论文标题
Swiftagg+：在安全汇总的联盟学习中实现渐近的最佳沟通负载
SwiftAgg+: Achieving Asymptotically Optimal Communication Loads in Secure Aggregation for Federated Learning
论文作者
论文摘要
我们提出了Swiftagg+，这是一种针对联合学习系统的新颖的安全聚合协议，其中中央服务器汇总了$ n \ in \ Mathbb {n} $分布式用户的本地型号，每个大小$ l \ in \ Mathbb {n} $中的每个型号都以私密性的方式对本地数据进行了培训。 Swiftagg+可以大大减少通信开销，而无需对安全性进行任何妥协，并在缩小间隙内实现最佳通信负载。具体而言，在最多存在$ d = o（n）$下去用户的情况下，Swiftagg+实现了$（1+ \ Mathcal {o}（\ frac {1} {n} {n}））L $符号的每用户通信负载，而服务器通信的负载和$（1+\ nathcal \ nathcal {of n $} n $ lac {n $ lac）信息理论安全保证，与最多$ t = o（n）$半honest用户的任何子集，他们也可能与好奇的服务器相关。此外，拟议的Swiftagg+允许在通信负载和主动通信链接的数量之间进行灵活的权衡。特别是，对于$ t <n-d $，对于任何$ k \ in \ mathbb {n} $，Swiftagg+可以实现$（1+ \ frac {t} {k} {k}）l $符号的服务器通信负载，并且每个用户的通信负载最高$（1+ \ frac \ frac \ frac \ frac {t+d} n $ simption insition insity Is Contrention ins Attive n syment ins contery n syment in Crome connection n syment n syment in Crome nornection。 $ \ frac {n} {2}（k+t+d+1）$。
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 $\frac{N}{2}(K+T+D+1)$.
