Energy-aware scheduling, compilation, and execution of hard-real-time multi-task Java programs

Abstract

Most of the real-world embedded hard real-time applications are multi-task programs. Reduction in energy consumption for these systems is essential, especially if they are battery-powered. For multi-task applications, energy management could be considered in the offline or online scheduler. We developed an offline non-preemptive rate monotonic scheduler which generates an energy-optimized schedule judiciously using the idle intervals inside the rate monotonic schedule to either put the processor to sleep or slow down the previous task while satisfying the hard real-time constraints. In addition to the energy-aware scheduling, we implemented a new inter-task hardware-based run-time technique for energy management. We use three different methods of intra/inter-task and compile/run time energy management schemes to find out the optimum method for a multi-task hard-real-time Java program. Our results show that the offline energy-optimized schedule when used with intra/inter-task and compile/run time energy optimization techniques, gives around 35 to 40% energy reduction for the experimental multi-task program.