On Resilient and Efficient Linear Secure Aggregation in Hierarchical Federated Learning

2026-01-19

AI Summary

This work addresses the problem of minimizing communication and randomness overhead in secure aggregation under unreliable communication environments in hierarchical federated learning, where both client–relay and relay–server links may fail. Leveraging information-theoretic techniques combined with linear coding, shared randomness, and hierarchical topology modeling, the study establishes the theoretical limits of secure aggregation in this setting for the first time and proposes an optimal linear protocol that achieves these limits. The protocol ensures privacy while enabling efficient and robust model aggregation, with both communication and randomness costs attaining the derived theoretical lower bounds. Its optimality is rigorously confirmed through a converse proof, effectively bridging the gap between theoretical guarantees and practical deployment.

Technology Category

• Reasoning under Uncertainty: Stochastic Optimization
• Machine Learning: Distributed Machine Learning & Federated Learning
• Multiagent Systems: Agent Communication

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In this paper, we study the fundamental limits of hierarchical secure aggregation under unreliable communication. We consider a hierarchical network where each client connects to multiple relays, and both client-to-relay and relay-to-server links are intermittent. Under this setting, we characterize the minimum communication and randomness costs required to achieve robust secure aggregation. We then propose an optimal protocol that attains these minimum costs, and establish its optimality through a matching converse proof. In addition, we introduce an improved problem formulation that bridges the gap between existing information-theoretic secure aggregation protocols and practical real-world federated learning problems.

Problem

• hierarchical federated learning
• secure aggregation
• unreliable communication
• resilience
• communication cost

Innovation

• hierarchical federated learning
• secure aggregation
• unreliable communication
• information-theoretic security
• optimal protocol