Keynotes

Sanjoy K. Baruah : Multiprocessor Real-time Scheduling Theory: questions (many) and answers (a few)

Due to various inherent advantages of multiprocessor platforms, real-time application systems are increasingly coming to be implemented upon such platforms. However, theoretical developments have not kept pace: currently, our formal understanding of the behavior of such multiprocessor systems is approximately where our knowledge of uniprocessor systems was in the early 1970's. There is consequently a need for developing a theory of multiprocessor real-time scheduling that is as complete and sophisticated as uniprocessor real-time scheduling theory currently is, and that will prove as useful to the designers of real-time systems as uniprocessor real-time theory does today. In this presentation, I will propose a research agenda for multiprocessor real-time scheduling theory that aims to address this need. I will also briefly outline the progress that has been made thus far towards achieving the goals in this agenda.