A Scalable and Decentral Approach to sustained System Monitoring

Future heterogeneous system are likely to employ reconfigurability. In order to manage the complexity of such systems, the use of so-called Self-X capabilities is required. Of these, Self-Awareness is the essential property, providing required information for autonomous system and application optimization. We therefore propose a scalable, hierarchical, and decentral infrastruc-ture for sustained monitoring, applicable to heterogeneous multi-core systems. By analyzing the gathered data, the actual state of the (sub-)system is classified and appropriate actions may be triggered