The infeasibility of experimental quantification of life-critical software reliability

摘要: This paper affirms that quantification of life-critical software reliability is infeasible using statistical methods whether applied to standard or fault-tolerant software. The key assumption fault tolerance|separately programmed versions fail independently|is shown be problematic. cannot justified by experimentation in the ultra-reliability region and subjective arguments its favor are not sufficiently strong justify it as an axiom. Also, implications recent multi-version experiments support this affirmation.