Secrecy Performance of Cognitive Vehicular Radio Networks: Joint Impact of Nodes Mobility and Imperfect Channel Estimates

In this paper, we investigate the joint impact of nodes mobility and imperfect channel estimates on the secrecy performance of a cognitive vehicular radio network under time-varying Rayleigh fading channels. The network consists of a secondary source, a secondary destination, a secondary passive eavesdropper, and a primary user receiver. We use the first-order auto-regressive process to model the impact of nodes mobility in the channel, and a linear minimum mean square error estimation process to estimate the channel state information. Under such consideration, we first formulate the end-to-end instantaneous signal-to-noise ratio (SNR) at the destination. Then, we derive the exact closed-form secrecy outage probability and intercept probability expressions for the considered system. Further, to gain more insights into the considered system, we derive an expression in asymptotically low-outage regime. We also present some key practical cases of interest to reveal valuable information about the system secrecy diversity order. Finally, the numerical and simulation studies corroborate our analytical findings.