This is the dataset for the paper M. Polese, L. Bonati, S. D’Oro, S. Basagni, T. Melodia, “ColoRAN: Design and Testing of Open RAN Intelligence on Large-scale Experimental Platforms,” IEEE Transactions on Mobile Computing, pp. 1–14, July 2022. The dataset can be found at this link.
We perform data-driven closed-loop control implemented using the ColO-RAN RIC and the SCOPE framework on the Colosseum wireless network emulator. We demonstrate the feasibility of a closed-control loop where DRL agents running in xApps on the near-real-time RIC select the best-performing scheduling policy for each RAN slice.
We emulate a 5G network with 7 base stations and 42 users in the dense urban scenario of Rome, Italy. The locations of the base stations have been extracted from OpenCelliD (a database of real-world cellular deployments) and cover an area of 0.11 km2. We consider a multi-slice scenario in which the users are statically assigned to a slice of the network and request three different traffic types: high-capacity enhanced mobile broadband (eMBB), ultra-reliable low-latency communications (URLLC), and machine-type communications (MTC). This reflects the case, for instance, of telecom operators providing different levels of service to different devices (e.g., MTC service to IoT-enabled devices, or URLLC to devices for time-critical applications). The base stations serve each slice with a dedicated (and possibly different) scheduling policy, selecting among proportionally fair (PF), waterfilling (WF), and round-robin (RR). We also consider the case where the number of physical resource blocks (PRBs) allocated to each slice varies over time.
The dataset can be found at this link.
If you use this dataset, 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]