Bazan P, Bolch G, German R (2006)
Publication Type: Conference contribution
Publication year: 2006
Publisher: VDE Verlag GmbH
Edited Volumes: 2006 13th GI/ITG Conference on Measuring, Modelling and Evaluation of Computer and Communication Systems, MMB 2006
City/Town: Berlin
Pages Range: 465-468
Conference Proceedings Title: Proc.13th GI/ITG Conference
Event location: Nürnberg, Germany
ISBN: 3-8007-2945-8
WinPEPSY-QNS is a tool for the easy construction of queueing network models and the evaluation of the performance measures. A graphical and a textual editor can be used for the construction process. The performance measures of the models can be derived by several analysis methods. For product-form networks the mean value analysis (MVA), the method of Jackson, and the method for BCMP networks are implemented. For non-product-form networks the methods of Marie or Chylla can be used. The state space based analysis is provided by a discrete event simulation and solvers for continuous time Markov chains. Some important non-Markovian distributions can be approximated by phase-type distributions. For large Markov chains an efficient solver has been implemented, based on multi-valued decision diagrams and a new approximation method for the stationary analysis.
APA:
Bazan, P., Bolch, G., & German, R. (2006). WinPEPSY-QNS-Performance Evaluation and Prediction System for Queueing Networks. In Proc.13th GI/ITG Conference (pp. 465-468). Nürnberg, Germany, DE: Berlin: VDE Verlag GmbH.
MLA:
Bazan, Peter, Gunter Bolch, and Reinhard German. "WinPEPSY-QNS-Performance Evaluation and Prediction System for Queueing Networks." Proceedings of the Measuring, Modelling and Evaluation of Computer and Communication Systems,, Nürnberg, Germany Berlin: VDE Verlag GmbH, 2006. 465-468.
BibTeX: Download