Fault Model-Based Testing

本文的研究重点是基于有限状态机(FSM)和输入输出转换系统(IoTS)模型的基于模型的测试，包括它们的扩展。它倡导基于故障模型的测试，这引起了研究人员和测试实践者不断增长的兴趣。造成这种情况的原因之一是，程序/规范覆盖范围不能为故障检测提供任何保证。这里可以提到Dijkstra的格言：测试可以用来显示错误的存在，但永远不能用来显示它们的不存在。这项拟议的研究旨在找到一个问题的答案：测试至少能保证没有留下某些类型的错误吗？那么，挑战就是如何对故障进行建模。我们将故障模型定义为规范、故障域和一致性关系的元组。作为给定不确定机器的所有可能子机的集合的故障域，称为给定规范机器的变异机。突变机器包含该规范，并使用大量模拟潜在故障的突变转换对其进行扩展。在我们最近的工作中，我们开发了一种基于逻辑编码和SMT求解的测试生成方法。以这种方法为起点，我们计划研究一个通用的基于故障模型的测试框架来处理各种面向状态的模型，从经典的FSM和IoTS模型到它们的符号扩展，达到在关键应用的模型驱动开发中使用的面向状态模型的扩展级别，例如Simulink Stateflow和UML状态机以及活动图和序列图。这个项目的目标是开发可扩展的测试方法(由原型工具支持)，用于开发测试，并保证领域专家根据产品开发历史、公司的细节及其测试假设选择的故障模型的覆盖范围。

The proposed research focuses on a model-based testing from Finite State Machine (FSM) and Input/Output Transition System (IOTS) models, including their extensions. It advocates fault model-based testing, which draws constantly growing interests of both, researchers and test practitioners. One of the reasons for this is the fact that the program/specification coverage does not provide any guarantee for fault detection. The dictum of Dijkstra could be mentioned here: testing can be used to show the presence of bugs, but never to show their absence. The proposed research is aimed at finding an answer to the question: can testing at least guarantee that no faults of certain types are left? The challenge then is how faults should be modeled. We define a fault model as a tuple of a specification, fault domain, and conformance relation. A fault domain as a set of all possible submachines of a given nondeterministic machine, called a mutation machine of a given specification machine. The mutation machine contains the specification and extends it with a number of mutated transitions modelling potential faults. In our recent work, we developed an approach for test generation which is based on logical encoding and SMT-solving. Considering this approach as a starting point, we plan to investigate a general fault model-based testing framework to deal with a variety of state-oriented models, from classical FSM and IOTS models to their symbolic extensions reaching the level of extensions found in state-oriented models used in model-driven development of critical applications, such as Simulink Stateflow and UML state machines as well as activity and sequence diagrams. The goal of this project is to develop scalable testing approaches (supported by prototype tools) for developing tests with a coverage guarantee for fault models chosen by the domain experts based on the history of a product development, company's specifics and their test assumptions.