Social-Based Link Reliability Prediction Model for CR-VANETs

Jing Wang, Aoxue Mei, Xing Tang, Bing Shi

Abstract: Cognitive radio technology can improve the spectrum efficiency, and solve the problem of spectrum scarcity for vehicular communications. Nevertheless, the link reliability of cognitive radio vehicular ad hoc networks (CR-VANETs) is not only related to primary users but also to secondary users, which are acted as vehicles. To address this issue, we propose a social-based link reliability prediction model by jointly considering social characteristics of primary users and secondary users. First, we analyze the probability of the available channel in CR-VANETs based on social characteristics of primary users. Second, we analyze social characteristics of secondary users through the friendliness, the similarity and the centrality. Third, we utilize the social characteristics of primary users and secondary users to propose a link reliability prediction model and to predict the probability of the available link between two neighboring vehicles. Simulation results show that the predicted number of active primary users is consistent with the corresponding value in the real dataset, and the proposed social-based link reliability prediction model is effective.