A Survey of Research Advances in Federated Learning for Anomaly Detection
DOI:
https://doi.org/10.61173/k0jdzg31Keywords:
Federated Learning, Anomaly Detection, Unsupervised Learning, Model CompressionAbstract
In fields such as industrial control systems, cybersecurity, and finance, detecting abnormal behaviors is of utmost importance to ensure operational reliability and prevent potential losses. In recent years, the advantage of Federated Learning (FL) in addressing privacy protection challenges and enabling collaborative model training across distributed data sources has made it an increasingly important solution for anomaly detection tasks. This paper provides a comprehensive review of the application of FL in various anomaly detection scenarios, including cybersecurity, intelligent industry, and financial services. It systematically analyzes the performance of popular models such as RNN/LSTM, AE/VAE, GNN, and Transformer-based methods under non-IID data distributions common in federated settings. Furthermore, this study identifies and discusses several critical challenges faced in FL-based anomaly detection, such as data heterogeneity, communication overhead, robustness, and security threats. Corresponding strategies to mitigate these issues are also summarized. Experimental evidence and case studies demonstrate that FL can achieve efficient, privacy-preserving, and scalable distributed anomaly detection, and thus holds significant promise for deployment in real-world applications.
References
in sequence modeling. Their fusion (e.g., Graphormer, IEEE Access, 2020, 8: 140699–140725. GTN) enables models to learn both local dependencies [2] Li L, Fan Y, Tse M, Lin K Y, et al. A review of applications and global attention mechanisms, significantly improving in federated learning. Computers and Industrial Engineering,
detection accuracy and generalization. Designing efficient 2020, 149: 106854. algorithms that support local graph structure updates, [3] Rieke N, Hancox J, Li W, Milletari F, et al. The future of cross-client graph sharing, and heterogeneous attention digital health with federated learning. NPJ Digital Medicine,
fusion under federated settings will become a research 2020, 3(1): 119. hotspot. [4] Lim W Y B, Luong N C, Hoang D T, et al. Federated learning Introduction of blockchain and trustworthy collaboration in mobile edge networks: a comprehensive survey. IEEE
mechanisms is building new security architectures for Communications Surveys & Tutorials, 2020, 22(3): 2031–2063. federated learning. In multi-party federated systems, the [5] LI Xiaoming, WANG Yifan, ZHANG Tao. Clustered authenticity of model updates, client trustworthiness, and federated learning architecture for network anomaly detection in overall fairness of collaboration are major challenges. large-scale heterogeneous IoT networks. Computers & Security,
Blockchain technology, with its immutability, distributed Elsevier, 2023, 126: 103299. ledger, and smart contract capabilities, can realize trans- [6] Alsulaimawi Z. Anomaly detection via gradient and parent auditing of the training process and automated autoencoder analysis in federated learning. arXiv preprint punishment rules. Future directions include designing on- arXiv:2403.10000, 2024. chain model aggregation mechanisms, combining with [7] Husnoo M A, Anwar A, Haque M E, et al. Decentralized differential privacy and secure multi-party computation federated anomaly detection in smart grids: A P2P gossip (SMC) to achieve multi-dimensional defense, and opti- approach. arXiv:2407.15879v2, 2025-01-09. mizing low-latency, high-throughput blockchain architec- [8] Zhang Y, Suleiman B, Alibasa M J, Farid F. Privacy-aware tures for federated systems. anomaly detection in IoT environments using FedGroup: a group-based federated learning approach. Journal of Network
and Systems Management, 2024, 32(1): 20. 4. Conclusion [9] Zhang C, Yang S, Mao L, Ning H. Anomaly detection and This paper reviews recent advances in federated learn- defense techniques in federated learning: a comprehensive
ing for anomaly detection, highlighting key applications review. Artificial Intelligence Review, 2024, 57(6): 150. including network security, industrial monitoring, and [10] Xu R, Miao H, Wang S, Yu P S, Wang J. PeFAD: A financial fraud detection. It examines major models and parameter-efficient federated framework for time series anomaly technical approaches, demonstrating how federated learn- detection. arXiv preprint arXiv:2406.02318, 2024. ing effectively overcomes data silos and privacy concerns, [11] NARDI Mirko, VALERIO Lorenzo, PASSARELLA thereby enhancing both the security and practicality of Andrea. Anomaly detection through unsupervised federated anomaly detection systems. learning. arXiv preprint, arXiv:2209.04184, 2022-09-09. Despite these advances, challenges such as Non-IID data [12] ZHANG Hengrun, ZENG Kai, LIN Shuai. Federated Graph modeling, communication overhead, vulnerability to mali- Neural Network for Fast Anomaly Detection in Controller Area cious attacks, and limited generalization in heterogeneous Networks. IEEE Transactions on Information Forensics and
environments persist. Future research directions include Security, IEEE, 2023, 18(5): 1566-1579. focusing on unsupervised learning methods, model com-
Downloads
Published
Issue
Section
License
Copyright (c) 2025 by the authors.
This work is licensed under a Creative Commons Attribution 4.0 International License.
