One-Periodic Template Marks Model of Normal Behavior of the Safety Parameters of Information Systems Networking Resources

The article is devoted to improvement cyber-attacks recognition systems on network resources of information systems. By analogy with the classification of technical systems failures, it is proposed to select a class of step-by-step cyber-attacks that are characterized by a long and expected, from the point of view of the information security system, process of changing security parameters to values that exceed the safe limits. The approach to the recognition of progressive cyber-attacks is grounded. It involves estimating the deviation of the current values of the safety parameters from the patterns of normal behaviour, calculated over a given operation interval. The need for improving the existing patterns by adapting them to the typical non-stationary nature of the dynamics of security parameters is shown. The Markov model of a one-period model of normal behaviour was developed and verified, which allows the approximation of dynamics safety parameters in case of its one-time non-stationary. The expediency of carrying out of further researches in the direction of development of a multi-periodic template Markov model of safety parameters normal behaviour is shown.