Characterization of the traffic on high-speed token-ring networks

With application toward modeling network interconnection and client/server computing environments, we present a characterization of the traffic on high-speed token-ring networks such as FDDI. The traffic passing by an arbitrary point on the token-ring network is modeled through a Markov packet train characterization. The transition probabilities for the size of successive packet trains are derived through use of the continuous polling model. A simulation study of a bridge for LAN/WAN interconnection is presented in order to investigate the accuracy of our traffic characterization. We find that our traffic characterization, when used to represent the output traffic from a token-ring network, provides significantly more accurate results for the mean packet waiting time at the bridge than when modeling the output traffic by a Poisson process. © 1994.