In this paper, we present a software reliability apportionment scheme using analytic hierarchy process which attempts to answer the question " how reliable should each system module be ? ". Reliability requirements determined by integrating user's view, software manager's view and programmer's view would be more realistic, consistent and economically attainable than those obtained through subjective or haphazard method. This model determines reliability goals at the planning and design stages of the software project, hence making reliability a singular measure for performance evaluation. The concept of frequency ratio is used which considers the probability of a module execution when the software is executed for a predefined number of input cases. Modules which are to be executed rather infrequently may be allocated relatively lower reliability values.
No takes yet. Share an insight, caveat, or question.
Aggarwal et al. (1995) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: