Briag Le Nabec, Belgacem Ben Hedia, Jean-Philippe Babau
{"title":"QuaRTOS-DSE: A Tool for Design Space Exploration of Embedded Real-Time System","authors":"Briag Le Nabec, Belgacem Ben Hedia, Jean-Philippe Babau","doi":"10.1109/ISORC.2018.00014","DOIUrl":null,"url":null,"abstract":"To improve the quality, the performance and the real-time correctness of embedded real-time systems, one main design activity is the deployment consisting on mapping functions to tasks and tasks to processors. Several deployment strategies have been proposed by the literature. Each of them is well adapted for one deployment step, one optimization criterion and one type of application. But for realistic systems and general case, none of them is optimal. This paper presents a model-based approach to facilitate the integration and the composition of different deployment strategies. The framework allows to combine different strategies for different parts of the system, and to apply consecutively different strategies. To compare different compositions of strategies, evaluations are proposed based on a set of real-time, performance and design quality criteria. The library of strategies and evaluation criteria may be easily extended by respecting a generic API. Evaluation on case studies shows the interest of combining several strategies to optimize different criteria like real-time and quality of design criteria.","PeriodicalId":395536,"journal":{"name":"2018 IEEE 21st International Symposium on Real-Time Distributed Computing (ISORC)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 IEEE 21st International Symposium on Real-Time Distributed Computing (ISORC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISORC.2018.00014","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
To improve the quality, the performance and the real-time correctness of embedded real-time systems, one main design activity is the deployment consisting on mapping functions to tasks and tasks to processors. Several deployment strategies have been proposed by the literature. Each of them is well adapted for one deployment step, one optimization criterion and one type of application. But for realistic systems and general case, none of them is optimal. This paper presents a model-based approach to facilitate the integration and the composition of different deployment strategies. The framework allows to combine different strategies for different parts of the system, and to apply consecutively different strategies. To compare different compositions of strategies, evaluations are proposed based on a set of real-time, performance and design quality criteria. The library of strategies and evaluation criteria may be easily extended by respecting a generic API. Evaluation on case studies shows the interest of combining several strategies to optimize different criteria like real-time and quality of design criteria.