GAN Neural Networks Architectures for Testing Process Control Industrial Network Against Cyber-Attacks

Protection of computer systems and networks against malicious attacks is particularly important in industrial networked control systems. A successful cyber-attack may cause significant economic losses or even destruction controlled processes. Therefore, it necessary to test the vulnerability process possible cyber-attacks. Three approaches employing Generative Adversarial Networks (GANs) generate fake Modbus frames have been proposed this work, tested for an network compared with classical approach known from literature. In first approach, one GAN generates byte a message frame. next two approaches, expert knowledge about frame structure used part frame, while remaining parts are generated using single multiple GANs. The single-GAN worst one. one-GAN-per-byte significantly more correct than method. Moreover, all i.e., selected bytes GANs methods. Finally, we describe effect cyber-attacks on operation process.