Self-Similarity in a multi-stage queueing ATM switch fabric

Recent studies of digital network traffic have shown that arrival processes in such an environment are more accurately modeled as a statistically self-similar process, rather than as a Poisson-based one. We present a simulation of a combination sharedoutput queueing ATM switch fabric, sourced by two models of self-similar input. The effect of self-similarity on the average queue length and cell loss probability for this multi-stage queue is examined for varying load, buffer size, and internal speedup. The results using two self-similar input models, Pareto-distributed interarrival times and a Poisson-Zeta ON-OFF model, are compared with each other and with results using Poisson interarrival times and an ON-OFF bursty traffic source with Ge ometrically distributed burst lengths. The results show that at a high utilization and at a high degree of self-similarity, switch performance improves slowly with increasing buffer size and speedup, as compared to the improvement using Poisson-based traffic