Service placement and migration algorithm utilizing precise positioning for connected and automated vehicles

In this paper, the goal is to reduce Multi-access Edge Computing (MEC) service placement and migration delay for Connected Automated Vehicles (CAV) by using the precise position of the vehicle. To reduce the migration process delay, the migration must start before the vehicle reaches the future serving node. Thus, an AI position-based scheme is proposed to predict candidate nodes for migration. Real-time precise positioning data is acquired from an RTK-GNSS measurements campaign. The obtained imbalanced raw data is treated and used with the prediction scheme, and the resulting prediction accuracy reaches up to 99.3%. Finally, we propose an algorithm to perform service placement and migration based on the position prediction, the algorithm shows around 50% latency reduction than core placement, and up to 29% compared to the benchmark prediction algorithm.