Wireless Network Selection System Using Federated Learning Considering Applications in Use

In recent times, the spread of mobile terminals and the escalating data communication volume have engendered connectivity challenges in densely populated locales such as urban stations. Previous studies have approached this issue by conceptualizing it as a misalignment between base stations and terminals, thereby proposing a matching approach base on graph theory. However, even after mitigating the mismatches, this solution remains computationally intensive and may not be conducive to practical implementation. This research presents a novel approach leveraging federated learning and introduce a meticulously designed system that ensures both precision and expeditious processing. As an initial phase, we conducted simulations employing machine learning techniques, yielding credible and substantiated results.