Intelligent Architectures for Open Radio Access Networks

Intelligent architectures for Open Radio Access Networks (O-RAN) represent a paradigm shift in mobile network design, moving from closed, monolithic systems to modular, vendor-agnostic frameworks. By disaggregating radio, distributed and centralised units, and exposing standardised north- and south-bound interfaces, O-RAN enables multi-vendor interoperability and programmability. Central to this evolution is the RAN Intelligent Controller (RIC), which hosts third-party applications (xApps and rApps) to implement data-driven, closed-loop control for functions such as traffic steering, energy management and fault remediation. Artificial intelligence and machine learning workflows are embedded at various timescales—near-real-time for radio resource allocation and non-real-time for model training and policy updates—ensuring continual optimisation of network performance. Open architectures foster innovation by lowering entry barriers for new vendors and research communities, with practical applications already visible in early 5G and proof-of-concept deployments. Global operators recognise O-RAN’s potential to reduce costs, accelerate service rollout and enhance sustainability, laying the foundation for future 6G systems.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Intelligent Architectures for Open Radio Access Networks publication trend

The graph below shows the total number of articles in intelligent architectures for open radio access networks across all publications each year (not limited to Nature Index journals).

Technical terms

Open Radio Access Network (O-RAN): A disaggregated RAN architecture with standardised interfaces enabling multi-vendor interoperability and software-based control.

RAN Intelligent Controller (RIC): A centralised software platform that hosts third-party applications (xApps and rApps) to perform intelligent control and optimisation of RAN functions.

xApp: A modular, near-real-time application deployed on the RIC, implementing AI/ML algorithms for tasks such as load balancing and interference mitigation.

Network slicing: The creation of multiple logical networks over a shared physical infrastructure, each tailored to specific service requirements (eMBB, URLLC, etc.).

Disaggregation: The separation of hardware and software components in network elements, allowing independent development, deployment and scaling.

Reinforcement learning: A machine learning paradigm in which agents learn optimal actions through trial-and-error interactions with the environment, guided by reward signals.

References

1. Understanding O-RAN: Architecture, Interfaces, Algorithms, Security, and Research Challenges. IEEE Communications Surveys & Tutorials (2023).
2. Delay-sensitive resource allocation for IoT systems in 5G O-RAN networks. Internet of Things (2024).
3. AI/ML Enabled Automation System for Software Defined Disaggregated Open Radio Access Networks: Transforming Telecommunication Business. Big Data Mining and Analytics (2024).