Efficient and Proactive V2V Information Diffusion Using Named Data Networking

Yang Wang,Hengchang Liu,Liusheng Huang,John Stankovic

DOI: https://doi.org/10.1109/iwqos.2016.7590391

2016-01-01

Abstract:Due to high mobility and intermittent connections in vehicular networks, reliable and efficient Vehicle-to-Vehicle (V2V) communication is a challenging task. The Named Data Networking (NDN) paradigm is recently being applied to achieve efficient V2V communication, however, proactive V2V information diffusion conflicts with the receiver-initiated nature of NDN. This paper bridges this gap by exploiting hierarchical data names to achieve efficient and proactive V2V information diffusion. We first identify a popular subgroup of vehicles, then select them as the diffusion seeds with 3G/4G capability, while others are only equipped with short-range V2V communication. We also design a namespace-based method to optimize data transmission when vehicles are close, in order to maximize the information distribution across geographical space. We evaluate our solution via a real-world taxicab dataset. Experimental results demonstrate that our approach significantly outperforms state-of-the-art solutions in terms of diffusion speed and success rate of data retrieval.