Skip to Main Content
The rapid increase in heat dissipation in real-time systems imposes various thermal issues. For instance, real-time constraints cannot be guaranteed if a certain threshold temperature is exceeded, as it would immediately reduce the system reliability and performance. Dynamic thermal management techniques are promising methods to prevent a system from overheating. However, when designing real-time systems that make use of such thermal management techniques, the designer has to be aware of their effect on both real-time constraints and worst-case peak temperature. In particular, the worst-case peak temperature of a real-time system with non-deterministic workload is the maximum possible temperature under all feasible scenarios of task arrivals. This study proposes an analytic framework to calculate the worst-case peak temperature of a system with general resource availabilities, which means that computing power might not be fully available for certain time intervals. The event and resource models are based on real-time and network calculus, and therefore, our analysis method is able to handle a broad range of uncertainties in terms of task arrivals and available computing power. Finally, we propose an indicator for the quality of the resource model with respect to worst-case peak temperature and schedulability.