Development on Open-RAN Simulator with 5G-LENA

Open radio access network (O-RAN) has gained great attention owing to openness and intelligent management. To orchestrate and optimize mobile access networks with artificial intelligence (AI) and machine learning (ML) techniques, the O-RAN community standardized the RAN intelligent controllers (RICs). Owing to RICs, the mobile network operator can obtain more optimization solutions for classical use cases (e.g., ondemand traffic steering, anomaly detection, Quality-of-Service (QoS) management, and base station on-off policy). To intelligently control the network, RICs need to collect the data to make and verify the intelligent application. Thus, the O-RAN simulator is necessary for obtaining RAN data collection and testing intelligent applications in RICs. In this work, we develop a network simulation-3 (ns-3)-based novel O-RAN simulator by benchmarking the ns-O-RAN project. A real-world near-real-time RIC and ns3-simulated 5G NR gNBs are integrated through the E2 interface by exchanging E2 messages, RAN key performance measurements (KPMs), and RIC Control Messages. We implement the novel O-RAN simulator with ns-3 and present the experimental validation results.