,
Northern Technical University , Mosul , Iraq
Northern Technical University , Mosul , Iraq
The rapid growth of the Internet of Things (IoT) brings new options to innovative healthcare, transportation, and industrial systems. However, this expansion also increases cyber threats to these infrastructures. Standard anomaly detection systems use fixed machine learning models. Such models require frequent retraining and are not very sensitive to concept drift which results in many false positives when used in adaptive IoT systems. In order to overcome these issues, this paper will present a bio-inspired, adaptive anomaly detector. It also presents a framework for selecting dynamic detectors via Artificial Immune Systems (AISs). The system architecture combines several immune-inspired concepts. Adverse selection separates normal from abnormal patterns, danger theory classifies anomalies in context, and clonal selection and mutation help detectors evolve. Immune memory supports long-term learning and quick response. The proposed model was tested on three benchmark IoT security datasets: UNSW-NB15, BoT-IoT, and TON_IoT. This allowed assessment against legacy and new attack scenarios. In the experiment, the approach achieved 97.5% accuracy, 96.9% precision, 97.8% recall, and a 97.3 F1-score. Compared to related 2023-2025 works, it performs 2.4-8.4% better across various measures. Detection latency decreased due to immune memory integration, and adaptation to zero-day attacks improved. These results confirm that AIS-based anomaly detection is a scalable and adaptive tool for securing future IoT environments.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
0
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.