Improving Responsiveness of Hard Real-Time Embedded Systems

Hard real-time systems are found in many critical embedded applications, for example aeroplane flight control, industrial production line control, and so on. The safe scheduling of tasks in these hard real-time systems is crucial to their correct operation, but the hard constraints of this type of scheduling reduce the responsiveness of the systems. In this paper we show the application of runtime analysis in informing the design of hard real-time embedded systems by allowing scheduled tasks to be dynamically re-ordered to improve the average responsiveness, while still meeting the hard constraints imposed by the system. The technique is semi-automated, and uses the reduce computer algebras system to precalculate a symbolic form of the runtime of scheduled tasks. The symbolic form is used to modify the source code of the scheduler.