An inter-task real time DVFS scheme for multiprocessor embedded systems

Abstract

In this paper 1 , we have addressed energy-efficient scheduling of real time applications intended to be executed on multiprocessor systems. Our proposed technique, called Deterministic Stretch-to-Fit (DSF) technique, is based on inter-task real time dynamic voltage and frequency scaling (RT-DVFS). It mainly comprises of three components. Firstly, we propose an online algorithm to reclaim energy by adapting to the variations in actual workload of target application tasks. Secondly, we extend our online algorithm with an adaptive and speculative speed adjustment mechanism. This mechanism anticipates early completion of future task instances based on the information of their average workload. Thirdly, we propose a one-task extension technique for multi-task multiprocessor systems. No real time constraints of target application are violated while applying our proposed technique. Simulation results show that our online slack reclamation algorithm alone gives up to 53% gains on energy consumption and our extended speculative speed adjustment mechanism, along with the one-task extension technique, gives additional gains, reaching a theoretical low-bound on the scalable frequency and voltage.

Hybrid power management in real time embedded systems: an interplay of DVFS and DPM techniques

“…Aydin and Yang (2003) propose their solution for periodic hard real time tasks on identical multiprocessors with DVFS support when only partitioned scheduling is used. Bhatti et al (2010) propose an inter-task multiprocessor DVFS technique in which execution of tasks in real schedule mimics the canonical schedule (produced at runtime) and dynamic slack is fully consumed by a single appropriate priority task.…”

“…DVFS experts: First DVFS expert used in our experiments is presented in Bhatti et al (2010) in which, authors suggest that their technique is based on an online slack reclamation algorithm which permits the dynamic slack, produced by the precedent task, to be fully consumed by single subsequent task at the appropriate priority level. Such greedy allocation of slack allows large variations in F op and V op which eventually results in larger gains on energy consumption.…”

Impact of operating points on DVFS power management

“…Other perspectives of this work are to implement and study other power strategies like [17] and [18] on real hardware platforms boards like the ARM1176JZF-S and the ARM11 MPCore in order to explore their effectiveness in real development world. In particular, we will further analyze the impact of operating points for such other DVFS strategies if generalization is possible and at which conditions.…”

“…Rabbits rely on QEMU for software execution and SystemC for hardware modeling [17]. The negative energy gains obtained are probably due to the fact that differences of power level in consecutive operating points are small for the target platform we have used (Versatile ARM1176JZF-S).…”

Accurate energy characterization of OS services in embedded systems

“…That is why, when calculating the overhead of OS services, we execute the application on the hardware platform while adapting a low power technique: The dynamic voltage frequency scaling (DVFS) technique, [19]. In this article, we adapt an instance of DVFS technique: the deterministic stretch-to-fit technique (DSF) proposed in [20], it is based on the slowdown strategy of reducing the processor power consumption. Slowdown is known to reduce the dynamic power consumption at the cost of increased execution time for a given computation task.…”