Standing Firm in 5G: A Single-Round, Dropout-Resilient Secure Aggregation for Federated Learning

2025-05-11

AI Summary

To address the bottlenecks of high communication overhead, poor robustness, and reliance on fixed infrastructure in secure aggregation protocols for federated learning under 5G’s high-mobility and high-packet-loss conditions, this paper proposes the first single-round, base-station-assisted, packet-loss-resilient secure aggregation protocol. Methodologically, it integrates key-homomorphic pseudorandom functions (KH-PRFs), t-out-of-k secret sharing, and precomputation optimizations to achieve end-to-end privacy preservation despite dynamic edge-device join/leave events. Experiments demonstrate a 62% reduction in communication overhead, a 5.3× decrease in aggregation latency, and guaranteed security and correctness even with up to 90% random client dropouts. The protocol significantly enhances the feasibility of deploying real-time federated learning in 5G networks.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Multiagent Systems: Agent Communication
• Search and Optimization: Learning to Search

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing

Abstract

Federated learning (FL) is well-suited to 5G networks, where many mobile devices generate sensitive edge data. Secure aggregation protocols enhance privacy in FL by ensuring that individual user updates reveal no information about the underlying client data. However, the dynamic and large-scale nature of 5G-marked by high mobility and frequent dropouts-poses significant challenges to the effective adoption of these protocols. Existing protocols often require multi-round communication or rely on fixed infrastructure, limiting their practicality. We propose a lightweight, single-round secure aggregation protocol designed for 5G environments. By leveraging base stations for assisted computation and incorporating precomputation, key-homomorphic pseudorandom functions, and t-out-of-k secret sharing, our protocol ensures efficiency, robustness, and privacy. Experiments show strong security guarantees and significant gains in communication and computation efficiency, making the approach well-suited for real-world 5G FL deployments.

Problem

• Secure aggregation in 5G FL faces dropout and mobility challenges
• Existing protocols lack efficiency due to multi-round communication
• Lightweight single-round protocol needed for 5G FL privacy

Innovation

• Single-round secure aggregation protocol
• Base stations for assisted computation
• Key-homomorphic pseudorandom functions