EXPERIMENTAL STUDIES OF INFORMATION SIGNS OF FRACTAL TRAFFIC

The problem of increasing the stability of critical information infrastructure systems under conditions of dynamic evolution of risk factors (as a rule, with statistical uncertainty) is based on the background of policies and management models. Managing the stability of critical network infrastructure is impossible without selecting and substantiating quantitative information characteristics of signal traffic. The task becomes non-trivial for heterogeneous signal traffic, which acquires self-similar (fractal) properties. The article considers information flows of signal traffic circulating in Triple Play and Quadruple Play wireless networks of critical application. Experimental studies of sufficient volume have been conducted, based on the results of which the most valuable information features of fractal traffic have been selected and substantiated. Indicators of information value are periodic and quasi-periodic intensity spikes inherent in fractal traffic with Hurst parameters that vary in a certain range of values. It is shown that quasi-periodic intensity spikes are sources of hidden periodicities, a comparative analysis of methods for detecting hidden periodicities is carried out, and statistical characteristics of the normalized range method (R/S analysis method) as one of the effective practical methods are provided.