Northeastern University launches Fully Automated and Virtualized O-RAN Private 5G Network with AI Automation
The Institute for the Wireless Internet of Things (WIoT) at Northeastern University and its Open6G R&D Center announce the availability of the first production-ready private 5G network fully automated through Artificial Intelligence (AI). The system is built on open-source components enabling a fully virtualized, programmable O-RAN compliant network in a campus environment.
The network provides connectivity to 5G devices (smartphones, cameras, and 5G dongles) for video conferencing, browsing, and streaming, key ingredients of experiential learning activities at Northeastern. “This industry-first O-RAN network provides WIoT researchers and industry partners with a fully customizable platform for research and development in new wireless use cases and technologies beyond 5G” says Tommaso Melodia, Professor of Electrical and Computer Engineering at Northeastern and WIoT Director.
The WIoT 5G network is built on open-source programmable components, compute solutions from vendors including WIoT partners Dell Technologies and NVIDIA, software-defined radios, and dedicated automation and orchestration pipelines through zTouch, Northeastern’s AI-based management, control, and orchestration framework. zTouch enables streamlined deployment of the software-based infrastructure in a matter of seconds, with automated configuration from high-level intents and management of the software-defined radio frontends. The network runs on Dell servers and uses OpenAirInterface and Open5Gs as the main radio access and core network implementations. It also features base stations based on the NVIDIA Aerial Research Cloud, integrating a GPU-based physical layer and OpenAirInterface. The system runs from a server room that terminates connections from 64 antennas in an indoor space to software-defined radios.
Northeastern’s private 5G deployment showcases for the first time key features of next-generation wireless systems:
- Openness and programmability – The network follows the O-RAN architecture in a multi-vendor environment, with open interfaces, near real time RAN Intelligent Controller (RICs) and non real time RIC. Components are based on open or open-source software that can be updated to track the latest research in algorithmic and data-driven control and optimization of the RAN.
- Resiliency and self-healing behavior – the zTouch automation framework extends the best practices of cloud-native systems to the RAN. Specifically, continuous integration and deployment pipelines re-provision base stations and other components in case of failures or malfunctions.
- Intelligence orchestration – the platform can automatically instantiate and orchestrate xApps, rApps, and dApps while managing conflicts. This is done through a simplified graphical interface that converts high-level intents into a set of intelligent network functions, data flows, and control outputs.
The network is deployed and serving indoor users in the Northeastern University campus in Boston, MA, with an extension to the Burlington, MA campus planned in the upcoming weeks. These two locations are part of Northeastern’s FCC Innovation Zone, thus enabling over-the-air experimentation in a number of frequency bands, including the CBRS band.
This private 5G deployment comes with unique opportunities for research in next-generation wireless networks. Its open, programmable, and virtualized nature simplifies the deployment and testing of new, advanced features, and it exposes telemetry and network performance to build AI and ML models that can truly generalize. The network can be used to develop and test advanced use cases including spectrum sharing mechanisms, AR/VR, end-to-end slicing solutions, and advanced security solutions, among others.
Use Case / Application