A software reliability growth model considering adaptive fault detection, fault removal efficiency, change point, and fuzzy-based optimal release time
TL;DR
This study proposes an SRGM considering change point, fault removal efficiency, and fault detection rate, unlike traditional models, considers that the fault detection rate depends not only on the number of faults detected but also on the remaining number of faults that have not been detected.
Abstract
Smart appliances and gadgets are becoming essential parts of our daily lives. So, there is an increasing need for reliable software that runs them, as even a small error can cause user dissatisfaction and serious problems. Software reliability growth models (SRGM) are often used to measure software reliability. This study proposes an SRGM considering change point, fault removal efficiency, and fault detection rate. This model, unlike traditional models, considers that the fault detection rate depends not only on the number of faults detected but also on the remaining number of faults that have not been detected. A change point is also incorporated to represent sudden changes in the testing process. With a high level of reliable software, firms must consider the development cost and time-to-market constraints. An earlier release may introduce reliability issues, whereas a delayed release may bring missed market opportunities. Factors such as resource allocation, testing strategies, and the experience of the testing team may impact reliability as well as development cost. To address this, considering factors like debugging cost, warranty cost, etc., a fuzzy optimization framework is used to determine optimal release time.
