Full text
| |
| Author(s): |
Adilson Luiz Bonifácio
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | Campinas, SP. |
| Institution: | Universidade Estadual de Campinas (UNICAMP). Instituto de Computação |
| Defense date: | 2009-12-16 |
| Examining board members: |
Arnaldo Vieira Moura;
Ana Cristina Vieira de Melo;
Jose Maria Parente de Oliveira;
Luiz Eduardo Busato;
Eliane Martins
|
| Advisor: | Arnaldo Vieira Moura |
| Abstract | |
This work aims to provide efficient test case generation methods for reactive and critical systems. In general, reactive and critical systems are real-time systems with context transformations. One of the most promising techniques for generating test suites is model-based testing. The formalisms supply the basis to perform a precise and dependable testing activity. In this scenery, our work proposes a construction of formal models, methods and techniques, as well as testing strategies, to support the process of automatically generating test suites for complex systems. However, the test generation process using formal models is usually infeasible in real applications, due to the state space explosion. Therefore, we need to find out suitable models to capture the system behaviors, and also to construct methods that can overcome the explosion problem, in a reasonable way, allowing the generation of test suites for complex systems. In this work we treat the following formal models: the conventional FSM; an extension of FSM using context variables (EFSM); the proposed extension of EFSM (TEFSM) to capture context variables and also clock variables; timed models, with disassociated input and output actions, called TIOA; and the proposed extension for TIOA, so-called TIOCA, to capture continuous time evolution and context transformations. In a first step of this work we proposed a technique to derive confirming sequences for TEFSM, using model-checking. Next, a classical method to generate complete test suites was generalized for FSM.We also proposed a new discretizationmethod for TIOA models, allowing the test case generation using test purpose and the synchronous product. Lastly, we extended the discretization method for TIOA to obtain more compact grid automata for TIOCA models, allowing the test case generation for systems with continuous time evolution and data flow transformations, using the notion of test purpose and the product of TIOCA (AU) | |