Performance Analysis of Stochastic Digraph Real-Time Task Model

Research output: Contribution to book/anthology/report/conference proceedingBook chapterResearchpeer-review

Abstract

Different models are available when scheduling tasks on real-time systems depending on the need for expressiveness and the properties of the system. The Digraph Real-Time (DRT) Task Model is one of the most expressive models, where the feasibility for hard real-time scheduling is computable in pseudo-polynomial time. We expand the DRT task model to the Stochastic Digraph Real-Time (SDRT) Task Model, which contains the stochastic behavior and not just worst-case values or minimum separations. We present the operational semantics of DRT task systems and expand those to fit SDRT task systems with soft deadlines. We implement these semantics using Uppaal SMC, which we use to simulate a set of SDRT tasks. The simulations are used in performance analysis to obtain metrics such as utilization and the ratio between missed deadlines and the number of released jobs. We show how the increase in size of the task set affects these different metrics of a scheduling system.
Original languageEnglish
Title of host publicationReal Time and Such : Essays Dedicated to Wang Yi to Celebrate His Scientific Career
EditorsSusanne Graf, Paul Petterson, Bernhard Steffen
Number of pages22
PublisherSpringer
Publication date23 Oct 2024
Pages98-119
ISBN (Print)978-3-031-73750-3
ISBN (Electronic)978-3-031-73751-0
DOIs
Publication statusPublished - 23 Oct 2024
SeriesLecture Notes in Computer Science
Volume15230
ISSN0302-9743

Keywords

  • Digraph Real-Time Task Model
  • Performance Analysis
  • Soft Real-Time Scheduling
  • Statistical Model Checking
  • Uppaal  SMC

Fingerprint

Dive into the research topics of 'Performance Analysis of Stochastic Digraph Real-Time Task Model'. Together they form a unique fingerprint.

Cite this