Blockchain based Secure Data Sharing for Glucose Prediction in IoMT Scenario using Deep Learning
Abstract
The growth of Internet of Medical Things (IoMT) systems makes the continuous health monitoring as the feasible task. Still, the challenges related to data security, privacy and requirement of efficient predictive outcome degrades the performance. To address the challenges, this research proposed the glucose prediction with blockchain technology based on cryptographic hashing and optimization-based key generation along with federated learning-based prediction. The proposed model utilized the Blockchain-based Data Acquisition and Storage mechanism integrated with the Optimal Bhattacharya Signed Tiger Hashing Algorithm (O-BHA) to ensure tamper-proof and secure data storage. In the O-BHA algorithm, the optimized key generation mechanism is employed using the Sinusoidal Beaver Behavior Optimization (SBbO) algorithm to improve global search capability and convergence. Finally, the Kolmogorov-Arnold Networks-based Federated Learning (KAN-FL) model is proposed for providing the decentralized training by preserving data privacy. The experimental evaluations demonstrate that the consideration of secure blockchain storage with federated learning based prediction improves the performance by ensuring data protection and system scalability.