PORT SCANNING MALWARE DETECTION IN IOT NETWORKS USING AN AI-BASED IDS FRAMEWORK
The fast growth of the Internet of Things (IoT) has greatly exposed the attack surface, making the IoT networks vulnerable to new malware attacks, especially port scanning malware. These advanced threats have been known to go beyond the standard security solutions like firewalls and signature-based Intrusion Detection Systems (IDS). In this work, an AI-Based IDS framework is proposed, which will utilize Machine Learning (ML), Deep Learning (DL), and Reinforcement Learning (RL) to identify and predict the presence of port scanning malware in IoT networks. The framework is tested with a publicly available set of data on IoT security by both balanced and imbalanced conditions, which guarantees a full performance test. In experiments, it is proven that DL models are more successful than other methods with precision of 98.5, recall of 99.2, F1-score of 98.8 and accuracy of 98.7. The ML traditional models are also very precise and recalls, but the DL models have overall better performance which detects port scanning malware effectively. The results support the potential of the DL-based solutions as the effective measure of boosting the security of IoT in response to the new cyber threats
Authors
- Mina Malekzadeh (ORCID: https://orcid.org/0000-0002-0733-3970)
- Mahmood Al-Okashi
Publication Details
- Journal
- Kufa Journal of Engineering
- Published
- 2026-11-14
- DOI
- https://doi.org/10.30572/2018/kje/170416
- Primary Topic
- Network Security and Intrusion Detection
- Type
- article
- Field-Weighted Citation Impact
- 0.00