Hardware node access for multi-UE real-radio experiment
Yoon Jae Hwang
Hi Powder Team!
I'm a researcher working on a 5G resource allocation project using the POWDER testbed. I'm currently running a single-node experiment with the ORAN profile (srsRAN Project gNB + FlexRIC nearRT-RIC + Open5GS) using ZMQ for RF simulation.
I'd like to extend this to a real over-the-air experiment with multiple UEs at different physical locations in order to study per-UE CQI variation over time. Specifically, I need:
- 1 compute node for the gNB, FlexRIC RIC, Open5GS 5GC, and an xApp (d430 or d740 is fine here as long as a USRP can be attached)
- 3 UE nodes, each with a USRP B210 (or equivalent) for real radio transmission — ideally nuc5300 nodes or fixed SDR endpoints at different physical locations
When I instantiate the ORAN experimental profile, the only available node type shown is d430 (compute-only, no radio hardware). I have a few questions:
1. Which POWDER profile and node type combination gives access to nodes with USRPs attached (e.g., nuc5300 or fixed indoor/rooftop SDR endpoints)?
2. Is there a way to request radio hardware nodes alongside a compute node within the same experiment topology?
3. Are there specific times or procedures for requesting fixed SDR endpoint allocations?
Our software stack is srsRAN Project (br-flexric compatible), FlexRIC (br-flexric branch), and Open5GS, all built from source. The UEs would run srsRAN_4G srsUE with UHD driver against USRP B210 hardware.
Thank you for your help.
Best regards,
Yoonjae Hwang
Dustin Maas
Hi Yoonjae,
I think if you want to move to real RF, you will need to use COTS UEs, and the best next step is probably the srs-rf-matrix profile (which actually deploys OCUDU now) and sets up a conducted RF environment with two COTS Quectel UEs. You can test your deployment there and then we can figure out how to handle the outdoor deployment, which is much more complicated!
Best,
Dustin
--
You received this message because you are subscribed to the Google Groups "Powder Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to powder-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/powder-users/ed01f660-4e61-4ba1-abdc-ead4a8653b6fn%40googlegroups.com.