The Evolution of AI and the Future of Satellite Communications

II SSPI Day de 2023

30/11/2023

This presentation, delivered by Waldo Russo at the II SSPI Day in 2023, explores how Artificial Intelligence is transforming the satellite communications sector. The work starts from the premise that, in the face of increasingly complex systems—multi-orbit hybrid satellites, NTN networks integrated with terrestrial networks, LEO constellations with thousands of satellites, and enabling technologies such as SD-WAN, NFV, and network slicing—the use of AI has become fundamental.

The presentation begins by reviewing basic AI concepts, distinguishing between Machine Learning, Deep Learning, and neural networks, as well as classifying the evolutionary stages of artificial intelligence (ANI, AGI, and ASI) and types of learning (supervised, unsupervised, and reinforcement learning).

Next, practical applications of AI in SATCOM are detailed, such as broadband optimization, predictive maintenance, antenna design, optical communications, massive MIMO, cognitive radios, satellite caching, and federated learning.

Finally, it discusses the impacts on the sector—greater efficiency, cost reduction, and new services—and addresses technical, ethical, and legal challenges related to data privacy, security, and reliability. The core message is that AI does not replace physics or humans, but amplifies their capabilities.

The presentation slides can be obtained by clicking here (attached PDF file).