Reliability Analysis of Exponential Models Based on Skewness and Kurtosis

Abstract

Every field in modern era is computerized. As the requirements of software increase, competitions among the manufacturers of software also increase. Thus, there is a need for reliable software. Software reliability is defined as the probability of failure-free operation of software in a specified environment for a specific period of time. Thus, if T denotes the failure time of software, then, its reliability, denoted by R(t), is given by R(t) = P(T *gt; t). Various models of software reliability have been developed. One such model is the exponential class model. For such a model, the reliability function is given by R(t) = фe-фt, where ф is the failure rate. Various estimates of reliability have been obtained for this class of models. The most commonly used method is the method of Maximum Likelihood Estimation (MLE). But it is not as efficient as the Minimum Variance Unbiased Estimation (MVUE). In our previous work, we obtained this minimum variance unbiased estimator for the reliability function R(t) and proved its efficiency by comparing it with the Maximum likelihood estimator. We used variance as a measure of comparison. But variance is only a second order measure. In this paper, we are trying to enhance our work further by comparing higher order measures. We are also trying to analyze the same using skewness and kurtosis. © Springer India 2015.