Federated Learning in Cybersecurity: Enhancing Decentralized Threat Detection
Sai Bhuvana Kurada · IOSR Journal of Mobile Computing & Application · 2025
Federated Learning (FL) is an emerging machine learning paradigm with transformative potential in the field of cybersecurity. By enabling decentralized model training across distributed devices without transferring raw data, FL addresses critical privacy concerns while supporting real-time threat detection. This review synthesizes recent advances in the application of FL to cybersecurity, focusing on its deployment in domains such as the Internet of Things (IoT), intrusion detection systems, and threat intelligence networks. The review highlights key advantages, including enhanced data privacy, collaborative learning across diverse sources, and improved anomaly detection. It also identifies significant implementation challenges, such as communication overhead, computational inefficiencies, and susceptibility to adversarial attacks. Furthermore, the paper outlines major research gaps, including the lack of standardized benchmarks, limited real-world deployment studies, and the need for personalized federated models that address non-IID data and heterogeneous device capabilities. To bridge these gaps, the review proposes future research directions such as developing secure and efficient aggregation methods, prototyping FL systems in live environments, and advancing lightweight, adaptive FL frameworks. Overall, this review underscores the potential of FL to become a foundational technology in next-generation cybersecurity systems, enabling scalable and privacypreserving threat mitigation across distributed infrastructures