A cloud Markov model for software reliability evaluation

摘要

There are two main types of fundamental uncertainties, randomness and fuzziness, in software reliability. Software module reliability and transition probabilities, which can't be determined with unerring precision in the general usage Markov chain, are expressed by virtual clouds. Then we present a novel approach based on cloud model theory and Markov process to software reliability evaluation. To solve the computational problem of cloud model, a simple statistical algorithm of algebra operations is put forward. The rationality of the novel approach is proved through comparative analysis with a notable example.