Open AccessApproach for parameter estimation in Markov model of software reliability for early prediction: a case study
Author(s) -
Singh Lalit K.,
Vinod Gopika,
Tripathi Anil K.
Publication year - 2015
Publication title -
iet software
Language(s) - English
Resource type - Journals
ISSN - 1751-8814
DOI - 10.1049/iet-sen.2014.0108
Subject(s) - reliability engineering , markov chain , reliability (semiconductor) , computer science , software quality , markov model , software , markov process , software reliability testing , software development , machine learning , power (physics) , engineering , mathematics , statistics , programming language , physics , quantum mechanics
Early prediction of software reliability may be used to evaluate design feasibility, compare design alternatives, identify potential failure areas, trade‐off system design factors, track reliability improvements, identify the cost overrun at an early stage and to provide optimal development strategies. Many researchers have proposed different approaches to predict the software reliability based on Markov model but the uncertainty associated with these approaches is to find the transition probabilities in between the two states of the Markov chain. The authors propose an approach to address this problem by modelling the software system through Petri Net, converting it into Markov chain and solving the linear system mathematically. The validation of the proposed approach has also been shown by comparing the predicted reliability, based on predicted transition probability, with computed reliability, based on operational profile of safety critical software of Nuclear Power Plant.