Network Intrusion Detection in Industrial Healthcare Environment using Ensemble Deep ...

Network Intrusion Detection in Industrial Healthcare Environment using Ensemble Deep Neural Decision Tree

The emerging industrial healthcare system leverages the concepts of Industry 4.0 to provide better services to patients. An integrated security tool is needed in an indus...

The emerging industrial healthcare system leverages the concepts of Industry 4.0 to provide better services to patients. An integrated security tool is needed in an industrial healthcare environment to ensure the integrity of network systems. The intrusion detection system (IDS) is a security tool that uses network traffic and payload statistics in the industrial healthcare environment to protect connected systems, including various devices, sensors, and networks, from cyberattacks. This work presents an ensemble deep neural decision tree (EDND)-based approach for intrusion detection in an industrial healthcare environment (IDIHE). The performance of IDIHE is evaluated using the network flows of the TCP/IP protocol and the pay-load flows of the IEC 60870-5-104 protocol. Experiments using statistics of network flows and payload flows show that the EDND showed better performance compared to the deep neural decision tree and the decision tree. In addition, EDND showed better performance using the payload flows of the IEC 60870-5-104 protocol compared to the network flows of the TCP/IP protocol. Using payload flows from the IEC 60870-5-104 protocol, EDND showed a performance improvement of 9.37% Accuracy and 9.66% F1-score compared to the existing methodology. The proposed methodology can be used by a network administrator to monitor and analyze network traffic in an industrial healthcare environment.

2026 14th International Conference on Intelligent Control and Information Processing (ICICIP)

The rapid advancement of Internet of Things (IoT)-enabled industrial technology and its integration into the healthcare environment has provided several valuable benefits to the delivery of services to the patients and healthcare operations [1]. An industrial healthcare environment is composed of various medical devices, IoT sensors, and networks. Devices and sensors communicate with others using various protocols using internet services in an interconnected industrial healthcare environment. Most commonly used protocol in an industrial healthcare environment is IEC 60870-5-104. The sectors’IEC 60870-5-104 protocol is insecure, but it is widely used in industrial critical systems, including healthcare 4.0. Rapid development and adoption of the Industrial healthcare environment with heterogeneous devices bring several vulnerabilities. An insecure communication property of the IEC 60870-5-104 protocol and other security vulnerabilities attract the cyber attackers to target industrial healthcare systems [1]. Recent literature shows that the number of attacks on the industrial healthcare environment is increasing over the years compared to the non-industrial sectors automated systems [2]. Cyber attacks on industrial healthcare systems have immediate consequences on patient care.

References is not available for this document.