The existence of software faults in safety-critical systems is not tolerable. Goals of software reliability assessment are estimating the failure probability of the program, /spl theta/, and gaining statistical confidence that /spl theta/ is realistic. While in most cases reliability assessment is performed prior to the deployment of the system, there are circumstances when reliability assessment is needed in the process of (re)evaluation of the fielded (deployed) system. Post deployment reliability assessment provides reassurance that the expected dependability characteristics of the system have been achieved. It may be used as a basis of the recommendation for maintenance and further improvement, or the recommendation to discontinue the use of the system. The paper presents practical problems and challenges encountered in an effort to assess and quantify software reliability of NASA's Day-of-Launch I-Load Update (DOLILU II) system DOLILU II system has been in operational use for several years. A Bayesian framework is chosen for reliability assessment, because it allows incorporation of (in this specific case failure free) program executions observed in the operational environment. Furthermore, we outline the development of a probabilistic framework that allows accounting of rigorous verification and validation activities performed prior to a system's deployment into the reliability assessment.
展开▼
