System Model and Work Flow

The Task Concurrency Management (TCM) methodology developed at IMEC and its university research partners since 1997 [36, 234] addresses the mapping of dynamic and concurrent tasks on multiple processors for real-time embedded systems, where energy consumption is often a major concern. The cornerstone of this methodology is a well-balanced two-phase scheduling method, which provides the required flexibility at a low run-time overhead. The rationale behind the two-phase scheduling is that the only way to handle very dynamic applications in a resource-efficient way is to postpone some design decisions till the run-time stage. Fixing everything at designtime according to the worst case estimation of the system would result in either a too energy-hungry platform or a poorly performing application most of the time. In the meantime, only the minimal amount of effort should be done at the run-time to minimize the overhead, and to still meet all hard realtime constraints.