Secure Fog Computing for Remote Health Monitoring with Data Prioritisation and AI-Based Anomaly Detection.

Area of Science:

• Cyber-physical systems
• Internet of Things (IoT)
• Healthcare technology

Background:

• Cloud architectures present latency, bandwidth, and privacy issues for smart health monitoring.
• Fog architectures offer proximity but face data management, security, and privacy challenges for sensitive patient data.

Purpose of the Study:

• To address limitations in current remote health monitoring systems.
• To propose an innovative fog-based framework integrating secure communication and intelligent data prioritization (IDP).
• To enhance anomaly and threat detection for real-time patient care.

Main Methods:

• Developed a fog-based framework with secure communication and IDP.
• Integrated an AI-based enhanced Random Forest model for anomaly and threat detection.
• Utilized a simulated smart healthcare scenario with synthesized data from wearable devices.

Main Results:

• Dynamically prioritized health features (heart rate, SpO2, breathing rate) using AI and rule-based thresholds.
• Achieved high predictive performance: 93.5% accuracy, 90.8% precision, 88.7% recall, and 89.7% F1-score.
• Demonstrated a successful proof-of-concept for real-time secure transmission of critical health anomalies.

Conclusions:

• The proposed fog-based framework effectively enhances secure remote health monitoring.
• Intelligent data prioritization and AI-driven anomaly detection improve real-time intervention capabilities.
• The framework supports clinicians with timely, accurate data for personalized patient care.