WSEAS Transactions on Communications
Print ISSN: 1109-2742, E-ISSN: 2224-2864
Volume 13, 2014
Recognition and Modelling of Bursty Period of Flow
Authors: ,
Abstract: This article presents some methods for recognition and modelling of bursty period of event flow. The flow of events represents generally 0/1 bit sequences. We used the Markov modulated Regular process (MMRP) and Bernoulli process (MMBP) for the description of flows with bursty period. We derived probability distribution of spaces of MMRP processes and used it for estimation of unknown parameters of process. Because the uncomfortableness of flows with MMP processes, we used a Effective bandwidth, to gain parameters of MMBP and obtain a process which will over design considered flow.