XAI for Maintainability Prediction of Software-Defined Networks
Gokul Yenduri, Thippa Reddy Gadekallu · 2023
Software-defined networking (SDN) refers to a method of computer networking that uses software abstractions rather than specific hardware. Network administrators may manage dynamic networks more efficiently by abstracting parts of the network’s low-level functions into a software program. If the maintainability of the SDN software is not taken into account, it may lead to bugs, security problems, and maybe the need for functional changes. As a consequence, the whole network needs to be reconfigured, which would raise the total cost of maintaining SDN. Additionally, to use the new reconfigured SDN properly, personnel must get training. As a result, it is crucial to take software maintainability into account while creating SDN. Maintainability is the ability of code or software to adapt to a change. In this study, we use artificial neural networks to predict the software maintainability of the dataset related to the interface management system and compared its performance over multivariate adaptive regression splines, step-wise regression, and support vector machines. The proposed model’s findings are interpreted using Explainable Artificial Intelligence.