Mobility Aware Task Migration in Vehicular Edge Computing Networks

With the emergence of autonomous vehicles and the ever-increasing data reported, providing the required latency and computational capabilities is becoming challenging. To address this issue, multi-access edge computing (MEC) for 3GPP 5G Cellular vehicles to everything (C-V2X) has been proposed recently. In this paper, we propose the Mobility Aware Task Migration (MATM) algorithm that strives the limitations of the benchmark algorithms that the MEC orchestrator utilizes for offloading tasks from vehicles. We provide a detailed simulation-based study for End-to-End (E2E) delay by serving the safety application as a function of different network densities. Firstly, we propose an extension to the location service defined in the ETSI MEC reference architecture; this extension enables the MEC service to send a migration notification to the MEC orchestrator to start the migration. Secondly, we introduce an enhancement to the MEC orchestrator module to enable task migration during vehicle mobility by selecting the most suitable MEC host with the closest proximity to the vehicle. Additionally, we use the Simu5G simulator to implement our scenario and conduct an evaluation of task offloading algorithms, and a detailed E2E delay analysis. Simulation results highlight the ability of our proposed algorithm to minimize E2E delay while ensuring fairness in resource allocation.