Detection of Botnet Attacks against Industrial IoT Systems by Multilayer Deep Learning Approaches

Abstract

Industry 4.0 is the next revolution in manufacturing technology that is going to change the production and distribution of goods and services within the following decade. Powered by different enabling technologies that are also being developed simultaneously, it has the potential to create radical changes in our societies such as by giving rise to highly-integrated smart cities. The Industrial Internet of Things (IIoT) is one of the main areas of development for Industry 4.0. These IIoT devices are used in mission-critical sectors such as the manufacturing industry, power generation, and healthcare management. However, smart factories and cities can only function when threats to cyber security, data privacy, and information integrity are properly managed. In this regard, securing IIoT devices and their networks is vital to preserving data and privacy. The use of artificial intelligence is an enabler for more secure IIoT systems. In this study, we propose high-performing deep learning models for the classification of botnet attacks that commonly affect IIoT devices and networks. Evaluation of results shows that deep learning models such as the artificial neural network (ANN), the long short-term memory (LSTM), and the gated recurrent unit (GRU) can successfully be used for classifications of IIoT malware attacks with an accuracy of up to 99%.