Pruning-based energy-optimal device scheduling for hard real-time systems

Dynamic power management (DPM) provides a simple, elegant and flexible method for reducing energy consumption in embedded real-time systems. However, I/O-centric DPM techniques have been studied largely for non-real-time environments. We present an offline device scheduling technique for real-time systems that generates an energy-optimal device schedule for a given task set while guaranteeing that all real-time deadlines are met. Our method takes as inputs a task set and a device-usage list for each task, and it schedules the tasks such that the energy consumed by the set of I/O devices is minimized. We compare our algorithm to an exhaustive enumeration method and show that the proposed algorithm is very efficient in terms of memory usage and computation time. We also present case studies to show that I/O-centric DPM methods can result in significant energy savings.