NEURO-FUZZY SYSTEM FOR DETECTING INTRUSIONS INTO THE INFORMATION NETWORK OF CRITICAL INFRASTRUCTURE

Abstract

 In the situation of Russia’s military aggression against Ukraine, the safety of people and the country largely depends on the reliability of critical infrastructure. In addition to physical attacks with weapons, Russia uses cyber weapons to attack the management systems of these facilities through cyberspace. Particularly alarming is the tendency for such facilities, which use modern technologies and operate in a single information environment, to remain vulnerable to new types of cyber threats, even with great efforts to protect them. This significantly complicates the task of ensuring long-term sustainability and security. Protecting information systems at such facilities is critical for the stable development of modern society. This article considers the task of detecting intrusions into critical infrastructure information networks. The main components of an intrusion detection system are identified and their functions are described. The article analyzes various approaches to detecting information security violations. The main methods of intrusion detection are characterized, their advantages and disadvantages are highlighted. It is shown that in order to increase the efficiency of detecting situations related to possible intrusion, it is necessary to use modern technologies of data mining. The features of technologies for use in intrusion detection systems were investigated, and based on the results of their comparative analysis, hybrid tools for detecting attacks were proposed. It is shown that the most promising for the task under consideration is the use of neuro-fuzzy methods. The architecture of a neuro-fuzzy system for detecting intrusions into the information network of critical infrastructure is proposed.