Reliability-Driven Energy-Efficient Task Scheduling for Multiprocessor Real-Time Systems

Abstract

Abstract-This paper proposes a reliability-driven task scheduling scheme for multiprocessor real-time embedded systems that optimizes system energy consumption under stochastic fault occurrences. The task scheduling problem is formulated as an integer linear program where a novel fault adaptation variable is introduced to model the uncertainties of fault occurrences. The proposed scheme, which considers both the dynamic power and the leakage power, is able to handle the scheduling of independent tasks and tasks with precedence constraints, and is capable of scheduling tasks with varying deadlines. Experimental results have demonstrated that the proposed reliability-driven parallel scheduling scheme achieves energy savings of more than 15% when compared to the approach of designing for the corner case of fault occurrences.

Variation-Aware Layer Assignment With Hierarchical Stochastic Optimization on a Multicore Platform

“…To solve the two-stage stochastic linear program, a commonly-used technique is the sample average approximation technique [12], [13], and [20]. In this method, roughly speaking, each time some random samples are generated according to the probability density function of ω and the expected value E[Q(x, ϑ(ω))] can be approximated by averaging the Q values among samples.…”

“…Further, such a new framework allows us to easily parallelize the algorithm. Our idea in this paper is motivated from and in the same spirit as the previous works [12] [13]. However, the works [12] [13] use the robust parameter idea to tackle the embedded system design problems but not the layer assignment problem.…”

“…(13) would still be too large to be efficiently solved for large signal nets. We use the following hierarchical optimization technique to tackle this difficulty.…”

“…This is accomplished by adding new constraints such as x e,l i = 1 to Eqn. (13) if x e,l i is equal to 1 in the solution of the relaxed mathematical program. In addition, In the solution, those variables close to 1 or 0 will be rounded and fixed.…”

Energy-aware Fault-tolerant Scheduling for Hard Real-time Systems

“…However, current commercial variable voltage processors only have a finite number of speeds [122], [111]. In addition, it takes time for a processor to change its running modes.…”