Dynamic Offloading Scheduling Scheme for MEC-enabled Vehicular Networks

Abstract

The incorporation of mobile edge computing (MEC) into vehicular networks is an inspiring approach to meet the explosive vehicular services demands. The mobile vehicles could execute the computing tasks locally or offload to the nearby MEC servers. However, the mobility of the vehicles brings great challenges to finish the offloading before the vehicles moving out of the coverage of original connected MEC servers. To avoid interruption in the offloading process, we investigate the problem and propose a dynamic offloading scheduling scheme for MEC-enabled vehicular networks. The limited resources and variable vehicle speeds are taken into consideration. The whole offloading task is partitioned into small task units (TUs). Then in every cell with respective MEC servers, the optimal offloading ratio and assigned TUs are derived according to the practical constraints like moving speed, cell coverage and transmit data rate. The proposed scheduling scheme is adaptive to the changes of the vehicles’ speeds and the wireless transmission environment. Simulation results show that the proposed scheme has a good performance in total offloading time and energy efficiency.