Introducing Research with NVIDIA Aerial Omniverse Digital Twin
The IIP Group is now integrating NVIDIA Aerial Omniverse Digital Twin as synthetic channel model in ongoing research activities.
The Integrated Information Processing (IIP) Group is now using channel-model data obtained with NVIDIA Aerial Omniverse Digital Twin in our machine learning (ML) research projects. Aerial Omniverse Digital Twin is an operator-scale digital twin for system-level simulation. Built on NVIDIA Omniverse, Aerial Omniverse Digital Twin scales to hundreds of base stations and thousands of user equipments. The framework is implemented using the full 5G NR physical layer from NVIDIA combined with RTX-accelerated ray-traced channels and can be used for the generation of spatially coherent channel model datasets for training in Sionna, PyTorch, or other ML frameworks. Our initial research is focusing on wireless positioning and channel charting using channel model vectors obtained through Aerial Omniverse Digital Twin.
Please check out our blog post for more details.