The Upper Bounds for Delay and Cell Loss Probability of ATM Traffic in a Finite Capacity Polling System

This paper focuses on the upper bounds for both the mean delay and the probability of cell loss that bursty arrivals incur in a finite capacity multiqueue system with nonexhaustive cyclic service. We compute the upper bounds for this system by considering a cell multiplexer with the same arrival processes and equal queue capacity. Under the ATM environment, the mean delay obtained from this multiplexer cannot only serve as an upper bound but also render a fairly accurate estimation for the mean delay of the polling system. For the cell loss probability, we consider a multiple urn model with uniform occupancy distribution which will guarantee the upper bound. A heuristic method is proposed to give better estimates for cases which have medium to high cell loss rate.