Application of Software Reliability Growth Models for Aerospace System Development Open Access
Downloadable ContentDownload PDF
As airborne systems become more complex and highly-integrated, greater effort is required to demonstrate that systems operate as intended. The design and verification processes outlined in DO-178B and DO-254 are based on requirements. The system development practice recommended in ARP4754 expands on the requirements based approach of the other standards by introducing scenarios to define operational conditions. The software development process DO-178C incorporated the concept of robustness by expecting the executable code to respond correctly to abnormal inputs and conditions. The challenge for system developers is to define an appropriate test space for what constitutes abnormal inputs and then to plan for the required test effort.This study investigated the application of software reliability growth models, specifically to software-centric airborne systems. A modification of the Goel-Okumoto model is presented that defines β as the product of the probability a software test case will fail to detect an error and the probability a software change will remove the observed failure. This research found that software development in safety-critical applications may now suffer from an over-emphasis on process and fail to fully demonstrate the reliability of the software product. The proposed model is shown to improve the reliability of the product and reduce total life-cycle costs while still maintaining strict compliance with the certification development process.