Nonparametric and simulated analysis of intensity for a queueing system (Q865583)

In this paper, the statistical inference of parameters in queueing systems with distribution-free interarrival and service times is proposed. The authors introduced a consistent and asymptotically normal (CAN) estimator of the intensity for three queueing systems, namely: M/E\(_{4}\)/1, E\(_{4}\)/M/1, E\(_{4}\)/H\(_{4}^{po}\)/1. Using this estimator and its associated estimated variance, an asymptotic confidence interval of intensity is obtained. Moreover, a rule for testing a statistical hypothesis of intensity is derived. To demonstrate the performance of the proposed estimator and its associated estimated variance applied to statistical inference of the intensity for all studied queueing systems, a numerical simulation was presented. In summary, using the statistical inference technique proposed in this paper, the CAN estimator of real queueing systems was obtained.