Efficient Secure Federated Learning via Information-Theoretically Secure Key Distribution: A Medical Imaging Case Study

2026-10-05

AI Summary

This study addresses the communication bottleneck in federated learning caused by the limited information-theoretically secure key generation rate. To overcome this, we propose an efficient framework that integrates physical-layer information-theoretic security with model compression. Through the synergistic design of backbone freezing, knowledge distillation, and model quantization, our approach transcends the limitations of conventional computational security, achieving deep integration between security mechanisms and the training pipeline, as validated on a real-world physical testbed. The proposed method reduces key consumption by approximately 35-fold while preserving predictive accuracy, thereby effectively preventing buffer depletion and key expiration. Consequently, it ensures both the sustainability and security of federated training in sensitive applications such as medical imaging.

Abstract

Federated Learning (FL) enables collaborative training of models across institutions without centralizing sensitive data, making it well-suited for privacy-concerned applications, such as medical imaging. To protect FL model updates during secure aggregation, additive masking is commonly employed. However, its underlying classical key establishment is only computationally secure. On the other hand, physics-based Information-Theoretically Secure (ITS) key exchange introduces practical constraints: finite key generation rates and time-limited storage severely limit throughput and sustained training of uncompressed models. In this work, we address this bottleneck by developing an FL framework that integrates frozen backbones, knowledge distillation, and quantization. These techniques reduce communication payload and, consequently, key material consumption. Moving beyond simulation, we benchmark this framework on a real physics-based key distribution testbed involving a chest X-ray classification application. Our results show that key usage can be reduced by $\sim$35$\times$ while maintaining predictive accuracy. This prevents buffer depletion and key expiration, enabling sustainable FL training under physical key generation constraints.

Problem

• Federated Learning
• Information-Theoretic Security
• Key Distribution
• Secure Aggregation
• Medical Imaging

Innovation

• Federated Learning
• Information-Theoretic Security
• Knowledge Distillation
• Quantization
• Medical Imaging