Search-Based Temporal Testing in an Embedded Multicore Platform

Komsan Srivisut*, John A. Clark, Richard F. Paige

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Multicore processors have now become the norm. However, for many embedded real-time systems their use introduces challenges in verification as their shared components are potential channels for interference. Of particular interest is the determination for each task of its worst case (longest) execution time (WCET). In this paper, we investigate the effectiveness of a variety of metaheuristic search algorithms for dynamically finding extreme execution times of tasks executing on a multicore processor. Over finite search spaces, these are shown to perform considerably better than randomly generated test inputs and the work reveals significant performance differences between the various algorithms.

Original languageEnglish
Title of host publicationApplications of Evolutionary Computation - 21st International Conference, EvoApplications 2018, Proceedings
PublisherSpringer
Pages794-809
Number of pages16
Volume10784 LNCS
ISBN (Print)9783319775371
DOIs
Publication statusPublished - 2018
Event21st International Conference on Applications of Evolutionary Computation, EvoApplications 2018 - parma, Italy
Duration: 4 Apr 20186 Apr 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10784 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Conference

Conference21st International Conference on Applications of Evolutionary Computation, EvoApplications 2018
Country/TerritoryItaly
Cityparma
Period4/04/186/04/18

Keywords

  • Genetic algorithms
  • Hill climbing
  • Multicore
  • Simulated annealing
  • Temporal testing

Cite this