ARTIFICIAL INTELLIGENCE BASED BYZANTINE FAULT TOLERANCE CONSENSUS ALGORITHM FOR HEALTH CARE MONITORING SYSTEM

Abstract

The rapid advancement of Internet of Things (IoT) and distributed computing has created unprecedented opportunities for real-time healthcare monitoring systems. However, these distributed systems face significant challenges related to fault tolerance, data integrity, and Byzantine failures — conditions where nodes may behave arbitrarily or maliciously, compromising the reliability of critical patient health data. This paper proposes a novel Artificial Intelligence Based Byzantine Fault Tolerance (AI-BFT) consensus algorithm specifically designed for healthcare monitoring systems. The proposed framework integrates machine learning-driven anomaly detection with a modified Practical Byzantine Fault Tolerance (PBFT) protocol to identify and isolate faulty or malicious nodes in real time. The AI component employs a Long Short-Term Memory (LSTM) neural network to model normal sensor behavior and detect deviations indicative of Byzantine faults. Experimental evaluations conducted on a simulated IoT-based healthcare network demonstrate that the AI-BFT algorithm achieves a fault detection accuracy of 97.4%, reduces consensus latency by 34% compared to classical PBFT, and maintains system availability above 99.2% under Byzantine fault conditions involving up to 33% faulty nodes. The proposed system provides a robust, scalable, and intelligent solution for ensuring data integrity and reliability in life-critical healthcare monitoring environments.