| Along with the development of computer software technology, software products have penetrated into every field of People's Daily life. In aviation, aerospace, medical and other fields, once invalidation happens to software,it will cause huge economic loss or lead people into death. In daily life, the invalidation of software would cause inconvenience to people's everyday life. In order to safeguard the software's reliability, we need to do reliability testing to get its mathematical index, and measure whether the reliability index meets user's requirement.The study of reliability testing based on Markov chain usage model mainly focuses on two aspects: first, how to construct Markov chain model to make the model accurately show the actual usage of software; second, the study of testing method based on Markov chain usage model, that refers to test cases generation and testing adequacy determination. In this thesis, we study the testing method and suppose that Markov chain usage model of software has been obtained.The primary contents are as follows:First introduce software testing and reliability in detail, and then introduce the software reliability testing. According to the target of software reliability testing, this thesis analyses its characters deeply, final explicitly states that the concept of software reliability testing.Second, this thesis analyses the methods of reliability testing based on Markov chain usage model from two aspects: test cases generation and testing adequacy determination, then point out their shortcomings. After that, it analyses the key causes of methods lack, and propose the concept of 2-step Markov chain usage model. According to the 2-step Markov chain usage model, this thesis proposes new methods of test cases generation and testing adequacy determination.At last, the thesis realizes an accessorial tool to generate the test case automatically. And through the results data from this tool, the methods proposed in front have been verified. |
PDF Full Text Request