Enhancing Network Intrusion Detection Using an Ensemble Voting Classifier for Internet of Things.

Area of Science:

• Cybersecurity
• Machine Learning
• Network Security

Background:

• The Internet of Everything (IoE) in 6G networks expands connectivity, increasing security risks from botnets and other attacks.
• Protecting IoT-enabled metaverse connections requires robust security measures to detect network anomalies.

Purpose of the Study:

• To propose a novel classification technique for enhanced network intrusion detection.
• To improve the accuracy and precision of security systems in the context of rapidly expanding IoT connectivity.

Main Methods:

• Developed a DRX ensemble voting classifier combining Decision Tree, Random Forest, and XGBoost algorithms.
• Evaluated the proposed technique using benchmark datasets: NSL-KDD, UNSW-NB15, and CIC-IDS2017.

Main Results:

• Achieved high accuracy rates: 99.88% (NSL-KDD), 99.93% (UNSW-NB15), and 99.98% (CIC-IDS2017).
• Significantly reduced false positive rates to 0.003%, 0.001%, and 0.00012% across the datasets.
• Demonstrated superior performance compared to other existing methods.

Conclusions:

• The DRX-based ensemble voting classifier is highly effective for network intrusion detection in IoE environments.
• The proposed method offers a robust solution for safeguarding 6G networks against emerging cyber threats.