An analysis of Intrusion detection systems in IIoT

Abstract

Enhancing manufacturing and industrial processes are achieved using smart sensors and actuators known as the industrial internet or things or IIoT. By implementing IIoT, machines transform into smart devices, communicating important information that enables driving business decisions faster and more accurately. Thus, IIoT enables the improvement of business processes that are agile and dependable. However, the widespread deployment of IoT networks across different industrial sectors has resulted in numerous security gaps, posing threats to the integrity of IIoT systems. An intrusion detection system (IDS) is a typical software that monitors and stimulates security explanations for computer networks. Deployment of the solutions aims to identify malicious activity and take measures that facilitate risk aversion. Due to IIoT’s peculiarities, standard IDS-based solutions are complicated to implement. This includes limited resources, sensitive data, and heterogeneous architecture. Researchers are implementing Fog/Edge computing, Machine Learning (ML), and Deep Learning for effective and flexible IDS deployments in different IIoT scenarios. This research focuses on how IDS emerges in specific manufacturing settings. To do this, we present an in-depth analysis of IDS deployment strategies, detection strategies, evaluation approaches, and data sources.