Reinforcement Learning Based Task Offloading and Take-Back in Vehicle Platoon Networks.
2019
In this paper, a platoon-assisted vehicular edge computing (PVEC) system is proposed to enhance the efficiency and success of offloading, in which task flows can be migrated to the platoon members. Due to the speed change of Intelligent Connected Vehicles (ICVs) in the platoon, a task offloading and take-back scheme is proposed which can avoid task processing failures by resulting in link disconnection. Considering the multitask offloading system, a multi-leader multi-follower Stackelberg game (MLMF-SG) is formulated to analyse the incentives for task flows and resource allocation for platoon members. In MLMF-SG, task flows as the offloading service consumers are the leaders and the offloading ICVs as the offloading service providers are followers. Specially, we propose an optimization scheme based on Reinforcement Learning (RL) to tackle the price strategies of task flows, which maximizes the player revenues by jointly optimizing the price decision and computing resource allocation. Simulation results verify the relationships among offloading service consumers and providers and demonstrate the excellent adaptability of RL algorithm.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
16
References
3
Citations
NaN
KQI