Review of terrestrial and satellite networks based on machine learning techniques

It has been broadly admitted that the upcoming networks will require to provide meaningfully more capacity than the existing ones to be able to deal with the growing traffic requirements of the users. Specifically, in the areas that optical fibers are improbable to be spread out because of the economical limitations; so, this would be a very crucial challenge. To address the above-mentioned issue, the combination of Terrestrial and Satellite Networks (TSNs) together would be an option. Satellite networks can cover enormous regions and current improvements have significantly raised the existing capacity while diminishing the cost. However, the characteristics of the geostationary satellite links are potentially various than the frequent terrestrial ones, essentially because the propagation time of the signal is high. The current study reviews the cutting-edge issues with respects to TSNs with machine learning methods.