Federated Learning (FL) is a distributed machine learning technique that allows model training among multiple devices or organizations by sharing training parameters instead of raw data. However, adversaries can still infer individual information through inference attacks (e.g. differential attacks) on these training parameters. As a result, Differential Privacy (DP) has been widely used in FL to prevent such attacks. We consider differentially private federated learning in a resource-constrained scenario, where both privacy budget and communication round are constrained. By theoretically analyzing the convergence, we can find the optimal number of differentially private local iterations for clients between any two sequential global updates. Based on this, we design an algorithm of differentially private federated learning with adaptive local iterations (ALI-DPFL). We experiment our algorithm on the FashionMNIST and CIFAR10 datasets, and demonstrate significantly better performances than previous work in the resource-constraint scenario.

翻译：联邦学习（FL）是一种分布式机器学习技术，它允许多个设备或组织通过共享训练参数而非原始数据来进行模型训练。然而，攻击者仍可通过针对这些训练参数的推理攻击（例如差分攻击）推断出个体信息。因此，差分隐私（DP）被广泛应用于联邦学习以防范此类攻击。我们考虑资源受限场景下的差分隐私联邦学习，其中隐私预算和通信轮次均受到限制。通过理论收敛性分析，我们能够找到任意两次连续全局更新之间客户端的最佳差分隐私局部迭代次数。基于此，我们设计了一种自适应局部迭代的差分隐私联邦学习算法（ALI-DPFL）。我们在FashionMNIST和CIFAR10数据集上对该算法进行了实验，结果表明，在资源受限场景下，其性能显著优于先前工作。