Issue No. 03 - March (1996 vol. 22)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/32.489081
<p><b>Abstract</b>—In this work a binary Markov process model is proposed for random testing for software. This model is suggested to replace the standard binomial distribution model, which is based on the easily-violated assumption of test runs being statistically independent of each other. In addition to a general result on the probability of having any specific number of software failures during testing, practical implications of the new model are also discussed. In particular, we demonstrate that in general the effect of possible correlation between test runs cannot be ignored in estimating software reliability.</p>
Binary Markov process, dependent test runs, random testing, software reliability, statistical testing, ultra-reliability application.
S. Mills and S. Chen, "A Binary Markov Process Model for Random Testing," in IEEE Transactions on Software Engineering, vol. 22, no. , pp. 218-223, 1996.