Synthesis of Behavioral Models from Scenarios
Synthesis of Behavioral Models from Scenarios
复制标题
DOI：
10.1109/tse.2003.1178048
复制
发表时间：
2003-02
期刊：
影响因子：
--
通讯作者：
Sebastián Uchitel;J. Kramer;J. Magee
中科院分区：
文献类型：
--
作者：
Sebastián Uchitel;J. Kramer;J. Magee
基于场景的规范，如消息序列图（MSC），作为需求规范的一部分非常有用。场景是一个局部故事，描述系统组件、环境和用户如何并发工作和交互，以提供系统级功能。需要将场景组合起来，以提供对系统行为的更完整描述。因此，情景综合是有效使用情景说明的核心。一组情景应该如何解释？他们是如何互相联系的呢？底层语义是什么？在综合多个场景的行为模型时，需要做哪些假设？在本文中，我们提出了一种基于清晰的声音语义的场景合成方法，它可以支持和集成许多现有的场景合成方法。本文的贡献有三个方面。我们首先定义了一个MSC语言的声音抽象语义标记的过渡系统和并行组合。该语言集成了现有的方法的基础上，通过使用高层次的MSC（hMSC）的情况组成和那些基于状态识别通过引入显式组件状态标记。这种组合允许利益相关者将场景规范分解为可管理的部分，并使用hMCS重用场景;它还允许他们使用状态标签将额外的特定于域的信息和一般假设显式地引入场景规范。其次，我们提供了一个声音合成算法，将场景转换为有限顺序过程形式的行为规范。这个规范可以用Labeled Transition System Analyzer使用模型检查和动画进行分析。最后，我们展示了有多少嵌入在现有的合成方法的假设，可以明确和建模在我们的方法。因此，我们提供了一个共同的方法，以规范，综合和分析的基础。
Scenario-based specifications such as Message Sequence Charts (MSCs) are useful as part of a requirements specification. A scenario is a partial story, describing how system components, the environment, and users work concurrently and interact in order to provide system level functionality. Scenarios need to be combined to provide a more complete description of system behavior. Consequently, scenario synthesis is central to the effective use of scenario descriptions. How should a set of scenarios be interpreted? How do they relate to one another? What is the underlying semantics? What assumptions are made when synthesizing behavior models from multiple scenarios? In this paper, we present an approach to scenario synthesis based on a clear sound semantics, which can support and integrate many of the existing approaches to scenario synthesis. The contributions of the paper are threefold. We first define an MSC language with sound abstract semantics in terms of labeled transition systems and parallel composition. The language integrates existing approaches based on scenario composition by using high-level MSCs (hMSCs) and those based on state identification by introducing explicit component state labeling. This combination allows stakeholders to break up scenario specifications into manageable parts and reuse scenarios using hMCSs; it also allows them to introduce additional domain-specific information and general assumptions explicitly into the scenario specification using state labels. Second, we provide a sound synthesis algorithm which translates scenarios into a behavioral specification in the form of Finite Sequential Processes. This specification can be analyzed with the Labeled Transition System Analyzer using model checking and animation. Finally, we demonstrate how many of the assumptions embedded in existing synthesis approaches can be made explicit and modeled in our approach. Thus, we provide the basis for a common approach to scenario-based specification, synthesis, and analysis.