Open AI Cellular (OAIC): An Open Source 5G O-RAN Testbed for Design and Testing of AI-based RAN Management Algorithms

Open Radio Access Network (O-RAN), defined by the O-RAN Alliance, is quickly emerging as one of the most critical RAN architectures for designing and deploying future cellular networks. O-RAN aims to use openness and intelligence to solve new research problems that arise from the broad set of services offered by 5G. However, the lack of end-to-end, stable, open-source platforms offering realistic environments has limited progress in developing effective solutions. This paper describes an open-source 3GPP-compliant O-RAN-based 5G testbed being developed at Virginia Tech called Open AI Cellular (OAIC). The OAIC testbed includes the OAIC-Control framework to facilitate the design of AI-based RAN management algorithms and the OAIC-Testing framework to facilitate automated testing of AI-based RAN management algorithms. While O-RAN defines non-Real Time RAN Intelligent Controller (RIC) and near-Real Time RIC, the OAIC testbed introduces a new Real-Time RIC, zApps, and Z1 interface to support new use cases that require latency of less than 10 ms. The current and target capabilities of the OAIC testbed, along with the challenges within the O-RAN framework, are discussed. An example use case (Key Performance Metrics) that utilizes the OAIC testbed is also narrated. The OAIC testbed integrated with CORNET infrastructure is remotely accessible via a portal and is expected to accelerate R&D in academia and the wireless industry.