版权所有:内蒙古大学图书馆 技术提供:维普资讯• 智图
内蒙古自治区呼和浩特市赛罕区大学西街235号 邮编: 010021
作者机构:Technische Universität Berlin Department of Electrical Engineering and Computer Science Berlin10587 Germany University of Minnesota Twin Cities Department of Electrical and Computer Engineering MinneapolisMN55455 United States The Hong Kong University of Science and Technology IoT Thrust Department of Computer Science and Engineering Sai Kung Hong Kong
出 版 物:《IEEE Journal on Selected Areas in Communications》 (IEEE J Sel Areas Commun)
年 卷 期:2023年第41卷第4期
页 面:977-989页
核心收录:
学科分类:0303[法学-社会学] 0810[工学-信息与通信工程] 0808[工学-电气工程] 080802[工学-电力系统及其自动化] 08[工学] 0835[工学-软件工程] 0812[工学-计算机科学与技术(可授工学、理学学位)] 081202[工学-计算机软件与理论]
基 金:European Research Council through the European Research Council (ERC) Advanced [Content-Aware Wireless Networks: Fundamental Limits, Algorithms, and Architectures (CARENET)] National Natural Science Foundation of China (NSFC) Foshan Hong Kong University of Science and Technology (HKUST) Projects
主 题:Information theory
摘 要:We propose SwiftAgg+, a novel secure aggregation protocol for federated learning systems, where a central server aggregates local models of $N \in \mathbb {N}$ distributed users, each of size $L \in \mathbb {N}$ , trained on their local data, in a privacy-preserving manner. SwiftAgg+ can significantly reduce the communication overheads without any compromise on security, and achieve optimal communication loads within diminishing gaps. Specifically, in presence of at most $D=o(N)$ dropout users, SwiftAgg+ achieves a per-user communication load of $\left({1+\mathcal {O}\left({\frac {1}{N}}\right)}\right)L$ symbols and a server communication load of $\left({1+\mathcal {O}\left({\frac {1}{N}}\right)}\right)L$ symbols, with a worst-case information-theoretic security guarantee, against any subset of up to $T=o(N)$ semi-honest users who may also collude with the curious server. Moreover, the proposed SwiftAgg+ allows for a flexible trade-off between communication loads and the number of active communication links. In particular, for $T © 1983-2012 IEEE.