Towards a secure Metaverse: Leveraging hybrid model for IoT anomaly detection.

Area of Science:

• Cybersecurity
• Artificial Intelligence
• Computer Networking

Background:

• The integration of the Internet of Things (IoT) and the metaverse blurs virtual and real worlds, necessitating robust security measures.
• Ethical, privacy, and security concerns arise from interconnected IoT systems within the metaverse.
• Anomaly detection is crucial for identifying and preventing malicious activities in complex, dynamic IoT networks.

Purpose of the Study:

• To propose and evaluate a hybrid machine learning model for anomaly detection in IoT-enabled metaverse environments.
• To compare the performance of the proposed hybrid model against various traditional machine learning techniques.
• To enhance the security and trustworthiness of connected devices in the metaverse.

Main Methods:

• A hybrid model integrating Random Forest (RF) and Neural Network (NN) was developed.
• The hybrid model and several other machine learning algorithms (Decision Tree, Naive Bayes, K-Nearest Neighbor, RF, Logistic Regression) were trained and tested.
• The CIC-IDS 2017 Network Intrusion Dataset was utilized for model evaluation.

Main Results:

• The proposed hybrid RF-NN model demonstrated superior performance compared to other machine learning models.
• The hybrid model achieved an exceptional accuracy rate of 99.99% in anomaly detection.
• The study highlighted the model's effectiveness in high-accuracy anomaly identification and minimizing false positives.

Conclusions:

• The hybrid RF-NN model offers a highly effective solution for anomaly detection in IoT-enabled metaverse systems.
• This approach significantly enhances the security and dependability of connected devices.
• The findings underscore the potential of advanced machine learning for safeguarding metaverse environments.