Software Test Case Generation Based on the Fault Propagation Path Coverage

In the process of software testing, correlated defects raise researchers' attention worldwide. Some potential defects are hard to be detected in the test. To address these potential defects, this paper adopts an error propagation model to describe the process of defect evolution and applies fault injection method to introduce known seed-defects. Subsequently, seed-defects are activated and related potential defects are induced through test case design. This method employs intelligent algorithms to constantly design test cases to cover paths of seed-defects and paths of propagation. As a result, more undetected and correlated defects can be detected in paths of propagation.