An efficient method for parameter estimation of software reliability growth model using artificial bee colony optimization

Abstract

One of the established trends of research areas and practices in software engineering that dealt with the measurement and enhancement of reliability is software reliability engineering. Stochastic software reliability models find typical usage for analysis. These are the models that perform modeling on failure process of the software and exploit other software metrics or failure data as cornerstone for parameter estimation. The ability of the models for estimating and predicting the current reliability and future failure behavior, respectively, is high. Due to any failures and faults in the system, a product becomes unreliable. The lack of understanding of nature of the software makes the measurement of software reliability as a challenging task. It is not possible to determine a best way to measure the reliability and other aspects of software. This paper proposes an efficient software reliability growth model (SRGM) in which logistic exponential TEF is exploited. This model offers increased failure rate recognition and suitable ways to resolve faults and so on. Our work estimates the SRGM parameters using optimization algorithm. Such estimation can aid in developing precise software reliability model. In order to accomplish the optimization, we use artificial bee colony (ABC) algorithm. As the parameters optimization considerably improves the quality of parameters to be used for reliability growth model, reliability growth can also be improved considerably.