AN ENSEMBLE COGNITIVE LEARNING-BASED INTRUSION DETECTION SYSTEM FOR SECURE IOT ENVIRONMENTS

DOI: https://doie.org/10.0113/Jbse.2025967844

Kushal Kumar B.N., Dr. Balakrishna R., Dr. M.V. Panduranga Rao

Keywords:

Internet of Things Intrusion detection, Ensemble learning, Sustainability

Abstract:

The Internet of Things (IoT) has significantly transformed various sectors, including smart homes, wearable technology, network applications, and autonomous systems. Despite its advantages, IoT networks face substantial challenges in detecting abnormal traffic patterns, which are crucial for safeguarding network infrastructure from cyber threats. Intrusions that exploit system vulnerabilities can enable attackers to use malicious traffic to gain unauthorized access. Attacks like Distributed Denial of Service (DDoS), Denial of Service (DoS), and Service Scans highlight the necessity of an automated system that can quickly identify anomalies and minimize damage. Although many automated techniques for abnormal traffic detection exist, current Intrusion Detection Systems (IDS) require improvements in efficiency, scalability, and flexibility to handle diverse IoT network environments effectively. This research focuses on addressing these gaps by introducing an Ensemble Cognitive Learning-Based Intrusion Detection System. The proposed system utilizes the Edge-IoT dataset as a benchmark to evaluate machine learning models for detecting network anomalies. A cognitive engine incorporating a meta-classifier dynamically selects the best-performing model for the ensemble classifier, ensuring precise detection of attacks. Experimental findings indicate that the proposed model surpasses existing methods, offering superior multi-class classification accuracy and robust detection capabilities.

PDF