I am a Ph.D. candidate at the Mobile Communication Laboratory (MCL), Kyung Hee University, South Korea, and a member of the Linux Foundation Aether SD-RAN Engineering Team.
Over the past 3+ years, we have developed an end-to-end O-RAN research platform integrating commercial UEs (Google Pixel 7a and Samsung Galaxy J5), OpenAirInterface (OAI), OCUDU, SD-RAN RIC, SMO, and Smart5G Non-RT RIC (a customized version of the O-RAN SC Non-RT RIC).
We are now extending this platform with NVIDIA Aerial to explore AI-RAN research and contribute to the AI-RAN community. In particular, we are interested in developing projects that combine NVIDIA Aerial’s L1/L2 capabilities with Aether’s cloud-native RIC platform.
As we are just beginning to build an NVIDIA Aerial-based testbed, we would greatly appreciate any guidance, suggestions, or lessons learned from the community.
We look forward to collaborating with Aerial community and sharing our experiences.
Thank you for the warm welcome!
We’ll definitely work through the NVIDIA AI Aerial documentation you shared. We’re also planning to build an NVIDIA Aerial testbed with DGX Spark, following the Aerial testbed hardware procurement guide ( Part 1. Procure the Hardware — Aerial Testbed ).
We really appreciate the resources and support from the Aerial community, and we’ve already learned a lot.
We look forward to sharing our progress and experiences as we integrate NVIDIA Aerial into our O-RAN testbed!
