Minimizing CPU energy in real-time systems with discrete speed management

This article presents a general framework to analyze and design embedded systems minimizing the energy consumption without violating timing requirements. A set of realistic assumptions is considered in the model in order to apply the results in practical real-time applications. The processor is assumed to have as a set of discrete operating modes, each characterized by speed and power consumption. The energy overhead and the transition delay incurred during mode switches are considered. Task computation times are modeled with a part that scales with the speed and a part having a fixed duration, to take I/O operations into account. The proposed method allows to compute the optimal sequence of voltage/speed changes that approximates the minimum continuous speed, which guarantees the feasibility of a given set of real-time tasks, without violating the deadline constraints. The analysis is performed both under fixed and dynamic priority assignments.