Vikram Prasad, Head of Program & Client Solutions at Amdocs, unveiled the latest advancements in AI-powered RAN solutions. Amdocs recently launched its Cognitive RAN Automation (CRA) platform, incorporating open-source standards and NVIDIA technology to optimize network operations. This innovation enables real-time root cause analysis, energy savings, and intelligent traffic steering—reshaping how operators enhance customer experience.
The CRA platform, already tested with UK telecom partners, supports energy-efficient cell site management, massive MIMO optimization, and digital twin network simulations. By leveraging AI, operators can reduce power consumption, improve coverage, and streamline operations.
Alejandro Piñero:
All right, here we are once again at the FNTV studio in Barcelona, covering MWC all week. My name is Alejandro Piñero, your host, and it might be Wednesday and there might be some tired eyes, but again, not here because we are having some great chats. And I'm so excited to welcome Vikram Prasad from Amdocs to the studio. Thanks for dropping by, Vikram.
Vikram Prasad:
Thank you, Alejandro.
Alejandro Piñero:
Now, I expected AI to be a big topic at Barcelona and certainly it has been, but the amount of information and use cases around how it works into the network has been really overwhelming. So I'd love for you to give me your point of view and of course from Amdocs in terms of what you're working on in that area.
Vikram Prasad:
Definitely. Thank you for this opportunity. So in Amdocs today, like recently just on Monday, we actually announced a general availability of our Cognitive RAN platform. It's a solution that's pretty much like a non-realtime break solution based on the open source community. And the cool thing about it is it actually incorporates the O-RAN standards as well as it incorporates NVIDIA's platform. So in terms of running heavy RAN workloads and providing root cause analysis and insights for planning engineers as well as operations teams, we are actually building some cool use cases on top of the CRA platform.
Alejandro Piñero:
So let's talk about those because it's always great to bring it to life and give folks a few examples and Amdocs have the breadth. You guys have the clients out there. Why don't you tell us a little bit about whatever you can share of how this works in the field?
Vikram Prasad:
Definitely. So before the launch of the GA solution, we already tested this with ARIANE in the United Kingdom. We already got certified from the Telecom infrastructure project for four of our rApps. So we actually currently have the energy savings rApp, which is pretty much looking at historical network RAN data and leveraging the ML algorithms inferencing when the cell site is actually having lesser utilization and power.
And depending upon the type of customer experience at that particular point in time, takes actions to reduce the power or turn off that particular cell. So this has been certified by TIP. In addition to the energy savings use case, we've actually built a massive MIMO optimization rApp as well. So if you look at the massive MIMO antennas today, it comes in different beam patterns. It comes in 4 by 4, 8 by 8, 32 by 32, 64 by 64 and beyond that. What Amdocs has actually done with the rApp is looking at the geolocation of subscribers, looking at the density of signal and quality of the users, analyzing those in real time and changing the beam patterns to actually serve capacity and coverage where it is actually needed. And this is actually leveraging the NVIDIA GPUs behind to produce a holistic customer experience for the operators.
So beyond these two, what we've actually also built is traffic steering. So traffic steering rApps, and we've also built a sell outage compensation. So we currently have four rApps available and we have a roadmap of many more. We also have announced partnerships with AirHop, Zincworks as well as [inaudible 00:03:49] to actually have their rApps also onto our marketplace. The other thing that we have also announced is we have built digital twin network simulation that's actually based on the NVIDIA aerial platform. So this actually gives the planning folks a huge opportunity to actually sit without actually having any drive testing or going out in the field. Sit in their offices and simulate a digital twin of their network for planning and customer experience optimization. So we really think that these cool use cases would really drive operators to actually gain that additional customer experience that they're actually after.
Alejandro Piñero:
Yeah, that's addressing a lot of challenges and it's fair to say you've been busy, but thinking ahead, just real quick while we close out our interview today, what are you working on now?
Vikram Prasad:
So we actually hear a lot about inferencing agentic RAN. So what we are actually doing is taking all of those workloads coming out of the RAN interface, core transport, training it in our models, and giving the operations the insight of end-to-end visibility of how faster they can actually get root cost analysis for their end customers. So that's something that I think is up in the horizons.
Alejandro Piñero:
Amazing. Well, we'll have to have you back to tell us all about that when it's ready. Vikram, thank you again for joining us here in Fierce Network TV and have a great rest of the show.
Vikram Prasad:
Thank you, Alejandro. Thanks for the opportunity.