Granular Traceability Between Requirements and Test Cases for Safety-Critical Software Systems

Traceability is mandatory in developing safety-critical systems as prescribed by safety guidelines, such as DO-178C, and it is vital for avionics industries. Testing is mandatory for requirement validation to ensure the safety and quality of a software product. Requirement traceability all along the development cycle is essential. Requirements’ traceability and test generation fields have been studied extensively. This paper presents a granular traceability approach between low level requirements (LLRs) and test cases that is supported by a model-based test case generation. From LLR specifications we use model-to-model transformation to obtain an extended finite state machine (EFSM) and its corresponding control flow graph (CFG) and data flow graphs (DFGs) that can support various granularity levels of traceability. The uniqueness of the proposed traceability approach is the creation of traceability elements at finer granularity during test case generation that satisfies MC/DC and Du path coverage criteria, and their retrieval for coverage analysis. The granularity level reached in this work corresponds to decision nodes, edges, predicates, and variables that are located on their source artifacts (CFG, DFG, EFSM). Several traceability techniques are used such as IDs and Links to create and retrieve traceability elements, coverage elements on models, transformation records, and artifacts that ensure both the forward traceability of requirements by construction and backward traceability using graph exploration techniques.