Apps-usage driven energy management for multicore mobile computing systems

This paper presents an idea to redesign the existing mobile computing system scheduler with Apps-usage awareness. By identifying the Apps-usage patterns, we model the relationship between system power consumption and different combinations of Apps. As observed some of the Apps combinations produce more power dissipation comparing to the others. The proposed scheduler reorders the Apps scheduling sequence according to the recorded Apps combination and power consumption values, hence minimizing the overall power consumption. The scheduling schemes are tested in a simulated environment built with GEM5 and McPat. The simulated result shows power consumption reduction up to 19%.