A Comparison Study of Connected Vehicle Systems between Named Data Networking and IP

Abstract

Connected vehicle (CV) system has been a very promising technology to improve transportation safety and efficiency, which is a crossing field of intelligent transportation systems (ITS) and the internet of things (IoT). The popular IP-based network protocol has been employed in most CV applications. However, the host-tohost feature of IP protocol is low-efficient for information dissemination. Further more, IP protocol is limited to support direct communication among vehicles with highly ad hoc connectivity. Recently, the next generation network, Named Data Networking (NDN), has been proposed to be a very promising technology to address the information dissemination problem of IP protocol, particularly in cloud computing-based applications and services. In this paper, SimIVC-NDN, a federated simulation platform with the capability of performing a microscopic traffic simulation with both NDN and IP-based networking is proposed. With SimIVC-NDN, we have conducted a quantitative simulation comparison of two CV systems powered by NDN and IP protocols respectively for image dissemination, a common cloud computing service. In the experiments CV systems are constructed based on a calibrated traffic model of Whitemud Drive at Edmonton, Canada. The simulation results show that the NDN-based CV system lowers the packet delay by two orders of magnitude compared with the IP-based one with low packet loss rate, indicating that a NDN-based networking is a promising alternative to the conventional IP-based one for cloud computing applications of CV systems.