Biography
Fan Dong was a Master’s student at DENOS Lab from 2022 to 2024, working on federated learning, AI, and edge computing. His research covers federated learning across aerial and space networks, where client heterogeneity is extreme, and model aggregation strategies that promote data diversity. He is a co-author of the lab’s ACM Computing Surveys topology-aware federated learning survey.
Projects
-
Federated learning in aerial and space networks
Aggregation and weighting strategies for federated training across satellites and aircraft, where client heterogeneity is high and connectivity is intermittent.
Publications
- ACM CSUR2024
Topology-aware Federated Learning in Edge Computing: A Comprehensive Survey
- FL4Data-Mining2023
A Privacy-Preserving Hybrid Federated Learning Framework for Financial Crime Detection
- IEEE ICC-W2024
FedGreen: Carbon-aware Federated Learning with Model Size Adaptation
- IEEE WF-IoT2024
Navigating High-Degree Heterogeneity: Federated Learning in Aerial and Space Networks
- JNCA2025
Optimizing Federated Learning with Weighted Aggregation in Aerial and Space Networks
- 2023
WeiAvg: Federated Learning Model Aggregation Promoting Data Diversity
Achievements
- Third place in the Privacy-Enhancing Technologies (PETs) Prize Challenge at the 2023 Summit for Democracy, with a UCalgary and Michigan State University team, for federated detection of financial crime across banks without exposing personal data.