Energy Management for Periodic Real-Time Tasks with Variable Assurance Requirements
Reliability-aware power management (RAPM) schemes, which consider the negative effects of voltage scaling on system reliability, were recently studied to save energy while preserving system reliability. However, in previous RAPM schemes, real-time tasks are occasionally treated unfairly and the selected jobs are determined in greedy fashion. In this paper, we study static flexible RAPM schemes for real-time periodic tasks, which consider the assurance requirements of tasks and manages a subset of jobs for every task accordingly. The problem is shown to be NP-hard in the strong sense and the upper bounds on energy savings are discussed. For a special case of tasks’ recovery patterns, a pseudo-polynomial static scheme is proposed. Dynamic schemes that explore dynamic slack for better energy savings and reliability enhancement are also discussed. The schemes are evaluated extensively through simulations. The results show that, compared to the previous RAPM schemes, the new flexible RAPM schemes can guarantee the assurance requirements and provide fairness for all tasks, but at the cost of decreased energy savings. However, when combining with dynamic schemes, such cost can be effectively recovered.