Batch poissonian arrival models of multiservice network traffic

The emergence of packet-switched communication networks has made it clear that Poissonian arrival flow models are not quite adequate and required the development of new models based on non-Poisson distributions. This paper is devoted to the analysis of a particular case of a batch Markovian flow, namely, batch (nonordinary) Poissonian arrivals. Such a flow is stationary and memoryless but not ordinary. We consider a class of queueing systems with constant service time. We present results of analytical computations of arrival flow parameters and also simulation results. We show that the variance of the queue depends on the third moment of the batch size in a batch Poissonian arrival flow.