Software Reliability Estimation of Schick-Wolverton Rayleigh Failure Time Model

Abstract

It is well known that the users and the developers of the software often encounter the problem of its quality and durability. Software reliability is a measure of the quality of the software. By estimating the reliability of the software, the developers can ensure that the reliability objectives as specified by the user are met. On the other hand, the reliability estimate also enables the users to decide whether or not to accept the software. Thus, reliability estimates are the key factors in decision making problems for both the developers and the users of the software. Software reliability models with specified failure time distributions are considered and failure data obtained during testing are used to estimate their reliability. Herein, Rayleigh failure time distribution of Schick-Wolverton is considered and reliability estimates have been obtained by the methods of Minimum Variance Unbiased Estimation (MVUE) and Maximum Likelihood Estimation (MLE). The two estimates have been compared. Case studies have also been considered and the above two estimates of reliability have been compared. It is observed that the estimator obtained using the method of minimum variance unbiased estimation provides a better estimator than that obtained using the method of maximum likelihood estimation. © 2021 Proceedings - 26th ISSAT International Conference on Reliability and Quality in Design, RQD 2021. All rights reserved.