Descripción
|
|
---|---|
Estimating dependability measures for a highly-dependable multi-component system is a problem of great interest in different areas of applications such as computer systems, telecommunications, aircraft design, power utilities, and many others. Increasing demand for system?s reliability cannot be satisfied with increasing reliability of components due to technological restrictions. The alternative is a fault-tolerant system which, through the use of redundancy, has the capability to sustain correct operation in the presence of faults. A system failure should be a rare event. Even with the assumption of exponential laws for failures and repairs of components, analytical or numerical solution methods are impractical due to the excessive number of states of the continuous time Markov chain (CTMC), and simulation becomes the only viable technique of analysis. However, the estimation of rare event probabilities with the ?crude? Monte Carlo technique requires a prohibitively large number of trials in most interesting cases and fast simulation methods for rare events are required. The two main acceleration techniques are importance sampling (IS) and Splitting/RESTART. The advantages of RESTART are that the extension to non-Markovian models is almost straightforward and that it is not so dependent as IS on particular features of the system. | |
Internacional
|
Si |
Nombre congreso
|
9th Workshop on Rare Event Simulation, RESIM?12 |
Tipo de participación
|
960 |
Lugar del congreso
|
Trondheim (Noruega) |
Revisores
|
Si |
ISBN o ISSN
|
|
DOI
|
|
Fecha inicio congreso
|
25/06/2012 |
Fecha fin congreso
|
27/06/2012 |
Desde la página
|
16 |
Hasta la página
|
27 |
Título de las actas
|
Proceedings of the 9th International Workshop on Rare Event Simulation |