performance evaluation of meta-heursitcs in energy aware real-time scheduling problems

摘要

Energy efficient real-time systems have been a prime concern in the past few years. Techniques at alllevels of system design are being developed to reduce energy consumption. At the physical level, newfabrication technologies attempt to minimize overall chipset power. At the system design level,technologies such as Dynamic Voltage and Frequency Scaling (DVFS) and Dynamic Power Management(DPM) allow for changing the processor frequency on-the-fly or go into sleep modes to minimizeoperational power. At the operating system level, energy-efficient scheduling utilizes DVFS and DPM atthe task level to achieve further energy savings. Most energy-efficient scheduling research efforts focusedon reducing processor power. Recently, system-wide solutions have been investigated. In this work, weextend on the previous work by adapting two evolutionary algorithms for system-wide energyminimization. We analyse the performance of our algorithms under variable initial conditions. We furthershow that our meta-heuristics statistically provide energy minimizations that are closer to the optimum85% of the time compared to about 30% of those achieved by simulated annealing over 500 unique testsets. Our results further demonstrate that in over 95% of the cases, meta-heuristics provide moreminimizations than the CS-DVS static method.