检索结果-内蒙古大学图书馆

SSRN 2024年

作者： Tang, Qiang Li, Bao Wang, Jin Song, Yun Alfarraj, Osama Tolba, Amr SchoolofComputer Science andCommunication Engineering Changsha University ofScience andTechnology Hunan Changsha China Hunan ProvincialKey Laboratory ofIntelligentProcessing ofBig Data on Transportation SchoolofComputer andCommunication Engineering Changsha University ofScience andTechnology Changsha410004 China Hunan University ofScience andTechnology Hunan Xiangtan China Computer Science Department Community College King SaudUniversity Riyadh11437 Saudi Arabia

Unmanned aerial vehicle(UAV) assisted federated learning(FL) can complete machine learning(ML) model training without relying on a large number of ground basestation(BS) facilities. However, how to efficiently and quickly select high-value user equipments(UEs) to participate in FL is the key to reducing the latency of FL training and improving the accuracy of FL model. Therefore, we proposed a fast decision-making framework for multi-UAV-assisted FL in this paper, which can ensure that users perform model training locally through distributed computing without uploading any private data. In order to perform efficient FL in UAV networks, an algorithm for selecting high-value UEs to participate in FL model training is proposed. On this basis, we designed a low-complexity algorithm to optimize the hovering position of UAVs to ensure that the UAV network can cover all UEs in each global round. For the purpose of applying the framework in actual scenarios, we formulated an optimization problem by considering dynamic changes in UE location and UE computing frequency, and solved it through the proposed low complexity algorithm. Numerical simulation results showed that our proposed algorithm has higher global model accuracy and lower total model training latency. © 2024, The Authors. All rights reserved.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：