Deep Learning and Blockchain-Enabled Healthcare Data Protection for Industrial Cyber-Physical Systems: A Hybrid Security Framework

The fast adoption of Industrial Cyber-Physical Systems (ICPS), Internet of Medical Things (IoMT), cloud computing, and telemedicine technologies has completely changed the way modern healthcare delivery systems work. But at the same time, the growing level of connectivity of medical devices and healthcare infrastructure increases the attack surface of the healthcare sector and makes it more vulnerable to ransomware, data breaches, unauthorized access and insider attacks. Centralized security approaches can hardly protect the system from new emerging cyber threats and maintain data integrity and privacy at the same time. Although Deep Learning algorithms have already shown great potential for intrusion detection and anomaly recognition, they do not provide any guarantees concerning safe data sharing and tampering-proof storage. In turn, although Blockchain provides decentralization, immutability and transparency, it does not have intelligent threat detection capabilities. In order to resolve these issues, this paper suggests using a hybrid framework based on Deep Learning and Blockchain for healthcare Industrial Cyber-Physical Systems. The architecture of the suggested framework incorporates Convolutional Neural Networks (CNN), LSTM neural networks, Autoencoders, Smart Contracts, Distributed Ledgers and access control via Blockchain.