计算机科学 ›› 2011, Vol. 38 ›› Issue (Z10): 230-238.

• 分布式与网络应用 • 上一篇    下一篇

一种基于条件Pi演算的组合服务柔性演化模型

刘涛,曾国荪   

  1. (同济大学计算机科学与技术系 上海 200092);(同济大学嵌入式系统与服务计算教育部重点实验室 上海 200092)
  • 出版日期:2018-11-16 发布日期:2018-11-16
  • 基金资助:
    本文受863项目(2007八八O1Z425,2009l1l1012201),973专项(2007CB316502) , 国家白然基金项目(90718015),NSFC-微软亚洲研究院联合资助项目(60990165,教育部博十点基金项目(C200900721L0035},上海市优秀学科带头人计划项目(10XD1404400),高效能服务器和存储技术国家重点实验室开放基金项目(2009HSSl106)资助。

Model for Flexible Evolution of Composite Services Based on Conditional Pi Calculus

LIU Tao ,ZENG Guo-sun   

  • Online:2018-11-16 Published:2018-11-16

摘要: 组合服务应当具有适应所处环境和商业规则不断变化的能力。现有的服务组合语言和执行机制缺乏应对动态变化所需的可变性和适应性。尽管已有很多扩展,但是组合服务的动态适应正确性还缺乏保障。提出了一种基于条件Pi演算的组合服务柔性演化机制。通过增加归属操作符和条件控制符对经典Pi演算进行扩展,使之与事件一条件一动作模式更好地结合起来,从而提出了一种描述组合服务流程的方法。从分析组合服务各种变化的场景出发,提出了11种基本的变化场景及其不同的柔性演化模式,对每种模式进行了形式化描述和分析。该方法能够保证组合服务的可变性和适应正确性。

关键词: 条件Pi演算,服务自适应,柔性演化模式

Abstract: Composite services ought to be adaptable to changing environments and business rules. Existing service composition languages and execution engines lack variability and adaptability needed to cater for dynamic changes. In spite of many extensions,how to ensure the correctness for the dynamic adaptation of composite services remains a challenge.We proposed a mechanism for flexible evolution of composite services based on conditional Pi calculus. We extended classical Pi calculus through introducing the conditional control operator and the belonging operator, which suits the Event Condition-Action(ECA) pattern better. A method to describe composite service processes was proposed based on conditional Pi calculus and ECA. I}hrough scenario analysis of various composite service changes, eleven basic dynamic scenarios and corresponding flexible evolution patterns were proposed. Each pattern was formalized and analyzed to ensure the variability and adaptation correctness of composite services.

Key words: Conditional Pi calculus, Service adaptation, Flexible evolution pattern

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!