A Hybrid Intrusion Detection Framework for Hypervisor-Based MitM Attack Detection in Medical Cyber-Physical Systems: Leveraging PCA, Anomaly Detection, and KNN

Cyber-Physical Systems for Medical devices revolutionized healthcare by integrating medical devices with intelligent cloud server and embedded systems. Recent development in virtualization within these systems invites threats to critical components like VM, Hypervisors. Threats like DDoS, MitM affect the systems and its sensitive data. This study proposes a lightweight machine learning-based detection model for MitM attack detection, which combines principal component analysis, k-nearest neighbors. The framework handles data imbalance through random under sampling and applies scaling to manage noise. Result shows strong accuracy and efficiency using an Edge-IIotset dataset. The approach offers a practical, interpretable solution for securing virtualized healthcare systems against evolving threats.