. However, the study on the physical layer security of vehicular networks is scarce. To bridge the knowledge gap, this paper proposes an analysis framework for studying the PLS of the vehicular communications, where the spatial distribution of vehicles are modeled by the Cox processes driven by a common Poisson line process (PLP). Based on the maximum signal-to-interference-ratio (SIR) association scheme, we characterize the maximum secrecy performance of the network, where the coverage probability and secrecy probability are derived to facilitate the evaluation of the effective secrecy throughput. Numerical results are given for validating the analytical results and showing the impact of the key system parameters on the secrecy performance.