arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2203.13060cs.ITcs.DCcs.LGmath.IT
收藏

SwiftAgg +:在安全聚合中实现渐进最优通信负载,用于联邦学习

SwiftAgg+: Achieving Asymptotically Optimal Communication Loads in Secure Aggregation for Federated Learning

Technische Universität Berlin(柏林工业大学);Sharif University of Technology(谢里夫理工大学);The Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州));The Hong Kong University of Science and Technology(香港科技大学)

Tayyebeh Jahani-Nezhad, Mohammad Ali Maddah-Ali, Songze Li, Giuseppe Caire

摘要:

我们提出了SwiftAgg+,一种新的安全聚合协议的联邦学习系统,其中中央服务器聚合本地模型的$N \in \mathbb{N}$分布式用户,每个大小$L \in \mathbb{N}$,训练他们的本地数据,在隐私保护的方式。SwiftAgg+可以在不影响安全性的情况下显著降低通信开销,并在减少间隙的情况下实现最佳通信负载。具体来说,在最多$D=o(N)$退出用户的情况下,SwiftAgg+实现了$(1+\mathcal{O}(\frac{1}{N}))L$符号的每用户通信负载和$(1+\mathcal{O}(\frac{1}{N}))L$符号的服务器通信负载,具有最坏情况的信息理论安全保证,针对可能与好奇服务器勾结的最多$T=o(N)$半诚实用户的任何子集。此外,所提出的SwiftAgg+允许在通信负载和活动通信链路的数量之间进行灵活的权衡。特别地,对于$T<N-D$和任何$K\in\mathbb{N}$,SwiftAgg+可以实现$(1+\frac{T}{K})L$符号的服务器通信负载,以及高达$(1+\frac{T+D}{K})L$符号的每用户通信负载,其中网络中成对活动连接的数量是$\frac{N}{2}(K+T+D+1)$。

英文摘要:

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 $(1+\mathcal{O}(\frac{1}{N}))L$ symbols and a server communication load of $(1+\mathcal{O}(\frac{1}{N}))L$ 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 $T<N-D$ and for any $K\in\mathbb{N}$, SwiftAgg+ can achieve the server communication load of $(1+\frac{T}{K})L$ symbols, and per-user communication load of up to $(1+\frac{T+D}{K})L$ symbols, where the number of pair-wise active connections in the network is $\frac{N}{2}(K+T+D+1)$.

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。

↑