Application of Markov chain approach for multi-attributes dynamic software reliability assessment under both AHP and gray correlation methods

Abstract

Measuring software trustworthiness is an important field in trusted software assessment. In this paper, a dynamic model of multi-attributes dynamic software reliability assessment is proposed based on analytic hierarchy process (AHP) and gray correlation, since current software assessing methods cannot fully meet the accuracy and objectivity of the assessment. Our novel model of software trustworthiness comprehensive evaluation index system is established, where AHP comprehensive evaluation method is introduced to determine the trustworthy indicators and the weight of the sub-attributes. Meanwhile, group expert judgments are used to modify group decision results and the Euclid distance is used to judge the value’s change of state checkpoints. In addition, the software properties probabilistic transformation model of each level of sub-properties is constructed to accomplish dynamic software reliability assessment using the methods of Markov chain process and gray correlation model. Finally, an example is used to illustrate the effectiveness of our dynamic software reliability assessment.