Mobile-edge CoMputing for VehiCular networks

Kecheng Zhang
Abstract:1556-6072/17©2017ieee ieee vehicular technology magazine | June 2017 Cloud-based vehicular networks are a promising paradigm to improve vehicular services through distributing computation tasks between remote clouds and local vehicular terminals. To further reduce the latency and the transmission cost of the computation off-loading, we propose a cloud-based mobileedge computing (MEC) off-loading framework in vehicular networks. In this framework, we study the effectiveness of the computation transfer strategies with vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communication modes. Considering the time consumption of the computation task execution and the mobility of the vehicles, we present an efficient predictive combination-mode relegation scheme, where the tasks are adaptively off-loaded to the MEC servers through direct uploading or predictive relay transmissions. Illustrative results indicate that our proposed scheme greatly reduces the cost of computation and improves task transmission efficiency.
What problem does this paper attempt to address?