The UltraSAN modeling environment
Performance Evaluation - Special issue: performance modeling tools
Stochastic Activity Networks: Structure, Behavior, and Application
International Workshop on Timed Petri Nets
Evaluating Dynamic Failure Probability for Streams with (m, k)-Firm Deadlines
IEEE Transactions on Computers
On the performance analysis of ABR in ATM LANs with stochastic petri nets
Journal of Systems Architecture: the EUROMICRO Journal
Consecutive customer losses in regular and oscillating MX/G/1/n systems
Queueing Systems: Theory and Applications
Stochastic model for cell movement in a knockout switch in computer networks
Journal of High Speed Networks
Hi-index | 0.00 |
Abstract: The proposed use of asynchronous transfer mode (ATM) in B-ISDN necessitates fast packet switches (FPS) capable of providing adequate quality of service (QoS). The knockout switch is an FPS with low delay and cell loss at the switch level. However, the performance perceived by a specific input is not simple to determine, since the often computed cell loss probability (CLP) is a time averaged value obtained with respect to the switch. An analysis of the distribution of consecutive cell losses seen at a tagged port would be more useful. In this paper, we examine this distribution under a wide range of traffic burstiness for the knockout switch using Markov processes generated automatically from a stochastic activity network (SAN) representation. The results provide useful information on the performance at a specific input, as well as illustrate the usefulness of SANs in modeling and analyzing telecommunication switch designs.