Using Statistical Model Checker for Schedulability Analysis of Real-Time Systems Under Uncertainty

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F26%3A0198918" target="_blank" >RIV/00216305:26230/26:0198918 - isvavai.cz</a>

Výsledek na webu

<a href="https://link.springer.com/chapter/10.1007/978-3-032-01377-4_13" target="_blank" >https://link.springer.com/chapter/10.1007/978-3-032-01377-4_13</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/978-3-032-01377-4_13" target="_blank" >10.1007/978-3-032-01377-4_13</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Using Statistical Model Checker for Schedulability Analysis of Real-Time Systems Under Uncertainty

Popis výsledku v původním jazyce

Schedulability analysis aims to decide whether it is possible to meet the timing constraints of the given subset of real-time tasks that will be scheduled by the given policy and executed on the given platform. For some situations (classes of systems and conditions), a guaranteed/proven schedulability analysis method exists. As such a method may be absent for other situations and there may be a practical need to cope with the schedulability analysis in such problematic (but realistic, burdened with uncertainty) situations, the analysis must rely on alternative means. This paper presents some of the situations (related, e.g., to the drift of a clock and OSTime, event/interrupt management, task design patterns or OS components such as a scheduler and task queues) that represent a problem for guaranteed approaches first. Then, it presents our approach for interruptible CPU-based systems; it builds on a simulation model over a network of stochastic timed automata, an instrument able to cope with such situations. To analyze schedulability, we apply the statistical model-checking technique to our model. The technique has already shown to easily scale to complex dynamic systems as well as to efficiently solve various problems. Our results indicate that the model and the technique are appropriate for schedulability analysis in realistic situations, where the computational complexity and result of the analysis are driven by predefined parameters such as the degree of confidence.

Název v anglickém jazyce

Using Statistical Model Checker for Schedulability Analysis of Real-Time Systems Under Uncertainty

Popis výsledku anglicky

Schedulability analysis aims to decide whether it is possible to meet the timing constraints of the given subset of real-time tasks that will be scheduled by the given policy and executed on the given platform. For some situations (classes of systems and conditions), a guaranteed/proven schedulability analysis method exists. As such a method may be absent for other situations and there may be a practical need to cope with the schedulability analysis in such problematic (but realistic, burdened with uncertainty) situations, the analysis must rely on alternative means. This paper presents some of the situations (related, e.g., to the drift of a clock and OSTime, event/interrupt management, task design patterns or OS components such as a scheduler and task queues) that represent a problem for guaranteed approaches first. Then, it presents our approach for interruptible CPU-based systems; it builds on a simulation model over a network of stochastic timed automata, an instrument able to cope with such situations. To analyze schedulability, we apply the statistical model-checking technique to our model. The technique has already shown to easily scale to complex dynamic systems as well as to efficiently solve various problems. Our results indicate that the model and the technique are appropriate for schedulability analysis in realistic situations, where the computational complexity and result of the analysis are driven by predefined parameters such as the degree of confidence.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

20205 - Automation and control systems

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Lecture Notes in Computer Science, Bridging the Gap Between AI and Reality

ISBN

978-3-032-01376-7

ISSN

—

e-ISSN

—

Počet stran výsledku

24

Strana od-do

233-256

Název nakladatele

Springer Nature Switzerland

Místo vydání

Cham

Místo konání akce

Crete, Greece

Datum konání akce

30. 10. 2024

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—