Model Driven Incremental Design and Validation of Software Systems

软件系统无处不在。它们是我们日常生活的一部分，它们控制着汽车、飞机、电信网络和服务以及其他家用电器。它们非常复杂，因为它们的固有特征：大尺寸，嵌入式，分布，并发性，实时约束和截止日期。这样的系统需要使用严格的和正式的方法，以确保其功能的正确性和其他非功能的质量，如实时性。模型驱动开发（MDD）是在过去二十年中出现的一种方法，并且正在获得动力。然而，MDD的采用和应用受到可伸缩性和形式化的挑战性问题的阻碍。 到目前为止，MDD的重点一直是开发概要文件，扩展/专门化统一建模语言（UML）和转换语言，只有少数技术被提出来实现。此外，这些技术没有跨越必要的阶段来链接和关联各种工件和阶段，没有规模或者不是正式的。因此，本研究建议的主要目标是有助于复杂的软件系统的MDD与技术，他们的建模，设计和验证。我们建议调查的建模框架，涵盖了几个阶段的开发过程：需求，设计和验证/测试。建模框架将由UML 2的一个子集和相关的配置文件（如MARTE、SysML和UML Profile for Testing（UPT））组成。使用这种形式化的框架，我们的目标是设计符合MDD的技术，增量设计和验证的软件系统。增量设计方法，考虑到功能和非功能的特点，以及增量测试用例生成技术将被设计出来。这些正式的增量技术肯定会扩展，并将有助于MDD的应用和采用。 这项建议旨在设计技术，以提高软件系统的质量，并降低其开发成本。本提案中所处理的研究问题对软件研究界和包括电信和汽车行业在内的行业具有高度重要性。拟议的研究计划将培养研究生谁将加入加拿大的劳动力，并在不同部门的加拿大经济作出贡献。

Software systems are pervasive. They are part of our daily life, they control cars, airplanes, telecommunications networks and services and other home appliances. They are very complex, because of their inherent characteristics: large size, embedded, distribution, concurrency, real-time constraints and deadlines. Such systems require usage of rigorous and formal methods in order to ensure their functional correctness and other non-functional qualities such as real-timeliness. Model Driven Development (MDD) is an approach that has emerged over the last two decades, and which is gaining in momentum. However, the adoption and the application of MDD is hindered by the challenging issues of scalability and formality. The focus of MDD, so far, has been on developing profiles, extending/specializing the Unified Modeling Language (UML) and languages for transformations, only a few techniques have been proposed for its realization. Moreover, these techniques do not span the necessary phases to link and relate the various artifacts and phases, do not scale or are not formal. The main objective of this research proposal is therefore to contribute to the MDD of complex software systems with techniques for their modeling, design and validation. We propose to investigate a modeling framework that covers several phases of the development process: requirements, design, and validation/testing. The modeling framework will consist of a subset of UML 2 and related profiles like MARTE, SysML and UML Profile for Testing (UPT). Using this formalized framework, we aim at devising MDD compliant techniques for the incremental design and validation of software systems. Incremental design approaches taking into account functional and non-functional characteristics as well as incremental test case generation techniques will be devised. These formal incremental techniques will certainly scale and will contribute to the application and adoption of MDD. This proposal aims at devising techniques for improving the quality of software systems and at reducing their development cost. The research issues tackled in this proposal are of high importance for the software research community and industry which includes the telecommunications and automobile sectors. The proposed research program will train graduate students who will join the Canadian work force and contribute to the Canadian economy in different sectors.