Securing IoT Communications via Anomaly Traffic Detection: Synergy of Genetic Algorithm and Ensemble Method

Area of Science:

• Computer Science
• Cybersecurity
• Machine Learning

Background:

• Internet of Things (IoT) systems face significant security challenges due to decentralized architectures and resource-constrained devices.
• Anomalous network behaviors and data manipulation threaten IoT security and reliability.
• Machine learning methods are increasingly used for intrusion detection and prevention in IoT.

Purpose of the Study:

• To propose an advanced, multi-phase anomaly detection framework for IoT networks.
• To enhance the security and reliability of IoT ecosystems against cyber threats.
• To develop a scalable and adaptable solution for diverse IoT scenarios.

Main Methods:

• Data preprocessing using Median-KS Test for noise reduction and data balancing.
• Optimal feature selection via a Genetic Algorithm with eagle-inspired search strategies.
• Ensemble classifier combining Decision Tree, Random Forest, and XGBoost algorithms.

Main Results:

• Achieved 98% accuracy, a 12.5% improvement over existing methods.
• Increased detection rate to 95% (14% improvement).
• Reduced false positive rate to 10% (9.3% reduction) and false negative rate to 5% (10.8% reduction).

Conclusions:

• The proposed framework demonstrates superior effectiveness, reliability, and scalability for securing IoT networks.
• The multi-step methodology ensures adaptability in handling diverse and evolving cyber threats.
• The results underscore the framework's potential for real-world IoT security applications.