On Using Secure Aggregation in Differentially Private Federated Learning with Multiple Local Steps
This work addresses communication bottlenecks for practitioners in privacy-sensitive federated learning, though it is incremental as it builds on existing techniques with a new analysis.
The paper tackles the problem of communication inefficiency in differentially private federated learning by enabling multiple local optimization steps while maintaining formal privacy guarantees through secure aggregation, resulting in higher utility models under limited communication rounds.
Federated learning is a distributed learning setting where the main aim is to train machine learning models without having to share raw data but only what is required for learning. To guarantee training data privacy and high-utility models, differential privacy and secure aggregation techniques are often combined with federated learning. However, with fine-grained protection granularities, e.g., with the common sample-level protection, the currently existing techniques generally require the parties to communicate for each local optimization step, if they want to fully benefit from the secure aggregation in terms of the resulting formal privacy guarantees. In this paper, we show how a simple new analysis allows the parties to perform multiple local optimization steps while still benefiting from using secure aggregation. We show that our analysis enables higher utility models with guaranteed privacy protection under limited number of communication rounds.