MoCFL: Mobile Cluster Federated Learning Framework for Highly Dynamic Network

2025-03-03

AI Summary

To address feature distribution drift and data heterogeneity caused by frequent client join/leave in highly dynamic mobile edge clusters, this paper proposes a lightweight and robust federated learning framework. Methodologically, it introduces: (1) an affinity-matrix-based local feature aggregation mechanism that explicitly models inter-client feature similarity; (2) an incremental global classifier trained jointly on historical and current features to concurrently mitigate data drift and catastrophic forgetting; and (3) a dynamic topology-aware aggregation strategy adaptable to time-varying communication topologies. Extensive experiments on the UNSW-NB15 dataset demonstrate that the proposed approach significantly improves classification accuracy and model robustness under dynamic conditions, while maintaining low computational and communication overhead—making it particularly suitable for resource-constrained mobile edge environments.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Multiagent Systems: Multiagent Learning
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• 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
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Frequent fluctuations of client nodes in highly dynamic mobile clusters can lead to significant changes in feature space distribution and data drift, posing substantial challenges to the robustness of existing federated learning (FL) strategies. To address these issues, we proposed a mobile cluster federated learning framework (MoCFL). MoCFL enhances feature aggregation by introducing an affinity matrix that quantifies the similarity between local feature extractors from different clients, addressing dynamic data distribution changes caused by frequent client churn and topology changes. Additionally, MoCFL integrates historical and current feature information when training the global classifier, effectively mitigating the catastrophic forgetting problem frequently encountered in mobile scenarios. This synergistic combination ensures that MoCFL maintains high performance and stability in dynamically changing mobile environments. Experimental results on the UNSW-NB15 dataset show that MoCFL excels in dynamic environments, demonstrating superior robustness and accuracy while maintaining reasonable training costs.

Problem

• Addresses data drift in dynamic mobile clusters
• Enhances feature aggregation with affinity matrix
• Mitigates catastrophic forgetting in mobile FL

Innovation

• Affinity matrix for feature similarity quantification
• Integration of historical and current feature information
• Enhanced robustness in dynamic mobile environments