Based Software Testing Software Quality Assurance Study

As Computers' becoming more and more popular, software has been omnipresent in many fields of the society and in our daily life. Therefore, more and more attention has been paid to software quality and assuring software quality has become a crucial task in software engineering field. This thesis conducts in-depth research on SQA based on software testing.Starting with the definition of software quality, the dissertation firstly discusses the topic on how to evaluate the software quality scientifically. It reviews the roadmap of software quality evaluation systems (SQEM) and analyzes the advantages and disadvantages of main SQEMs then introduces the most popular international software quality standards. Considering the process technology and user satisfactory measurement, combining the methods and theories in software metrics subject, a new SQES which is a total quality measurement system based on three-layer model has been proposed by the dissertation.Then the thesis deals with the basic means to realize SQA - the software testing technology. It includes software inspection (static testing) and software testing in narrow sense (dynamic testing). In order to effectively arrange testing work and better assure the software quality, the scientific testing strategy need to be studied. The risk-based testing strategy is considered as a practical method to optimize limited resources' collocating and improve testing efficiency. Due to the shortages of present risk-based model, Software quality measurements and classic McCabe's cyclomatic complexity metrics have been introduced to the risk evaluation model, furthermore, the capture-recapture model has also been used to estimate the faults remaining in each function which will contribute to adjusting testing effort to make them focus on the high-risk functions.