Babich Y. Functional improvement to monitoring of dynamic characteristics of information-communication networks

Object of research - process of monitoring for an information and communication network. Subject of research - methods and technologies of monitoring the parameters of dynamic characteristics of information and communication network with the ability to predict their values. Research methods - statistical forecasting methods, probability theory, method of multiple regressions, simulation. Theoretical and practical results: the concept of predictive monitoring with the ability to make short-term, situation, and long-term forecasts for parameters describing dynamic characteristics of an information and communication network was formulated for the first time; it provides possibility to forest the moments when emergency situation may occur as well as the deadline moments for performing reconfiguration or reconstruction of an object under monitoring, which improves functional capabilities of monitoring and widens the application area of the TMN and TINA concepts; the proposed concept was used as a basis for developing a method of predictive monitoring formalized by means of a cognitive algorithm that implies the procedures of predictive monitoring for parameters describing dynamic characteristics of an information and communication network and allows to determine right moments to reconfigure network resources or to perform reconstruction; the algorithm includes statistical forecasting methods and, in contrast to the known approaches, doesn't require learning and relearning during operation; the architecture of an information and communication network management system was improved by adding a cognitive analysis subsystem capable of identifying, classifying and forming messages about regularities in monitored data; the messages can be used as a criterion for preventive maintenance of convergent network equipment; for the first was set and solved the problem of implementing the predictive monitoring to states of a convergent mobile aggregation/distribution using integral parameter of frame alignment loss probability in the scenario of structure agnostic Circuit Emulation Service over Ethernet for TDM-flow transportation through asynchronous network segment, which required formation of the reconfiguration rule for CESoETH equipment that considered predictive monitoring results as well as the new mathematical expressions for probabilities of frame alignment loss and recovery along with the coefficient of emulated channel throughput beneficial utilization as a function of TDM bytes encapsulated into CES-frame; the proposed approach allowed to increase the average value of the mentioned coefficient by 10,49%; forecasts quality estimation over 6368 forecasts has shown 97% of correct forecast.