arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2203.01008cs.ITcs.LGeess.SPmath.IT
收藏

基于Mesh网络的无人机辅助分散学习

UAV-Aided Decentralized Learning over Mesh Networks

EURECOM(欧洲通信学院)

Matteo Zecchin, David Gesbert, Marios Kountouris

摘要:

分散式学习使无线网络设备能够协作地训练仅依赖于设备到设备(D2D)通信的机器学习(ML)模型。众所周知,分散优化算法的收敛速度严重依赖于网络的连通性程度,网络拓扑结构越密集,收敛时间越短。因此,由于其无线节点的通信范围有限,现实世界的网状网络的局部连接性破坏了分散式学习协议的效率,使其可能无法实现。在这项工作中,我们调查的作用,无人驾驶航空器(UAV),用作飞行接力,在促进分散的学习程序,在这种具有挑战性的条件。我们提出了一个优化的无人机轨迹,这是定义为一系列的航路点,无人机访问顺序,以便在稀疏连接的用户组之间传输情报。然后,我们提供了一系列的实验,突出了无人机在网状网络分散学习的背景下的重要作用。

英文摘要:

Decentralized learning empowers wireless network devices to collaboratively train a machine learning (ML) model relying solely on device-to-device (D2D) communication. It is known that the convergence speed of decentralized optimization algorithms severely depends on the degree of the network connectivity, with denser network topologies leading to shorter convergence time. Consequently, the local connectivity of real world mesh networks, due to the limited communication range of its wireless nodes, undermines the efficiency of decentralized learning protocols, rendering them potentially impracticable. In this work we investigate the role of an unmanned aerial vehicle (UAV), used as flying relay, in facilitating decentralized learning procedures in such challenging conditions. We propose an optimized UAV trajectory, that is defined as a sequence of waypoints that the UAV visits sequentially in order to transfer intelligence across sparsely connected group of users. We then provide a series of experiments highlighting the essential role of UAVs in the context of decentralized learning over mesh networks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑