Effects of software test efficiency on software reliability demonstration testing effort

To solve the problem that in the traditional software reliability demonstration testing(SRDT), software testability was not considered as one of the possible influence factors on test effort, a method which introduced test efficiency to improve the quantity model of minimal sample size in SRDT was put forward. The definition of software testability and its effect on software reliability test suite were analyzed. The concept of test efficiency was proposed. The difference between the numbers of test cases with different test efficiency and whether efficiency is considered were compared. It was demonstrated in detail that the traditional method and the Bayesian method with no prior knowledge are all the conditions where the test efficiency equaled to zero, so both the methods gave the most conservative value. The improved method combined with the test efficiency was recommended. Based on the principle of statistical fault injection, the quantitative estimation method of test efficiency was proposed and the process and the estimation method were shown by a case study.