Analyzing effects of SIP-messages batch arrival on queuing time for server processing

Development of multimedia services in modern telecommunication networks requires performance measures estimation of each network node. It is necessary to take into account the associated signaling traffic, such as signaling messages of Session Initiation Protocol (SIP), which today is one of the major signaling protocols in the next generation networks. The paper formulated an approach to analyze the dependence of probabilistic characteristics of SIP server mathematical model with batch arrival and vacations of the batch size distribution. Using non-parametric statistical methods we have proved the dependence of probabilistic characteristics of the distance between batch size distribution functions in the selected norm. The elasticity coefficients for probabilistic characteristics of the model for four distributions (Zipf, logarithmic, geometric and uniform) were found. The proposed approach provides additional factors to estimate the probability characteristics using a uniform batch size distribution.