系统仿真学报 ›› 2018, Vol. 30 ›› Issue (9): 3399-3403.doi: 10.16182/j.issn1004731x.joss.201809021

• 仿真应用工程 • 上一篇    下一篇

组合服务的时序鲁棒性仿真分析

叶恒舟1,2, 李陶深3, 关云慧2   

  1. 1. 广西大学电气工程学院,广西 南宁530004;
    2. 桂林理工大学信息科学与工程学院,广西 桂林 541000;
    3. 广西大学计算机与电子信息学院,广西 南宁530004
  • 收稿日期:2016-05-10 出版日期:2018-09-10 发布日期:2019-01-08
  • 作者简介:叶恒舟(1980-),男,湖北汉川,博士生,教授,研究方向为Web服务组合、优化设计。
  • 基金资助:
    国家自然科学基金(61762031),广西自然科学基金(2014GXNSFBA118269),广西嵌入式技术与智能系统重点实验室开放基金(2018A-05)

Simulated Analysis of Temporal Robustness for Composite Service

Ye Hengzhou1,2, Li Taoshen3, Guan Yunhui2   

  1. 1.College of Electrical Engineering, Guangxi University, Nanning 530004, China;
    2.College Of Information Science and Engineering, Guilin University Of Technology, Guilin 541000, China;
    3.School of Computer and Electronic Information, Guangxi University, Nanning 530004, China
  • Received:2016-05-10 Online:2018-09-10 Published:2019-01-08

摘要: 服务的执行时间可能受到网络、CPU、磁盘I/O及负载变动的影响,进而影响组合服务的完工时间。研究组合服务的完工时间的分布规律与影响因素对于保证组合服务的性能至关重要。采用有向无环图描述组合服务,提出了一种评估组合服务的完工时间的波动程度的鲁棒性指标。通过理论分析给出了该指标的界限;通过仿真分析,探索了服务的执行时间的分布与波动情况、组合服务中简单服务的个数等因素对该指标的影响。为分析鲁棒服务组合的必要性和设计鲁棒服务组合方法奠定了基础。

关键词: 组合服务, 时序鲁棒性, 有向无环图, 正态分布

Abstract: The execution time of a service may fluctuate due to the variety of network, CPU, disk I/O and workloads, which will affect the makespan of the composite service. It is crucial to study the probability distribution and influencing factors of the real makespan to guarantee the performance of the composite service. This paper adopts directed acyclic graph (DAG) to describe the composite service. Also, with the goal of evaluating the variety degree of the makespan, we propose a temporal robustness metric for composition services. Through the rigorous analysis, we can find the boundary of this metric. By the analysis of simulation, we explore the variety of several factors to this metric, such as the distribution and fluctuation of service execution time, the number of simple services in the composition service. The experiment results lay the theoretical foundation for the analysis and design of robustness service composition.

Key words: composite service, temporal robustness, directed acyclic graph, normal distribution

中图分类号: