Generating test cases using metamorphic testing and genetic algorithm for integer bugs detection

Abstract

In Software Testing, the oracle problem is well understood i.e. the test oracle is either not accessible or if it exists, it is troublesome to apply. Testing software using special values don’t guarantee the correctness of the software. Metamorphic Testing is used to alleviate the oracle problem by developing follow up test cases from initial test cases and software is tested using random values. This testing is very much effective in detecting bugs because by using certain properties of the objective function we develop some metamorphic relations and test the software with follow up test cases against these metamorphic relations. Also in case of testing, it is very important that how effectively the test cases are generated which will test the software using all the execution paths. So to increase the number of test cases, the operators of Genetic algorithm has been applied on the output of Metamorphic Testing to develop more profound test cases which will ensure the correctness of the software under consideration to the maximum extent.