A. Patel and C. L. Williamson. Proceedings of the IASTED Conference on Applied Modeling and Simulation (AMS'97), Banff, AB, pp. 298-302, July 1997.
Abstract: This paper studies the effective bandwidth required when multiple self-similar (i.e., long range dependent) traffic sources are multiplexed together in a network. In particular, an ATM network simulation tool is used to assess the accuracy of a theoretical effective bandwidth formula developed by Ilkka Norros for self-similar traffic. Synthetic traffic models with controllable workload parameters are used in the study, using a one-factor-at-a-time experimental design.
In general, the simulation results show good qualitative and quantitative agreement between the effective bandwidths predicted by the formula and by the ATM network simulator, provided that the number of multiplexed sources and the switch buffer size are not too small. Where the results are in good agreement, the formula tends to err on the "safe side", thus making the theoretical formula useful for call admission control and network dimensioning in an ATM network.
Keywords: Computer networks, telecommunication systems, e ective bandwidth, ATM