Reliability assessment measures based on software reliability growth model with normalized method

It is very important to assess software reliability quantitatively by using failure time data observed during software testing. This paper discusses a method of assessing software reliability on the basis of a software reliability model described by a nonhomogeneous Poisson process, that is to say, an exponential reliability growth model. The mean time between software failures and the software reliability function are adopted as reliability assessment measures. Since the time-intervals between software failures for the model are improper, it is shown that the means of the distributions can be obtained from the normalized distributions. Numerical examples of the mean time between software failures and the software reliability function are given by applying the method proposed here to actual data.