An automatic approach to find the most suitable configurations for software systems test

Software testing is one of the software improvement phases and as the software sensitivity is being more, it is requested more time and effort. The fuzz testing approach is created for continuing the software test process in the development environment. Configuration Fuzzing is an approach that randomly makes different configurations and tests them by using InVivo which is one of the fuzz testing implementations and is monitoring security breaches in them. Configuration Fuzzing is not able to find optimal configurations for testing due to random search in configurations in huge software. In this article, we are presenting an approach to select a more suitable configuration for examination by using fuzzy logic. This approach attends to three main parameters as fuzzy system inputs for each configuration which include security risk, test cost, and error existence possibility. Finally, the probability parameter of the human error rate is attended as a coefficient in each configuration. At last, a number between ۰ and ۱ is attributed as test priority in the fuzzy system output of each configuration as this number is bigger, its test priority will be higher.

Keywords:

software testing , fuzz testing , configuration fuzzig , human error rate , CCSS