In this paper, we study the problem on how to reduce the overall energy consumption while at the same time ensuring the timing and maximum temperature constraints for a real-time system. We incorporate the interdependence of leakage, temperature and supply voltage into analysis and develop a novel method to quickly estimate the overall energy consumption. Based on this method, we then propose a scheduling technique to minimize the overall energy consumption under the maximum temperature constraint. Our experimental results show that the proposed energy estimation method can achieve up to four-order-of-magnitude speedup compared with existing approaches while keeping the maximum estimation error within 4.8%. In addition, simulation results also demonstrate that our proposed energy minimization method consistently outperforms previous related approaches significantly.
展开▼