Aiding Traffic Prediction Servers through Self-Localization to Increase Stability in Complex Vehicular Clustering

Abstract

The integration of cellular networks and vehicular networks is complex and heterogeneous. Synchronization among vehicles in heterogeneous vehicular clusters plays an important role in effective data sharing and the stability of the cluster. This synchronization depends on the smooth exchange of information between vehicles and remote servers over the Internet. The remote servers predict road traffic patterns by adopting deep learning methods to help drivers on the roads. At the same time, local data processing at the vehicular cluster level may increase the capabilities of remote servers. However, global positioning system (GPS) signal interruption, especially in the urban environment, plays a big part in the detritions of synchronization among the vehicles that lead to the instability of the cluster. Instability of connections is a major hurdle in developing cost-effective solutions for deriving assistance and route planning applications. To solve this problem, a self-localization scheme within the vehicular cluster is proposed. The proposed self-localization scheme handles GPS signal interruption to the vehicle within the cluster. A unique clustering criterion and a synchronization mechanism for sharing traffic information system (TIS) data among multiple vehicles are developed. The developed scheme is simulated and compared with existing known approaches. The results show the better performance of our proposed scheme over others.