ONE Summit Regional Day: AI Networks, Private 5G, and more

Blog recap by Yogendra Pal and Sriram Rupanagunta

LF Networking’s ONE Summit Regional Day was hosted by Infosys, on Nov 30th, at their beautiful campus in Bengaluru, India. As part of this event, Sriram Rupanagunta was asked to present his thoughts around “AI for Networks and Networks for AI” as a panelist along with other technical leaders in this space. This was very much aligned with the event theme “AI powered Networks”. The summit also focused on various technical talks around LFN open source projects and how these are enablers for enterprise, 5G and other use cases. Yogendra Pal presented insights on “The LFN 5G Super Blueprint: Overview & AI Exploration”. We would like to thank LFN and Infosys for their invitation and congratulate them on hosting this great event to foster open source networking collaboration in India. 

Insights on the “AI for Networks and Networks for AI” panel
  • In our experience, the use cases for AI in networking are mainly in the Troubleshooting and Optimization areas, which include examples such as RAN failure detection (using RAN data & weather data), and Slice optimization
  • Some of the challenges being faced are the availability of real life data for training the models, and converting POCs to production deployments (by proving the effectiveness of the predictions)
  • In terms of the Infrastructure requirements for running these AI workloads, in the recent past, our NVIDIA GTC demo was showcased, wherein RAN & AI workloads are orchestrated and switched on the same hardware with potential use cases for workload and compute optimization. This leads to another challenge in orchestrating the necessary infrastructure and switching the workloads, where the LFN projects such as Nephio will play a key role
  • In terms of trends using GenAI for the networking and for Enterprises, there is increasing usage of retrieval-augmented generation (RAG) for augmenting the model, which reduces the need for training the models using enterprise data or tweaking the LLMs
  • The format of the data, which is another challenge, needs standards such as ORAN (for providing data in a vendor-neutral form)
Insights on “The LFN 5G Super Blueprint: Overview & AI Exploration” presentation

The slides from this session are available here.

The 5G Super Blueprint ecosystem and landscape are shown below, wherein various open source components have been part of continuous development in the orchestration space inclusive of Nephio, EMCO, and ONAP to name a few.

LF Networking has recently compiled a list of on-going and completed blueprints into a 5G Super Blueprint library, which helps to enable enterprise and Private 5G networks use cases across the ecosystem:

In the LFN 5G Super Blueprint, there is ongoing effort towards enterprise to showcase the value of 5G for connectivity and enable various use cases. Aarna Networks has been a key contributor to this effort over the last 5 years providing the orchestration and management for several interactions and keynote demonstrations that have been shown live to thousands of community members. To see our latest contribution in this regard, refer to this presentation from the recent LFN D&TF event: 5G SBP: Orchestration of OAI Core and Amarisoft gNB with EMCO.

Orchestrating various components like O-RAN, 5G core, and edge applications requires an integration of management elements and a 5G strategy. Balancing performance with cost, and integrating multiple technologies into a unified system, are also key concerns. 

To address these challenges, Aarna Networks offers AMCOP, an open source orchestration, lifecycle management, real-time policy, and closed loop automation platform for edge and 5G services.

Are you experiencing challenges pulling together a Private 5G Network? Aarna Networks expert team is here to help. Request a free consultation to discuss how to create value with Private 5G for your specific requirements and use cases or request a Free Trial of AMCOP today.

We use cookies to enhance site navigation, analyze site usage, and assist in our marketing efforts. For more information, please see the Aarna Networks Cookie Policy.
Accept cookies