Embedded AI for 6G NTN Applications

NTN and AI have been among the most rapidly developing enablers of the envisaged 6G communication systems. In this talk, we identify promising applications that fall within their intersection, such as onboard and mobile edge processing. The main challenge in such use cases is the size, weight and power limitations, which force us to adopt embedded AI platforms. In this context, we provide an overview of research results from the SIGCOM group, capturing conventional neural network architectures, as well as neuromorphic concepts. The talk concludes by listing the grand research challenges in this domain.