Integrating Homomorphic Encryption and Synthetic Data in FL for Privacy and Learning Quality

2026-03-03

AI Summary

This work proposes Alternating Federated Learning (Alt-FL), a novel approach that addresses the trade-off between model performance and computational overhead in privacy-preserving federated learning, particularly when homomorphic encryption (HE) is employed. Alt-FL innovatively integrates synthetic data generation with a selective HE mechanism, alternating between real and synthetic data during local training and selectively transmitting either plaintext or encrypted model parameters across communication rounds. This strategy effectively mitigates data leakage risks, including deep leakage from gradients (DLG) attacks, while significantly improving model accuracy and reducing HE-related computational costs. Compared to existing selective HE methods, Alt-FL achieves a 13.4% increase in model accuracy and reduces homomorphic encryption computation overhead by up to 48%.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Search and Optimization: Learning to Search
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Security and Privacy: Privacy-enhancing technologies
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Federated learning (FL) enables collaborative training of machine learning models without sharing sensitive client data, making it a cornerstone for privacy-critical applications. However, FL faces the dual challenge of ensuring learning quality and robust privacy protection while keeping resource consumption low, particularly when using computationally expensive techniques such as homomorphic encryption (HE). In this work, we enhance an FL process that preserves privacy using HE by integrating it with synthetic data generation and an interleaving strategy. Specifically, our solution, named Alternating Federated Learning (Alt-FL), consists of alternating between local training with authentic data (authentic rounds) and local training with synthetic data (synthetic rounds) and transferring the encrypted and plaintext model parameters on authentic and synthetic rounds (resp.). Our approach improves learning quality (e.g., model accuracy) through datasets enhanced with synthetic data, preserves client data privacy via HE, and keeps manageable encryption and decryption costs through our interleaving strategy. We evaluate our solution against data leakage attacks, such as the DLG attack, demonstrating robust privacy protection. Also, Alt-FL provides 13.4% higher model accuracy and decreases HE-related costs by up to 48% with respect to Selective HE.

Problem

• Federated Learning
• Privacy Protection
• Learning Quality
• Homomorphic Encryption
• Resource Consumption

Innovation

• Homomorphic Encryption
• Synthetic Data
• Federated Learning
• Privacy Preservation
• Interleaving Strategy