Online Speed Fine Tuning Technique for Energy Minimization of Mixed Task Set

This paper aims to present a two phase scheduling algorithm that offers lesser energy consumption for real time systems modeled periodic as well as aperiodic tasks by utilizing the concept of dynamic voltage scaling and dynamic power down. The first phase is same as used in Sim et al. whereas, second phase has been modified by utilizing the concept of idle slot, available slack and power down in case idle slot is enough to switch off and save energy. Besides reduction in energy modified approach also utilizes the system better by reduces number of idle slot and length of idle slot. The extensive examples and simulation results illustrate that our approach can effectively reduce the energy consumption with better system utilization.