Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Efficient on-line schedulability tests for heterogeneous distributed real-time systems

: Becker, K.
: Zeller, M.; Kerstan, T.

Volltext urn:nbn:de:0011-n-1341288 (3.0 MByte PDF)
MD5 Fingerprint: 290513bcb56253949b0cede2159d03cb
Erstellt am: 28.6.2011

Paderborn, 2010, 149 S.
Paderborn, Univ., Master Thesis, 2010
Master Thesis, Elektronische Publikation
Fraunhofer ESK ()
automotive; scheduling; dynamic; schedulability

In this thesis it is investigated, how the schedulability of a set of tasks can be analyzed in an efficient manner. It is assumed, that the tasks are running in a heterogeneous distributed environment with hard real-time constraints, and that the task sets are dynamic. The main focus is laid on the situation, in which new tasks are desired to be added to an existing task set during runtime. The aim is to determine, whether or not the task set is still schedulable, if the new tasks would be added to the existing ones. Hence, the investigated algorithms should be efficiently applicable on-line as acceptance test for the new tasks. It is shown in this thesis, how this special situation can be exploited via adapting some existing schedulability tests. All this is done including the consideration of the heterogeneity aspects and some further restrictions, that appear in the automotive domain.