CorBin-FL: A Differentially Private Federated Learning Mechanism using Common Randomness

TL;DR

CorBin-FL introduces a novel differentially private federated learning mechanism using correlated stochastic quantization, achieving a better privacy-utility trade-off and outperforming existing methods on standard datasets.

Contribution

The paper presents CorBin-FL, a new privacy mechanism employing correlated binary stochastic quantization for differential privacy in federated learning.

Findings

CorBin-FL achieves superior model accuracy compared to Gaussian and Laplacian mechanisms.

Theoretical analysis confirms optimal privacy-utility trade-off.

Experimental results on MNIST and CIFAR10 validate effectiveness.

Abstract

Federated learning (FL) has emerged as a promising framework for distributed machine learning. It enables collaborative learning among multiple clients, utilizing distributed data and computing resources. However, FL faces challenges in balancing privacy guarantees, communication efficiency, and overall model accuracy. In this work, we introduce CorBin-FL, a privacy mechanism that uses correlated binary stochastic quantization to achieve differential privacy while maintaining overall model accuracy. The approach uses secure multi-party computation techniques to enable clients to perform correlated quantization of their local model updates without compromising individual privacy. We provide theoretical analysis showing that CorBin-FL achieves parameter-level local differential privacy (PLDP), and that it asymptotically optimizes the privacy-utility trade-off between the mean square error utility measure and the PLDP privacy measure. We further propose AugCorBin-FL, an extension that, in addition to PLDP, achieves user-level and sample-level central differential privacy guarantees. For both mechanisms, we derive bounds on privacy parameters and mean squared error performance measures. Extensive experiments on MNIST and CIFAR10 datasets demonstrate that our mechanisms outperform existing differentially private FL mechanisms, including Gaussian and Laplacian mechanisms, in terms of model accuracy under equal PLDP privacy budgets.