Logo des Repositoriums
 

Machine learning in run-time control of multicore processor systems

Kurzanzeige
dc.contributor.authorMaurer, Florian
dc.contributor.authorThoma, Moritz
dc.contributor.authorSurhonne, Anmol Prakash
dc.contributor.authorDonyanavard, Bryan
dc.contributor.authorHerkersdorf, Andreas
dc.date.accessioned2025-01-30T14:13:50Z
dc.date.available2025-01-30T14:13:50Z
dc.date.issued2023
dc.description.abstractModern embedded and cyber-physical applications consist of critical and non-critical tasks co-located on multiprocessor systems on chip (MPSoCs). Co-location of tasks results in contention for shared resources, resulting in interference on interconnect, processing units, storage, etc. Hence, machine learning-based resource managers must operate even non-critical tasks within certain constraints to ensure proper execution of critical tasks. In this paper we demonstrate and evaluate countermeasures based on backup policies to enhance rule-based reinforcement learning to enforce constraints. Detailed experiments reveal the CPUs’ performance degradation caused by different designs, as well as their effectiveness in preventing constraint violations. Further, we exploit the interpretability of our approach to further improve the resource manager’s operation by adding designers’ experience into the rule set.en
dc.identifier.doihttps://doi.org/10.1515/itit-2023-0056
dc.identifier.issn2196-7032
dc.identifier.urihttps://dl.gi.de/handle/20.500.12116/45639
dc.language.isoen
dc.pubPlaceBerlin
dc.publisherDe Gruyter
dc.relation.ispartofit - Information Technology: Vol. 65, No. 4-5
dc.subjectlearning classifier tables
dc.subjectmachine learning
dc.subjectmulticore processor systems
dc.subjectrun-time control
dc.titleMachine learning in run-time control of multicore processor systemsen
dc.typeText/Journal Article
mci.conference.sessiontitleArticle
mci.reference.pages164-176

Dateien

Sammlungen

it - Information Technology 65(4-5) - November 2023