Leakage Aware Scheduling on Maximum Temperature Minimization for Periodic Hard Real-Time Systems

Thermal management is becoming a critical issue in computing system design as the processor power continues to grow exponentially. Since high power consumption leads to high temperature, and high temperature in turn dramatically increases the leakage power consumption, a thermal management technique becomes ineffective if this temperature/leakage relation is not properly addressed in the deep sub-micron domain. This paper incorporates the leakage/temperature dependency into real-time scheduling analysis and presents a novel real-time scheduling method that can reduce the peak temperature when scheduling a hard real-time periodic task set. We formally prove the correctness of the proposed algorithm based on a processor model that can effectively account for the leakage/temperature relationship. The experimental results validate the assumptions of our scheduling method and also demonstrate its effectiveness in terms of feasibility improvement and peak temperature reduction.