远洋测量舰船具有覆盖范围广、机动灵活性强等优势，在通信和测控领域被广泛使用。然而，远洋测量舰船的硬件资源是受限的。当前远洋测量舰船电子系统采用传统任务间的最晚截止时间处理的调度方式粒度粗、灵活性差，导致执行任务的性能下降、延误率升高。因此，如何在硬件限制条件下设计一个高效的任务调度算法，满足远洋通信和测控需求，就成了一个重要而有挑战性的问题。为了解决这一问题，本文提出一种资源感知的任务内部调度设计方法来提高远洋测量舰船的任务执行效率。与传统的粗粒度的任务间调度算法不同，在任务内部调度方法中，任务可以在执行进行中被打断。基于远洋测量舰船实际通信和测控数据的仿真分析及硬件实验的结果表明，与EDF方式相比，本文的算法可以有效降低任务延误率30%，提升利用效率20%。
With the advantages of large covered range and strong flexibility, oceanic tracking ships are widely used for data transmission and processing. However, the hardware resources of electronic equipment on the oceanic tracking ships are quite limited. The traditional inter-task earlier deadline first (EDF) scheduling algorithm used on the oceanic tracking ships, which is coarse-grained and suffers from low flexibility, leads to low performance and high deadline miss rates (DMR) of executing tasks. Hence, the question on how to design an effective algorithm to perform the ever-increasing telemetry tasks well under the restriction of limited hardware resources is of great importance and challenge. To solve the problem, this paper proposes a resource-aware intra-task scheduling algorithm for the oceanic tracking ships to improve the effectiveness of task execution. Different from the traditional coarse-grained inter-task scheduling algorithm, in the fine-grained intra-task scheduling algorithm, tasks can be interrupted and scheduled during execution. Based on the real transmission and processed data on the oceanic tracking ships, the results of simulation analysis and hardware experiments show that the proposed scheduling algorithm reduces 30% DMRs and improves 20% resource utilization comparing with the inter-task method.
2018,40(11): 148-152 收稿日期：2018-09-15
DOI：10.3404/j.issn.1672-7649.2018.11.030
分类号：U615
作者简介：张大铭(1987-),男,工程师,主要从事电气系统测控设计工作
参考文献：
[1] KANSAL A, HSU J, ZAHEDI S, et al. Power management in energy harvesting sensor networks[J]. ACM Transactions on Embedded Computing Systems (TECS)-Special Section LCTES'05, 2007, 6(4):1-35.
[2] AUDET D, OLIVEIRA L C, MACMILLAN N, et al. Scheduling recurring tasks in energy harvesting sensors[C]//2011 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), Shanghai, China, 2011.
[3] LIU C L, LAYLAND J W. Scheduling algorithms for multiprogramming in a hard-real-time environment[J]. Journal of the ACM (JACM), 1973, 20(1):46-61.
[4] MOSER C, CHEN J, THIELE L. Reward maximization for embedded systems with renewable energies[C]//14th IEEE International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA'08), Kaohisung, Taiwan, China, 2008.
[5] MOSER C, THIELE L, BRUNELLI D, et al. Adaptive power management in energy harvesting systems[C]//Design, Automation and Test in Europe Conference and Exhibition (DATE'07), Nice Acropolis, France, 2013.
[6] ZHANG D, LI S. Intra-task scheduling for storage-less and converter-less solar-powered non-volatile sensor nodes[C]//International Conference on Computer Design (ICCD), Seoul, Korea, 2014.
[7] LI H, LIU Y. Performance-aware task scheduling for energy harvesting non-volatile processors considering power switching overhead[C]//ACM/EDAC/IEEE Design Automation Conference (DAC), Austin, Texas, USA, 2016.
