A survey on Hybrid learning-based Anomaly Detection in IoT networks
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
IoT networks face increasing risks from zero-day attacks, scalability issues, and limited resources.These challenges reduce the effectiveness of traditional intrusion detection systems.Predictive modeling has come forward as a promising approach to address these problems.Sixteen recent studies published between 2021 and 2025 have been reviewed and grouped into five categories: deep learningbased intrusion detection, hybrid predictive models, blockchain-enabled frameworks, federated and edge learning, and semantic or active learning methods.Each category is analyzed based on methodology, datasets, detection performance, and limitations.The survey points out significant improvements in detection accuracy, flexibility, and transparency while also identifying ongoing challenges like energy efficiency, resilience against new threats, and large-scale deployment.A structured taxonomy and comparative analysis bring together these diverse efforts.The findings provide a valuable reference for researchers and practitioners aiming to build secure, scalable, and smart IoT anomaly detection systems.
Fields of Study & Subjects
Network Security and Intrusion Detection
Internet of Things and AI
Artificial Immune Systems Applications
Publication details
- Journal / source
- Advances in intelligent systems research/Advances in Intelligent Systems Research
- Pages
- 1320–1335
- Publication date
- 1 April 2026
- Language
- EN
- Document type
- Book chapter