Enhancing Software Reliability Through Hybrid Metaheuristic Optimization

Several software dependability models have been designed and developed over the last three to four decades. They were crucial in ensuring the dependability of the software. Researchers have identified a number of software dependability metrics, and performance behavior is displayed in accordance with the software metrics. For their suggested software reliability model, the researchers used a number of different methodologies, all of which are effective in their respective fields. The majority of models have the drawback of being limited to a single domain and not performing better in other contexts. Many of them are based on assumption. A strong software reliability model can help in this regard, though. The selection of an adequate failure dataset and selection are crucial for this activity. A metaheuristic approach can be fruitful to find the well suited parameter estimation of the chosen model that can lead to better results. It is critical to select an adequate failure dataset and a strong software reliability model for this endeavor. Finding the parameter estimation of the selected model may be successful using a hybrid metaheuristic technique. This piece of work takes into account several failure datasets and a traditional Goel-Okumoto (G-O) model. The parameters of the G-O model have been calculated using a hybrid metaheuristic approach. In order to reduce the relative error (RE) and determine the goodness of fit, the G-O model’s Mean Value Function (MVF) is used. The performance of a few software measures has been determined, and it has been contrasted with those of current metaheuristic techniques. The suggested method outperforms the present metaheuristic methods in terms of performance. With the aid of MATLAB, the comparison analysis is carried out.