2026-07-25T09:51:05Z https://drops.dagstuhl.de/oai

oai:drops-oai.dagstuhl.de:20309 2024-07-04T05:10:01Z ddc:004 open_access

Optimizing Per-Core Priorities to Minimize End-To-End Latencies Paladino, Francesco Biondi, Alessandro Bini, Enrico Pazzaglia, Paolo Cause-Effect Chains Logical Execution Time End-to-End Latency Design Optimization Task Priorities Data Age Reaction Time Logical Execution Time (LET) allows decoupling the schedule of real-time periodic tasks from their communication, with the advantage of isolating the communication pattern from the variability of the schedule. However, when such tasks are organized in chains, the usage of LET at the task level does not necessarily transfer the same LET properties to the chain level. In this paper, we extend a LET-like model from tasks to chains spanning over multiple cores. We leverage the designed constant latency chains to optimize per-core priority assignment. Finally, we also provide a set of heuristic algorithms, that are compared in a large-scale experimental evaluation. Schloss Dagstuhl – Leibniz-Zentrum für Informatik Francesco Paladino and Alessandro Biondi and Enrico Bini and Paolo Pazzaglia 2024 Is Part Of LIPIcs, Volume 298, 36th Euromicro Conference on Real-Time Systems (ECRTS 2024) InProceedings Text doc-type:ResearchArticle publishedVersion application/pdf doi:10.4230/LIPIcs.ECRTS.2024.6 urn:nbn:de:0030-drops-203094 https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.ECRTS.2024.6 eng https://creativecommons.org/licenses/by/4.0/legalcode