AdeptHEQ-FL: Adaptive Homomorphic Encryption for Federated Learning of Hybrid Classical-Quantum Models with Dynamic Layer Sparing

2025-07-09

AI Summary

Addressing the challenge of simultaneously ensuring privacy preservation, model performance, and communication efficiency in non-IID federated learning, this paper proposes the first classical-quantum hybrid federated learning framework. The framework integrates a convolutional neural network (CNN) with parameterized quantum circuits (PQCs) and incorporates three key mechanisms: differential privacy–based weighted aggregation, selective homomorphic encryption, and dynamic layer freezing. These components jointly guarantee rigorous privacy—namely, ε-differential privacy and semantic security—while enhancing quantum representation capability and training efficiency. Extensive experiments on CIFAR-10 demonstrate that our method achieves 25.43% and 14.17% higher accuracy than Standard-FedQNN and FHE-FedQNN, respectively, while reducing communication overhead by approximately 40%. To the best of our knowledge, this is the first work to achieve a unified optimization of strong privacy guarantees, high model accuracy, and low communication cost in non-IID quantum-enhanced federated learning.

Technology Category

• Machine Learning: Quantum Machine Learning
• Search and Optimization: Non-convex Optimization
• Computer Vision: Learning & Optimization for CV

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Security and Privacy: Security and privacy of machine learning and AI applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Federated Learning (FL) faces inherent challenges in balancing model performance, privacy preservation, and communication efficiency, especially in non-IID decentralized environments. Recent approaches either sacrifice formal privacy guarantees, incur high overheads, or overlook quantum-enhanced expressivity. We introduce AdeptHEQ-FL, a unified hybrid classical-quantum FL framework that integrates (i) a hybrid CNN-PQC architecture for expressive decentralized learning, (ii) an adaptive accuracy-weighted aggregation scheme leveraging differentially private validation accuracies, (iii) selective homomorphic encryption (HE) for secure aggregation of sensitive model layers, and (iv) dynamic layer-wise adaptive freezing to minimize communication overhead while preserving quantum adaptability. We establish formal privacy guarantees, provide convergence analysis, and conduct extensive experiments on the CIFAR-10, SVHN, and Fashion-MNIST datasets. AdeptHEQ-FL achieves a $approx 25.43%$ and $approx 14.17%$ accuracy improvement over Standard-FedQNN and FHE-FedQNN, respectively, on the CIFAR-10 dataset. Additionally, it reduces communication overhead by freezing less important layers, demonstrating the efficiency and practicality of our privacy-preserving, resource-aware design for FL.

Problem

• Balancing model performance, privacy, and efficiency in Federated Learning
• Addressing quantum-enhanced expressivity and privacy in hybrid FL models
• Reducing communication overhead while preserving privacy in FL

Innovation

• Hybrid CNN-PQC architecture for expressive learning
• Selective homomorphic encryption for secure aggregation
• Dynamic layer-wise adaptive freezing reduces overhead