ISSN 1000-1239 CN 11-1777/TP

计算机研究与发展 ›› 2015, Vol. 52 ›› Issue (12): 2695-2706.doi: 10.7544/issn1000-1239.2015.20150745

所属专题: 2015面向“互联网+”的应用技术

• 网络技术 • 上一篇    下一篇


葛永琪1, 董云卫1, 顾斌1,2   

  1. 1(西北工业大学计算机学院 西安 710129); 2(北京控制工程研究所 北京 100190) (
  • 出版日期: 2015-12-01
  • 基金资助: 

Preemption Threshold Scheduling for Energy Harvesting Based Cyber-Physical Systems

Ge Yongqi1, Dong Yunwei1, Gu Bin1,2   

  1. 1(School of Computer Science, Northwestern Polytechnical University, Xi’an 710129); 2(Beijing Institute of Control Engineering, Beijing 100190)
  • Online: 2015-12-01

摘要: 在能量收集信息物理融合系统(energy harvesting based cyber-physical systems, EHCPS)中,其能量管理体系结构不同于传统电池供电嵌入式系统,任务调度策略需要考虑能量收集单元的能量输出、电池的能量存储和计算任务的能量消耗.实时任务在满足能量约束的情况下,才能满足时间约束.传统抢占阈值调度的可调度性分析没有考虑任务的能量属性,其阈值分配算法也不适用于EHCPS.针对此问题,提出了一种能量相关抢占阈值调度策略(energy related preemption threshold scheduling, ERPT),在可调度性分析中融入任务能耗属性和能量补充能力,并给出了阈值分配算法,为抢占阈值调度在EHCPS中的应用提供了一种解决方法.通过与目前现有的2个经典调度策略进行比较,验证了ERPT策略能够有效减少任务抢占.

关键词: 信息物理融合系统, 能量收集, 抢占阈值, 可调度性, 最坏情况响应时间

Abstract: In energy harvesting based cyber-physical systems (EHCPS), energy management architecture is different from traditional battery powered embedded systems. The task scheduling of EHCPS should take account of the output power of energy harvesting unit, the energy level of battery and the energy consumption of computing tasks. A real-time task may meet time constraint only if its energy constraint is satisfied. Against this background, the schedulability analysis of conventional preemption threshold scheduling doesn’t consider the tasks energy consumption, thus the preemption threshold assignment algorithm is not suitable for EHCPS. An energy related preemption threshold scheduling (ERPT) for EHCPS is proposed in this paper. It integrates tasks energy consumption and energy supply ability into schedulability analysis, and the preemption threshold assignment algorithms are also presented. ERPT presents a solution for applying preemption threshold scheduling in EHCPS. The proposed scheduling strategy is validated compared with other two existing classical algorithms. The experimental results show that the proposed scheduling can reduce the tasks preemption effectively.

Key words: cyber-physical systems (CPS), energy harvesting, preemption threshold, schedulability, worst-case response time (WCRT)