IDS PROTOTYPE FOR INTRUSION DETECTION WITH MACHINE LEARNING MODELS IN IOT SYSTEMS OF THE INDUSTRY 4.0

Abstract

Industry 4.0 significantly improves productivity by collecting and analyzing data in real time. This, combined with remote access functions, and cloud processing that allows Internet of Things IoT, provides information that optimizes processes and decision support. Also involves a great growth of new networks and systems with special features, which mean that they are vulnerable to different attacks. So new security requirements are emerging in the IoT network. To improve the security of an IoT system for a transparent way, it is proposed the development of a prototype intrusion detection system IDS, which detects anomalies in IoT environments using the MQTT protocol (Message Queuing Telemetry Transport), widely used in IoT systems. For this purpose, it is generated a dataset of an IoT system in which perform different attacks on the MQTT protocol. This dataset is used to train a machine learning model, which is implemented in the IDS that captures the network frames in real time from the system to classify and detect the different attacks. Keywords: IoT, industry 4.0, cybersecurity, IDS, MQTT protocol, Machine Learning.