Software reliability prediction using SPN

Abstract

Reliability is an important software quality parameter. In this research for computation of software reliability, component reliability model based on SPN would be proposed. An isomorphic markov chain is obtained from component SPN model. A quantitative reliability prediction method is proposed. The component reliability value is calculated according to the transition cumulative probability distribution of markov chain, obtained from the software SPN model. By means of reliability prediction of the whole software, we'll introduce CRMPN. In CRMPN states are component reliability model and transition are marked with components reliability. With this research more complex software could be simplified and reliability of the software could be evaluated effectively. An example is provided for demonstrating the feasibility and applicability of our method.

Keywords: Reliability, SPN, Markov Chain, Component based-software