Energy-Efficient Fault-Tolerant Scheduling of Reliable Parallel Applications on Heterogeneous Distributed Embedded Systems

Xie, Guoqi; Chen, Yuekun; Xiao, Xiongren; Xu, Cheng; Li, Renfa; Li, Keqin

doi:10.1109/tsusc.2017.2711362

Cited by 68 publications

(59 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There has been a large body of works investigating and trying to optimise reliability, energy consumption, as well as timing performance, in embedded systems [10], [6], [11], [12], [4]. On reliability, the target is to reduce the occurrence of faults, transient or permanent, during task execution.…”

Section: Related Workmentioning

confidence: 99%

Dynamic DAG Scheduling on Multiprocessor Systems: Reliability, Energy, and Makespan

Huang

Jiao

et al. 2020

IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

Self Cite

View full text Add to dashboard Cite

Multiprocessor systems are increasingly deployed in real-time applications, where reliability, energy consumption, and makespan are often the main scheduling objectives. In this work, we investigate dynamic scheduling of tasks modelled by directed acyclic graphs (DAGs), which is an NP-hard problem with all existing methods being heuristics. Our contributions have two steps: (i) Assuming that the allocation of DAG nodes to processors is given, we propose OEA (Optimal Energy Allocation) and SOEA (Search-based OEA)-the first optimal methods that minimise the energy consumption whilst satisfying the reliability requirement-for homogeneous and heterogeneous systems, respectively; (ii) We present a novel scheduling algorithm ODS (Out-Degree Scheduling) that allocates the DAG nodes according to their out-degrees, and considering energy consumption, reliability, as well as dynamic finish time. ODS dominates the widely applied HEFT (Heterogeneous Earliest Finish Time) in makespan. Combining SOEA with ODS makes a complete solution to the problem of dynamic DAG scheduling on multiprocessor systems, and achieves generally better results compared to the existing approaches. Specifically, in most cases, we are better on all the three objectives, i.e., reliability, energy as well as makespan, and in other cases we are better on some of the objectives.

show abstract

Section: Related Workmentioning

confidence: 99%

Dynamic DAG Scheduling on Multiprocessor Systems: Reliability, Energy, and Makespan

Huang

Jiao

et al. 2020

IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Res ← Res ∪ ERPOT(Λi, Pj, T k ) 9: Fig. 10 shows the resulting Pareto front in 3D for three different values of P obj , 1.3 W , 2.5 W , and 4.0 W .…”

Section: B Pareto Fronts Obtained With Erpotmentioning

confidence: 99%

“…In [9], Xie et al present an energy-efficient fault-tolerant list scheduling heuristic. The application model is a DAG of tasks, and the architecture model is a distributed memory multiprocessor with a CAN network.…”

Section: Related Workmentioning

confidence: 99%

“…Considering these four criteria simultaneously during the design phase is very difficult because they are antagonistic [1], [2], [4]- [9]. For instance, the total execution time and reliability are antagonistic because increasing the reliability requires some form of redundancy (be it spatial or temporal), which negatively impacts the execution time.…”

Section: Introductionmentioning

confidence: 99%

“…Although several studies have addressed some of these parameters, none have considered these four criteria jointly in an optimization problem. For instance, some studies completely ignore the reliability [4], [14] or the temperature [2], [9]. Other studies tackle the problem as a hardware/software co-design problem, jointly optimizing the floorplan of the multicore and the schedule of the application task graph to minimize the peak temperature [14], but without considering the reliability.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

ERPOT: A Quad-Criteria Scheduling Heuristic to Optimize Execution Time, Reliability, Power Consumption and Temperature in Multicores

Abdi

Girault

Zarandi

2019

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

We investigate multi-criteria optimization and Pareto front generation. Given an application modeled as a Directed Acyclic Graph (DAG) of tasks and a multicore architecture, we produce a set of non-dominated (in the Pareto sense) static schedules of this DAG onto this multicore. The criteria we address are the execution time, reliability, power consumption, and peak temperature. These criteria exhibit complex antagonistic relations, which make the problem challenging. For instance, improving the reliability requires adding some redundancy in the schedule, which penalizes the execution time. To produce Pareto fronts in this 4-dimension space, we transform three of the four criteria into constraints (the reliability, the power consumption, and the peak temperature), and we minimize the fourth one (the execution time of the schedule) under these three constraints. By varying the thresholds used for the three constraints, we are able to produce a Pareto front of non-dominated solutions. We propose two algorithms to compute static schedules. The first is a ready list scheduling heuristic called ERPOT (Execution time, Reliability, POwer consumption and Temperature). ERPOT actively replicates the tasks to increase the reliability, uses Dynamic Voltage and Frequency Scaling to decrease the power consumption, and inserts cooling times to control the peak temperature. The second algorithm uses an Integer Linear Programming (ILP) program to compute an optimal schedule. However, because our multicriteria scheduling problem is NP-complete, the ILP algorithm is limited to very small problem instances. Comparisons showed that the schedules produced by ERPOT are on average only 10% worse than the optimal schedules computed by the ILP program, and that ERPOT outperforms the PowerPerf-PET heuristic from the literature on average by 33%.

show abstract

Energy-Aware Partial-Duplication Task Mapping Under Real-Time and Reliability Constraints

Cui

Kritikakou

et al. 2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

An efficient task execution on multicore platforms can lead to low energy consumption. To achieve that, an Integer Non-Linear Programming (INLP) formulation is proposed that performs task mapping by jointly addressing task allocation, task frequency assignment, and task duplication. The goal is to minimize energy consumption under real-time and reliability constraints. To provide an optimal solution, the original INLP problem is safely transformed to an equivalent Mixed Integer Linear Programming (MILP) problem. The comparison of the proposed approach with existing energy-aware task mapping approaches shows that the proposed approach is able to find solutions when other approaches fail, achieving an overall lower energy consumption.

show abstract

Energy-Efficient Fault-Tolerant Scheduling of Reliable Parallel Applications on Heterogeneous Distributed Embedded Systems

Cited by 68 publications

References 41 publications

Dynamic DAG Scheduling on Multiprocessor Systems: Reliability, Energy, and Makespan

Dynamic DAG Scheduling on Multiprocessor Systems: Reliability, Energy, and Makespan

ERPOT: A Quad-Criteria Scheduling Heuristic to Optimize Execution Time, Reliability, Power Consumption and Temperature in Multicores

Energy-Aware Partial-Duplication Task Mapping Under Real-Time and Reliability Constraints

Contact Info

Product

Resources

About