Multiprocessor Systems-on-Chips

6.3. Process Execution Modeling

So far, we have focused on the timing of a single process execution or communication activity. To determine the load of a component, the number of such executions per time is needed.

6.3.1. Activation Modeling

To model the execution of a process, we must know how the process is activated. We can distinguish two very different types of processes. A process can be started once and then run forever, following an internal loop. The execution time of such a process is infinite, and the communication volume is, in principle, infinite, as well. In this case, the process and communication take whatever resources are available. If such a process uses blocking read then the timing depends on other processes and, eventually, ...