The analysis of a multiserver queue fed by a discrete autoregressive process of order 1

We analyze a multiserver queue with a discrete autoregressive process of order 1 (DAR(1)) as an input. DAR(1) is a good mathematical model for VBR-coded teleconference traffic. Based on matrix analytic methods and the theory of Markov regenerative processes, we obtain the stationary distributions of the system size and the waiting time of an arbitrary packet. Numerical examples illustrate the quantitative effect of the stationary distribution and the autocorrelation function of input traffics on system performance. keywords DAR(1), multiserver queue, Teleconference traffic, multiserver ATM multiplexer, matrix analytic methods, Markov regenerative theory, performance analysis This work was supported by Korea Research Foundation Grant (KRF-2001-042-D00009). 1