Creative Commons Attribution 3.0 Unported license
A system is said to be resilient if slight deviations from expected behavior during run-time does not lead to catastrophic degradation of performance: minor deviations should result in no more than minor performance degradation. In mixed-criticality systems, such degradation should additionally be criticality-cognizant. The applicability of control theory is explored for the design of resilient run-time scheduling algorithms for mixed-criticality systems. Recent results in control theory have shown how appropriately designed controllers can provide guaranteed service to hard-real-time servers; this prior work is extended to allow for such guarantees to be made concurrently to multiple criticality-cognizant servers. The applicability of this approach is explored via several experimental simulations in a dual-criticality setting. These experiments demonstrate that our control-based run-time schedulers can be synthesized in such a manner that bounded deviations from expected behavior result in the high-criticality server suffering no performance degradation and the lower-criticality one, bounded performance degradation.
@InProceedings{papadopoulos_et_al:LIPIcs.ECRTS.2018.14,
author = {Papadopoulos, Alessandro Vittorio and Bini, Enrico and Baruah, Sanjoy and Burns, Alan},
title = {{AdaptMC: A Control-Theoretic Approach for Achieving Resilience in Mixed-Criticality Systems}},
booktitle = {30th Euromicro Conference on Real-Time Systems (ECRTS 2018)},
pages = {14:1--14:22},
series = {Leibniz International Proceedings in Informatics (LIPIcs)},
ISBN = {978-3-95977-075-0},
ISSN = {1868-8969},
year = {2018},
volume = {106},
editor = {Altmeyer, Sebastian},
publisher = {Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
address = {Dagstuhl, Germany},
URL = {https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.ECRTS.2018.14},
URN = {urn:nbn:de:0030-drops-89899},
doi = {10.4230/LIPIcs.ECRTS.2018.14},
annote = {Keywords: mixed criticality, control theory, run-time resilience, bounded overloads}
}