Open Radio Access Network

Open Radio Access Network research examines how open interfaces and disaggregated components can make radio networks more programmable, interoperable, and efficient. A central focus is the RAN Intelligent Controller, where xApps and rApps use methods such as deep reinforcement learning, predictive models, federated learning, and digital twins to manage resources, slices, handovers, energy, and faults. Newer approaches add large language models and coordinated agents to translate operator intent into control actions, while work on anomaly detection and defenses addresses unreliable telemetry and malicious applications. A key challenge is making these adaptive systems safe, stable, and responsive on real networks while ensuring they generalize beyond simulations and remain compatible with open interfaces.