Dobrovolskyy Y. IContext methods usage for modeling of a packet-network traffic

The object is modeling of a packet-network traffic. The goal is improvement of a network traffic modeling methods, in particular, modeling of the traffic generated by network applications and services. Used methods include general principles of imitating modeling, context analysis, theory of Markov processes, queueing theory, probability theory and statistics. The novelty – based on a general idea of context modeling techniques, the work is the first to use context dependencies in measured sequences of network traffic for its imitating modeling. Results - the universal dynamic imitating model was build for implementation of the aforesaid paradigm. The model has a self-learning capability implemented by parsing of application’s traffic log files, captured in a network. Applied – in the educational process and research efforts of the Odessa national academy of telecommunications implemented for joint stock companies “Ukrtelecom” and “Telecominvest”; Department of telecommunications of the Ministry of Internal Affairs of Ukraine. Branch – packet networks, telecommunications, higher education