An Adaptive Algorithm to Offload Task for User's QoE in Vehicular Edge System.

With the rapid development of vehicle edge offloading technology, smart device owners' quality of experience (QoE) requirements for computing offloading are gradually increasing. However, some dynamic or uncertain factors in the edge offloading scenario, such as the fluctuating network environment, battery energy consumption and other conditions, will affect users. Based on the multi-armed bandit theory, this paper proposes an adaptive learning algorithm that can dynamically sense environmental changes. Finally, the failure rate and offloading performance of large, medium and small tasks are simulated on the EdgeCloudSim simulation platform. The results show that the proposed algorithm has lower latency and energy consumption performance.