GroupSecAgg: Information Theoretic Secure Aggregation with Uncoded Groupwise Keys
GroupSecAgg: Information Theoretic Secure Aggregation with Uncoded Groupwise Keys
复制标题
DOI：
10.1109/icc45041.2023.10279171
复制
发表时间：
2023-05
期刊：
影响因子：
--
通讯作者：
Kai Wan;Yin Yao;Hua Sun;Mingyue Ji;Giuseppe Caire
中科院分区：
文献类型：
--
作者：
Kai Wan;Yin Yao;Hua Sun;Mingyue Ji;Giuseppe Caire
安全聚合是联邦学习的核心组成部分，它在中央服务器上聚合来自分布式用户的本地训练模型，而不泄露有关本地用户数据的任何其他信息。本文遵循最新的带有用户丢失的信息论安全聚合问题，其目标是表征模型聚合期间从 $\mathrm{K}$ 用户到服务器的最小通信成本。所有现有的安全聚合协议都允许用户共享和存储编码密钥以保证安全。基于未编码的分组密钥更方便共享并且可以在大范围的实际应用中使用的动机，本文首次考虑未编码的分组密钥，其中密钥是相互独立的，并且每个密钥由一组$\mathrm{S}$用户共享。我们表明，如果 $\mathrm{S}$ 超出阈值，则使用未编码分组密钥的新安全聚合协议（称为 GroupSecAgg）可以实现与使用编码密钥的最佳协议相同的最佳通信成本。 Amazon EC2 上的实验表明，与最先进的技术相比，密钥共享和模型聚合时间有了相当大的改进。
Secure aggregation, which is a core component of federated learning, aggregates locally trained models from distributed users at a central server, without revealing any other information about the local users' data. This paper follows a recent information theoretic secure aggregation problem with user dropouts, where the objective is to characterize the minimum communication cost from the $\mathrm{K}$ users to the server during the model aggregation. All existing secure aggregation protocols let the users share and store coded keys to guarantee security. On the motivation that uncoded groupwise keys are more convenient to be shared and could be used in large range of practical applications, this paper is the first to consider uncoded groupwise keys, where the keys are mutually independent and each key is shared by a group of $\mathrm{S}$ users. We show that if $\mathrm{S}$ is beyond a threshold, a new secure aggregation protocol with uncoded groupwise keys, referred to as GroupSecAgg, can achieve the same optimal communication cost as the best protocol with coded keys. The experiments on Amazon EC2 show the considerable improvements on the key sharing and model aggregation times compared to the state-of-the art.