Real-Time System Constraints: Where do They Come From and Where do They Go?

Within the real-time community a great deal of research has been, and is being, conducted on scheduling. There exist a diversity of different scheduling algorithms as well as constraints that should be satis£ed by a feasible schedule. What is not commonly discussed though is how the constraints relate to the requirements on the actual real-time system. In fact, many constraints are mere artifacts that cannot be traced to the system design. This often results in an over-constrained and infeasible scheduling problem. To avoid this we are conducting a survey on industrial realtime system requirements. The aim is to give a better understanding of how to derive adequate constraints. Furthermore, we believe that by considering constraints imposed by industrial requirements such as product cost or hardware limitations, the time complexity of schedule generation can be greatly reduced. In this paper we present some early results from our survey. This includes a description of realtime system requirements and how they relate to commonly used real-time constraints.