Securing the Industrial IoT: A Novel Network Intrusion Detection Models
Mangayarkarasi Ramaiah, Mohemmed Yousuf Rahamathulla
Abstract
The growth of automation in the industry makes securing IoT networks a critical priority. A resilient network intrusion detection system (NIDS) can reduce cyber threats. Detecting network traffic irregularities with deep learning (DL) and machine learning (ML) techniques can improve IoT network security and reduce cyber threats. This study analyzes the merits of LSTM and other ML models in mitigating anticipated security breaches in Industrial IoT (IIoT) networks. To design robust NIDS, including an appropriate dataset is essential. Therefore, to identify dynamic threats targeting Industrial Internet of Things (IIoT) networks, deep learning (DL) and machine learning (ML) models underwent training and evaluation utilizing the EdgeIIoT-2021 dataset. The experimental findings demonstrate that the proposed approach surpasses the performance of the current state-of-the-art Network Intrusion Detection Systems (NIDS). The ERT-based IIoT-NIDS achieved a cyber-attack detection accuracy of 99.93%, while the LSTM-based IIoT-NIDS achieved an accuracy of 99.85%.