Traffic big data analysis supporting vehicular network access recommendation

With the explosive growth of Internet of Vehicles (IoV), it is undoubted that vehicular demands for real-time Internet access would get a surge in the near future. Therefore, it is foreseeable that the cars within the IoV will generate enormous data. On the one hand, the huge volume of data mean we could get much information (e.g., vehicle's condition and real-time traffic distribution) through the big data analysis. On the other hand, the huge volume of data will overload the cellular network since the cellular infrastructure still represents the dominant access methods for ubiquitous connections. The vehicular ad hoc network (VANET) offloading is a promising solution to alleviate the conflict between the limited capacity of cellular network and big data collection. In a vehicular heterogeneous network formed by cellular network and VANET, an efficient network selection is crucial to ensure vehicles' quality of service. To address this issue, we develop an intelligent network recommendation system supported by traffic big data analysis. Firstly, the traffic model for network recommendation is built through big data analysis. Secondly, vehicles are recommended to access an appropriate network by employing the analytic framework which takes traffic status, user preferences, service applications and network conditions into account. Furthermore an Android application is developed, which enables individual vehicle to access network automatically based on the access recommender. Finally, extensive simulation results show that our proposal can effectively select the optimum network for vehicles, and network resource is fully utilized at the same time.