Attack and Anomaly Detection in IIoT Networks Using Machine Learning Techniques
Pradeep Kumar, Indrajit Banerjee
Abstract
Sensor-driven edge devices, known as things, connect with control systems like intelligent machines and analytic applications, making a network known as the Industrial Internet of Things (Industry 4.0) work synchronously and communicate. The massive amount of data is generated by edge devices, making them vulnerable to attack if an intrusion detection system (IDS) is implemented in the cloud. So an intermediate layer is required to provide seamless and uninterrupted IoT cloud services to edge (end-user) devices. The fog layer technique is a collection of servers distributed across the zones where edge devices (stationary or mobile) reside. The machine learning methods suggested in this study operate on real-time traffic data produced by the network. The algorithms are thus previously trained on a well-known NSL KDD, an improved dataset of the formerly known KDD Cup ‘99 dataset. The algorithms proposed in this paper have achieved up to 99.8% accuracy in the model. This work is showing better than the existing algorithm.