Artificial Intelligence for 3-D Satellite Networks: A Comprehensive Overview

Day-by-day, wireless devices and diverse radio services require extended spectrum. Additionally, sixth generation (6G) mobile communication systems that are introducing new challenging use cases, need deep focus on latencies, number of connections, and so on. Hence, the combination of advanced design and working methodologies leads to enhance the necessary infrastructure, that in some cases can incorporate satellite networks as well. Artificial intelligence (AI) technique has been applied for various electromagnetic devices aiming to accelerate their overall design and analysis. In recent years, this paradigm has brightened the way of their application for space applications as well. In this framework, the present paper discusses a review on machine learning techniques employed toward design of 3-D satellite networks where it provides the use cases, requirements and enablers for these networks. Most importantly, the AI technology is employed for estimating the dynamic radio channel, signal detection and demodulation, network security, predicting the microwave signal attenuation, recognizing appropriate beam hopping patterns, and also for increasing efficiency in the places with the dust and sand storms. By preparing this paper, the authors has targeted to clarified that developing state-of-the-art methods including AI techniques would be a fundamental step towards the development of high-dimensional satellite network systems.