This website requires JavaScript.

Bayesian Optimization Enhanced Deep Reinforcement Learning for Trajectory Planning and Network Formation in Multi-UAV Networks

Shimin GongMeng WangBo GuWenjie ZhangDinh Thai HoangDusit Niyato
Dec 2022
摘要
In this paper, we employ multiple UAVs coordinated by a base station (BS) tohelp the ground users (GUs) to offload their sensing data. Different UAVs canadapt their trajectories and network formation to expedite data transmissionsvia multi-hop relaying. The trajectory planning aims to collect all GUs' data,while the UAVs' network formation optimizes the multi-hop UAV network topologyto minimize the energy consumption and transmission delay. The joint networkformation and trajectory optimization is solved by a two-step iterativeapproach. Firstly, we devise the adaptive network formation scheme by using aheuristic algorithm to balance the UAVs' energy consumption and data queuesize. Then, with the fixed network formation, the UAVs' trajectories arefurther optimized by using multi-agent deep reinforcement learning withoutknowing the GUs' traffic demands and spatial distribution. To improve thelearning efficiency, we further employ Bayesian optimization to estimate theUAVs' flying decisions based on historical trajectory points. This helps avoidinefficient action explorations and improves the convergence rate in the modeltraining. The simulation results reveal close spatial-temporal couplingsbetween the UAVs' trajectory planning and network formation. Compared withseveral baselines, our solution can better exploit the UAVs' cooperation indata offloading, thus improving energy efficiency and delay performance.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答