@inproceedings{bf1590031fa5417a8741274551ddafab,
title = "Resource-aware parameter tuning for real-time applications",
abstract = "Executing multiple applications on a multi-core system while the workload of all applications varies brings the challenge of dynamically adapting resource allocations and parametrizing with respect to constraints e.g. timing limits of real-time applications. We present a hybrid approach which extracts a set of Pareto-optimal operating points during design time which are used to dynamically parameterize the periodic application during run-time. The setup is done at the beginning of each iteration of the execution and exclusively allocates processing elements from the system depending on the current workload. The parametrization is performed with the observed information about workload complexity and allocated resources. Therefore guarantees on time limits can be granted for all iterations including situations when the number of available processing elements has been decreased sharply.",
keywords = "Parameter tuning, Reliability, Resource reservation, Resource-aware application, Self-aware application",
author = "Dirk Gabriel and Walter Stechele and Stefan Wildermann",
note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2019.; 32nd International Conference on Architecture of Computing Systems, ARCS 2019 ; Conference date: 20-05-2019 Through 23-05-2019",
year = "2019",
doi = "10.1007/978-3-030-18656-2\_4",
language = "English",
isbn = "9783030186555",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "45--55",
editor = "Martin Schoeberl and Christian Hochberger and J{\"u}rgen Brehm and Thilo Pionteck and Sascha Uhrig",
booktitle = "Architecture of Computing Systems - ARCS 2019 - 32nd International Conference, Proceedings",
}