The emergence of augmented reality (AR), autonomous driving and other new applications have greatly enriched the functionality of the vehicular networks. However, these applications usually require complex calculations and large amounts of storage, which puts tremendous pressure on traditional vehicular networks. Mobile edge computing (MEC) is proposed as a prospective technique to extend computing and storage resources to the edge of the network. Combined with MEC, the computing and storage capabilities of the vehicular network can be further enhanced. Therefore, in this paper, we explore the novel collaborative vehicular edge computing network (CVECN) architecture. We first review the work related to MEC and vehicular networks. Then we discuss the design principles of CVECN. Based on the principles, we present the detailed CVECN architecture, and introduce the corresponding functional modules, communication process, as well as the installation and deployment ideas. Furthermore, the promising technical challenges, including collaborative coalition formation, collaborative task offloading and mobility management, are presented. And some potential research issues for future research are highlighted. Finally, simulation results are verified that the proposed CVECN can significantly improve network performance.

Additional Metadata
Keywords architecture design, computation offloading, edge computing, Vehicular network
Persistent URL dx.doi.org/10.1109/ACCESS.2019.2957749
Journal IEEE Access
Citation
Xie, R. (Renchao), Tang, Q. (Qinqin), Wang, Q. (Qiuning), Liu, X. (Xu), Yu, F.R, & Huang, T. (Tao). (2019). Collaborative Vehicular Edge Computing Networks: Architecture Design and Research Challenges. IEEE Access, 7, 178942–178952. doi:10.1109/ACCESS.2019.2957749