Computer Simulation of Chaotic Phenomena in High-Speed Communication Networks

Chaos and self-similarity are the state-of-the-art problems in various areas of modern science and technology, thus network systems are not an exception from this trend. An increasing number of various network protocols, applications and services lead to more complex and unpredictable network traffic. The phenomenon of self-similarity is caused by the properties of network traffic whose origin is the behavior of TCP protocol. All these properties become more significant with the appearance of high-speed data transmission technologies. This behavior results into a congestion in network and packet losses. However, even modern congestion control mechanisms handle such kind of situations quite unfairly. As shown by W. feng et al., TCP Reno loss rate exceeds 5% in a heavily congested network so it"s easy to calculate that over a Gigabit Ethernet link, such a loss rate being translated into a loss of over 50 Mb/s. Obviously this level of loss rate is unacceptable. Models that describe the behavior of information systems sufficiently, giving an avenue for scientists to apply all sets of classical methods of chaos theory and analyze nonlinear dynamic systems, have not been offered so far. Phase portraits of the studied system are built and Lyapunov exponents for different values of the basic system parameters are calculated in this paper while a new approach in the analysis of packet switching network behavior involving TCP protocol is proposed. Such a network is analyzed as a nonlinear dynamic system that shows chaotic properties within a certain value of parameters.