Soft Computing Optimization of Stealth Data Loss Attack to Industrial Control Systems

The panorama brought by Industry 4.0 demonstrates the advantages of using communication networks in the control of physical systems, ranging from manufacturing processes to critical infrastructures. However, this trend also exposes the controlled physical systems to cyber threats. Recent literature presents a class of stealth attack to Industrial Control Systems (ICS) based on Real-Time Ethernet, where the attacker is able to affect the functioning of a plant by causing the selective loss of few frames on the ICS network. In this paper we demonstrate that the referred attack can be optimized in order to produce the same impact to the plant, but causing the selective loss of an even smaller number of frames in the network. The results show that the attack, which is based on Soft Computing tools, presents the same accuracy as its original version, however causing 13.8% less loss of frames on the network. The results show that the proposed approach increases the efficiency of the attack due to the lower frame loss it causes, which reduces the chances of it being perceived.