Scholarlink

This chapter explores the role of graph-based methods in strengthening resilience and security within FL systems. Given the highly distributed and heterogeneous nature of FL, graph representations provide a powerful tool to model relationships among clients, data distributions, and model updates [10]. By capturing structural dependencies through graph neural networks (GNNs), VGAEs, and attention mechanisms, adversarial behaviors such as poisoning or inference attacks can be more effectively detected and mitigated. This chapter examines how graph-based modeling enhances the robustness of aggregation, supports anomaly detection, and facilitates secure knowledge transfer in dynamic and resource-constrained environments.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Resilience and Security of Graph-Based Federated Learning

  • 章节

  • 发表日期：

    2026年

查看全文
  • Kai Li,
  • Xin Yuan,
  • Wei Ni

摘要

This chapter explores the role of graph-based methods in strengthening resilience and security within FL systems. Given the highly distributed and heterogeneous nature of FL, graph representations provide a powerful tool to model relationships among clients, data distributions, and model updates [10]. By capturing structural dependencies through graph neural networks (GNNs), VGAEs, and attention mechanisms, adversarial behaviors such as poisoning or inference attacks can be more effectively detected and mitigated. This chapter examines how graph-based modeling enhances the robustness of aggregation, supports anomaly detection, and facilitates secure knowledge transfer in dynamic and resource-constrained environments.