Predicting the Frequency Bands and the Path Loss in Wireless Communication Systems using Random Forests

Abstract

Proactive and predictive design in the next wireless generation is critical to avoiding the flaws of prior generations and achieving the 5G goal services pillars. Base stations are needed to perform and make judgments to maintain communication dependability as wireless devices become more commonplace. Machine Learning (ML) is used in this research to help base stations anticipate the frequency bands and the route loss. There is a comparison between the ML algorithms Multilayer Perceptron and Random Forests. In order to keep up with the demands of the new radios, systems that use various bands need an immediate reaction from devices to change bands quickly. For this reason, ML approaches are required to learn and help a radio base station in shifting between multiple bands in response to data-driven decisions. Afterwards, the findings are compared to those of different deep learning approaches. To guarantee that the projected works would succeed, these strategies are used to several case studies. Unsupervised algorithms were added to the random forests in order to improve the accuracy of the learning process.