Advancing Intelligent RAN: inside WWT's AI Aerial Innovation Lab
In this blog
The shift to intelligent RAN: Unlocking performance, efficiency and new revenue
The evolution of 5G and the path toward 6G are driving a fundamental shift in how wireless networks are designed and operated. Service providers are under pressure to deliver higher performance, ultra-low latency, and new revenue-generating services while improving efficiency and controlling costs.
To meet these demands, operators are exploring AI-native, software-defined radio access network architectures that can scale, adapt and monetize more effectively. NVIDIA AI Aerial™ plays a critical role in this transformation.
NVIDIA AI Aerial is a suite of platforms, software libraries and tools to build, train, simulate and deploy AI-RAN networks. It enables:
- Software-defined and AI-native RAN deployments
- Converged AI and telecom workloads
- Cloud-scale orchestration and automation
- Enhanced network performance and operational efficiency
To support ecosystem adoption of AI-powered RAN architectures, World Wide Technology (WWT) has deployed an advanced AI Aerial Innovation Lab within its Advanced Technology Center (ATC). The lab provides a production-realistic environment where service providers, technology partners and developers can evaluate AI-native RAN architectures, validate multi-vendor integration and demonstrate high-value use cases
WWT AI Aerial lab architecture at a glance
The WWT AI Aerial Innovation Lab represents a multi-vendor, cloud-native 5G RAN environment built on GPU-accelerated infrastructure designed to support both telecom and AI workloads.
Here is the high-level lab Architecture:
At the core of the lab is the AI Aerial CUDA-based RAN platform powered by NVIDIA GH200 Grace Hopper™ Superchip GPUs and NVIDIA® BlueField® DPUs, leveraging the NVIDIA Spectrum™-X Ethernet Networking platform for high-performance fronthaul and backhaul connectivity. Server infrastructure is provided by Supermicro, delivering a scalable compute foundation optimized for accelerated workloads.
Infrastructure and workload orchestration is managed by Armada Bridge (formerly Aarna Networks), which provides automation, lifecycle management, multi-tenancy and operational visibility across the environment.
A critical element of the deployment is the 5G timing and synchronization layer delivered by Net Insight through its Zyntai timing nodes. These nodes provide highly accurate and redundant time synchronization using PTP and GNSS sources, helping maintain reliable network operation even in challenging or degraded conditions. By operating over existing IP/MPLS networks, the solution enhances resilience for mission-critical services.
The foundational platforms in the lab are currently operational, with infrastructure readiness testing using a radio unit emulator. In the next phase, the deployment will integrate OpenAirInterface (OAI) Core and RAN software with radios from Foxconn to establish full end-to-end 5G communication.
High-value use cases enabled by the lab
The WWT AI Aerial Innovation Lab focuses on four strategic use case categories aligned to operator priorities.
Use case #1: Orchestration and automation
Modern telecom networks require cloud-like agility. The lab demonstrates cloud-native infrastructure management capabilities, including automated workload provisioning, multi-tenancy, lifecycle automation and real-time observability. These capabilities enable faster service deployment while reducing operational complexity.
Use case #2: Resilient 5G timing
Accurate synchronization is essential for advanced 5G features such as massive MIMO and beamforming. The lab validates redundant timing mechanisms that maintain service continuity during GNSS disruptions, interference or signal loss, helping ensure reliability in real-world operating environments.
Use case #3: AI on RAN for revenue enablement
The lab showcases converged infrastructure where GPU resources can dynamically shift between RAN processing during peak demand and AI inference workloads during off-peak periods. This approach maximizes asset utilization while enabling new revenue opportunities at the edge.
Use case #4: AI for RAN optimization
AI-driven analytics are applied to RAN and core network KPIs to improve performance, efficiency and reliability. The platform enables proactive traffic forecasting, energy optimization, fault detection and automated performance tuning while providing insights into feature utilization to maximize return on investment.
Advanced capabilities such as spectral efficiency optimization will be introduced as xApps, rApps and dApps are integrated into the environment. Together, these capabilities move networks toward autonomous, self-optimizing operations with minimal manual intervention.
Conclusion
The AI Aerial Innovation Lab enables service providers to accelerate network transformation by reducing deployment risk, improving infrastructure efficiency and enhancing operational resilience through AI-native architectures and advanced timing.
By combining accelerated computing from NVIDIA, cloud-native orchestration, resilient synchronization and open RAN integration, World Wide Technology (WWT) delivers a practical blueprint for intelligent, software-defined and revenue-optimized wireless networks that are ready for 5G today and 6G tomorrow.