A large-scale O-RAN Testing Framework for AI/ML Development
ColO-RAN is a publicly-available, large-scale O-RAN testing framework with software-defined radios-in-the-loop. Building on the scale and computational capabilities of the Colosseum wireless network emulator, ColO-RAN enables ML research at scale using O-RAN components, programmable base stations, and a “wireless data factory”. ColO-RAN features a near real time RIC (based on the O-RAN Software Community RIC), deployed on Colosseum, that controls a softwarized RAN (based on our SCOPE framework) with a standard-compliant E2 termination.
It also features an SDK for the swift integration and testing of AI/ML-based xApps for run-time RAN inference and/or control. At a high level, ColO-RAN xApps are made of two building blocks, shown below: (i) the Service Model (SM) connector, which handles the communications to/from the near-RT RIC (e.g., to communicate with the base stations), ASN.1 message encoding/decoding, and queries to the RIC Redis database, and (ii) the data-driven logic unit that performs tasks based on Key Performance Metrics (KPMs) received from the RAN at run time, e.g., traffic prediction and/or control of the base stations.
If you use ColO-RAN, please reference the following paper:
M. Polese, L. Bonati, S. D’Oro, S. Basagni, and T. Melodia, “ColO-RAN: Developing Machine Learning-based xApps for Open RAN Closed-loop Control on Programmable Experimental Platforms,” IEEE Transactions on Mobile Computing, July 2022. [pdf] [bibtex]
You can find more information about ColO-RAN at the following repositories: near-RT RIC and E2 termination. The ColO-RAN dataset can be found at this link.