Challenges and Machine Learning Solutions for Optical Communications in Space-Air- Ground Integrated Networks for 6G

• Published in: IEEE wireless communications pp. 1 - 8
• Main Authors: Ariyoshi, Masayuki, Funada, Junichi, Gabory, Emmanuel Le Taillandier de, Asai, Shigeru, Rodrigues, Tiago Koketsu, Kawamoto, Yuichi, Kato, Nei
• Format: Journal Article
• Language: English
• Published: IEEE 15.07.2024
• Subjects: 6G mobile communication; Optical attenuators; Optical diffraction; Optical distortion; Optical fiber networks; Satellite broadcasting; Satellites
• DOI: 10.1109/MWC.015.2300598
• ISSN: 1536-1284

Future networks, including Beyond 5G and even 6G, have ambitious requirements where all devices must stay permanently within connection range and receive reliable service with low delay. To achieve this, high-speed links must be established even in remote areas, which is best done through satellite-air-ground integrated networks and their combination of satellites, high altitude platform stations, and ground base stations, which can reach all users, regardless of environment. To increase the speed and transmission rate of these systems, the use of optical communication links, at high frequency bands, has been studied. However, existing works focus on single-link scenarios due to the complexity of a system-level approach. While viable and simple to implement, these approaches are not the most efficient as they underrepresent the impact of interactions between devices. In this work, we will explain the challenges of this optical- based satellite-air-ground integrated network system and propose the use of machine learning and smart networking solutions to address these issues, including theoretical discussions and numerical estimates of the benefits possible with those same solutions. This article thus provides a foundation for global-level control of optical communications in satellite-air-ground integrated networks.