Software reliability growth model considering power-law testing effort using optimization on hybrid PSO-GWO algorithm
Anup Kumar Behera, Priyanka Agrawal · 2024
Software reliability is a crucial aspect when it comes to assessing the quality of software. To ensure that software is free of errors before its release in the market, it undergoes thorough testing multiple times. Software reliability growth models (SRGMs) are valuable tools for software developers and testers as they help in analyzing stochastic failures and ensure software quality. Over the past four decades, several SRGMs have been proposed, but there is a constant need for upgraded versions in the software industry and research organizations. To address this need, it is essential to incorporate more updated testing effort functions and various fault detection rates into SRGMs to achieve more suitable software than ever before. This study presents a novel SRGM incorporating a power-law testing effort function (PL-TEF) and a non-linear fault detection rate (FDR), which are more practical in a realistic context. This research discusses a model that is compared by optimization techniques, using hybrid PSO-GWO algorithms conducted using a real dataset of failures to validate the model. The findings of the study reveal that the suggested model is the best fit for the data, and the ideal release time for the proposed model is also determined. These outcomes provide favorable results for software developers, offering valuable insights for decision-making processes.