Toward an Optimal Power-Aware Scheduling Technique

With the rapid demand of portable and mobile embedded systems, reducing energy consumption has become a vital research area in the real-time embedded systems community. There exist many research works which describe scheduling algorithms minimizing the total energy of a system. However, very few works describe scheduling techniques for which the energy consumption is optimal. This paper describes a scheduling technique which provides an optimal schedule in terms of power consumption on a uniprocessor platform running non-preeemptive jobs. To the best of our knowledge, there is no analytical formula of the minimum energy for an arbitrary job set with n jobs, where n ≥ 2 is the number of jobs. A multi-case analytical formula for the minimum energy of job sets with two jobs is provided here as an original contribution.