Streichert T, Koch D, Haubelt C, Teich J (2008)
Publication Status: Published
Publication Type: Book chapter / Article in edited volumes
Publication year: 2008
Publisher: Springer
Edited Volumes: Organic Computing
Series: Springer Series Understanding Complex Systems
Pages Range: 241-260
ISBN: 9783540776567
DOI: 10.1007/978-3-540-77657-4_11
Networked embedded systems which operate in unattended areas with rare maintenance often make use of redundant resources for guaranteeing reliable service. In this paper, we will present novel concepts for dynamically partitioning and assigning functionality to software as well as hardware reconfigurable resources in a network. As a result, self-adaptive and self-healing systems emerge with a good tradeoff between redundancy and reliability. The proposed concepts are embedded in a three step approach, which 1.) reestablishes the functionality after a resource defect, 2.) optimizes the binding of the running tasks and 3.) creates new replicas of the tasks in the network. In this contribution, we will give an overview over all three parts, but focus on the second step. For this second step, called dynamic hardware/ software partitioning, we will present algorithms, theoretical optimality bounds for workload distributions as well as experimental results. © 2009 Springer-Verlag Berlin Heidelberg.
APA:
Streichert, T., Koch, D., Haubelt, C., & Teich, J. (2008). Concepts for self-adaptive and self-healing networked embedded systems. In Rolf Würtz (Eds.), Organic Computing. (pp. 241-260). Springer.
MLA:
Streichert, Thilo, et al. "Concepts for self-adaptive and self-healing networked embedded systems." Organic Computing. Ed. Rolf Würtz, Springer, 2008. 241-260.
BibTeX: Download