T. V. JAMGHARYAN, Lieutenant Colonel, Senior Officer, Signal and AMS Department, the RA Armed Forces

SUMMARY

One of the indicators of the fourth industrial revolution is the “digital block”, which presents a bridge between the physical and digital realities. The entire digital reality is a series of devices connected with data networks. In this regard, threats to the network infrastructure were investigated, in particular, taking into account some peculiarities of network-centric wars. Software and hardware solutions, based on different principles and operation algorithms, are used in order to neutralize various threats to the network infrastructure. Tasks of both general and particular type, arising in the study of intrusion detection systems (IDS) and operating via machine learning were identified. In addition, the article formulated those tasks, the comprehensive solution of which will ensure the minimization of the “attack surface” on network infrastructures.

As a discrete object of study, the article discusses research problems of measuring the effectiveness (efficiency) of applying machine learning in intrusion detection systems for both mobile and embedded systems. The article dwells on the enhancements made to the research work on the application of machine learning in IDS. The solution of such problems will increase the degree of stability of the network infrastructure in case a possible or probable network-centric conflict occurs.