Intelligent Biometric Authentication-Based Intrusion Detection in Medical Cyber Physical System Using Deep Learning

The current generation of technology is evolving at a rapid speed and gaining a prominent position in the hearts of individuals. For instance, when internet-connected gadgets link to other devices, they create a large system that solves complicated issues and makes people's lives simpler and longer. The Cyber Physical System (CPS) is a very essential advancement in technology. The rapid and important growth of CPS influences several facts of people's lifestyles and allows a more comprehensive selection of services and applications, including smart homes, e-Health, smart transport, and e-Commerce, etc. In the advanced medical field, a medical cyber-physical system (MCPS) is a one-of-a-kind cyber-physical system that integrates networking capability, embedded software control devices and the complicated health records of patients. Medical cyber-physical data are digitally produced, electronically saved, and remotely accessible by medical personnel or patients through the process of MCPS's interaction between communication, devices, and information systems. MCPS is based on the concept that biometric readings can be used as a way to verify a user's identity in order to protect their security and privacy. Several studies have revealed that Machine Learning (ML) algorithms for CPS technology have achieved significant advancements. Interactions between real-time physical systems and dynamic surroundings have been significantly simplified by the use of more effective ML techniques in CPS. In this study, we have suggested a convolutional neural network (CNN)-based intrusion detection system for identifying anomalies in MCPS. The ECU-IoHT dataset has been used for our research. The experimental findings outperform the conventional ML baseline models, demonstrating the efficacy of our proposed approach.