A requirements inspection method based on scenarios generated by model mutation and the experimental validation

The requirements phase is the most critical phase of the software development life cycle. The quality of the requirements specification affects the overall quality of the subsequent phases and hence, the software product. An effective and efficient method to qualify the software requirements specification (SRS) is necessary to ensure the reliability and safety of software. In this paper, a requirements inspection method based on scenarios generated by model mutation (RIMSM) is proposed to detect defects in the functional requirements of a safety-critical system. The RIMSM method models software requirements using a High Level Extended Finite State Machine (HLEFSM). A method that executes the HLEFSM model is defined. The method uncovers the behaviors and generates the outputs of the system for a given scenario. To identify an adequate set of scenarios in which the model shall be executed, an analogue to mutation testing is defined which applies to the requirements phase. Twenty-one mutation operators are designed based on a taxonomy of defects defined for the requirements phase. Mutants of the HLEFSM model are generated using these operators. Further, an algorithm is developed to identify scenarios that can kill the mutants. The set of scenarios is considered to be adequate for detecting defects in the model when all mutants generated are killed. The HLEFSM model is then executed for the scenarios generated. The results of execution are used to detect defects in the model. A Requirements Inspection Tool based on Scenarios Generated by Model Mutation (RITSM) is developed to automate the application of the RIMSM method. The performance and usability of the RIMSM method are studied and demonstrated in an experiment by comparing the RIMSM method to the checklist-based reading method.