Coding-Enforced Resilient and Secure Aggregation for Hierarchical Federated Learning
This work addresses privacy and reliability issues in hierarchical federated learning, which is incremental as it builds on existing HFL methods with coding enhancements.
The paper tackles the challenge of maintaining model accuracy and privacy in hierarchical federated learning under unreliable communication by proposing H-SecCoGC, a scheme that integrates coding strategies for structured aggregation, achieving significant improvements in robustness, privacy, and efficiency as demonstrated by theoretical and experimental results.
Hierarchical federated learning (HFL) has emerged as an effective paradigm to enhance link quality between clients and the server. However, ensuring model accuracy while preserving privacy under unreliable communication remains a key challenge in HFL, as the coordination among privacy noise can be randomly disrupted. To address this limitation, we propose a robust hierarchical secure aggregation scheme, termed H-SecCoGC, which integrates coding strategies to enforce structured aggregation. The proposed scheme not only ensures accurate global model construction under varying levels of privacy, but also avoids the partial participation issue, thereby significantly improving robustness, privacy preservation, and learning efficiency. Both theoretical analyses and experimental results demonstrate the superiority of our scheme under unreliable communication across arbitrarily strong privacy guarantees