SMART: simulation and Markovian analyzer for reliability and timing

Author

Ciardo, Gianfranco ; Miner, Andrew S.

Author_Institution

Dept. of Comput. Sci., Coll. of William & Mary, Williamsburg, VA, USA

fYear

1996

fDate

4-6 Sep 1996

Firstpage

60

Abstract

SMART is a new tool for performance, reliability, availability, and performability modeling. Numerical solution algorithms are available for both continuous- and discrete-time Markov chains. Mixed-time non-Markovian models can be studied using simulation. Multiple interacting models and fixed-point iterative techniques for the decomposition and solution of complex models can be easily specified. To assist in the model specification and help in avoiding common mistakes, the input language is strongly typed

Keywords

Markov processes; digital simulation; formal specification; performance evaluation; software tools; Markov chains; SMART; availability; fixed-point iterative techniques; model specification; multiple interacting models; numerical solution algorithms; performability modeling; performance; reliability; simulation and Markovian analyzer; timing; Analytical models; Availability; Computational modeling; Computer science; Computer simulation; Educational institutions; Iterative algorithms; Performance analysis; Stochastic processes; Timing;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Performance and Dependability Symposium, 1996., Proceedings of IEEE International

Conference_Location

Urbana-Champaign, IL

ISSN

1087-2191

Print_ISBN

0-8186-7484-9

Type

conf

DOI

10.1109/IPDS.1996.540204

Filename

540204