Modeling and Clustering Analysis of Broadband Convergence Networks

In contemporary telecommunication systems Markov processes are seldom observed, and the widely used Markovian models don't represent precisely the real system. In order to omit the need of modeling the system with a Markov chain we apply different clustering approaches for obtaining the steady state probabilities, which are represented by the data clusters. Some widely used data clustering methods are applied for performance evaluation of different telecommunication networks. However, in order to accomplish our investigation, we conduct our research with Markovian models, so that we have a solid ground for comparison, although the benefits of applying clustering techniques lie in the domain of the non-Markovian processes.