Ray Le Maistre, TelecomTV (00:05):
So we're in Dublin for Fyuz 24. I'm here with Rob Hughes. He is head of wireless marketing and product strategy at Fujitsu. Rob, great to see you. Thanks so much for joining us today. So can you just tell us what Fujitsu is highlighting here at the Fyuz show?

Rob Hughes, Fujitsu (00:22):
Well, we've got a lot of really exciting things happening with Open Ran. We've got a new massive MIMO radio as part of our growing our U portfolio and introduced a new RAN intelligent controller. There's also a lot of new exciting things happening with AI and ml. So some new developments with AI ran as well as an AI powered accelerated root cause analyzer, as well as a machine learning powered design tool for optical networks.

Ray Le Maistre, TelecomTV (00:53):
Okay. Wow, that's quite the roll call there, but can you tell us more about that accelerated root cause analysis tool? Because these kinds of things are really important to operators, aren't they?

Rob Hughes, Fujitsu (01:04):
Yes. Yeah. The root cause analyzer tool is quite exciting because for operators, a network alarm storm is when you have a sudden increase in alarms where a large proportion of them are severe or critical. And this causes a lot of challenges for operators that have to sift through large data sets, especially if they're using traditional tools. So with this solution, it uses AI and ML to improve the root cause analysis. It quickly identifies the root cause, performs root cause analysis, and then provides recommendations to the operator on how to address 'em. So that's one of the cool things that we're showing off at the show today.

Ray Le Maistre, TelecomTV (01:54):
Okay. And that sounds, I'm really excited about these kind of solutions because they don't get maybe as much airtime as maybe they should, but the operators of course really appreciate them. Now Fujitsu obviously is making a name for itself in Open, ran with quite a few deals already. You mentioned earlier on a few of the things you're doing. Can you elaborate a bit more on what's new in Open ran for Fujitsu? Sure.

Rob Hughes, Fujitsu (02:25):
We recently announced our RAN Intelligent control, so that includes both near realtime Rick as well as non-Real time Rick. It is supported with a robust software development kit to access AI and ML functionality. This includes some off the shelf machine learning models that can be adapted to and trained on and operator's data to really enable the app developers to come up with new apps rather quickly. The non-real Rick is already deployed in a network acceleration project in the United States as well as will be trialed with an operator later this year. And the near real time Rick will be out shortly in 2025. For our massive MIMO radio, we're showing our new 64 transmit 64 receive radio that's fully open ran compliant and it's a high power, high performance solution that supports advanced beam forming to really improve spectral efficiency and energy efficiency.

Ray Le Maistre, TelecomTV (03:36):
Okay. You've mentioned AI a few times already and of course that's an incredibly hot topic and AI in the radio access network looks like it's going to be a pretty big thing. Is Fujitsu doing anything specific in this area?

Rob Hughes, Fujitsu (03:53):
Yeah, we've been working with NVIDIA for about four years now, and now we're happy to share that. We've also been working with SoftBank during this time, so we're going to be announcing an MOU that we've signed with SoftBank to increase the joint r and d that we're already doing with them to develop software that will use AI to improve RAN performance and really improve the communications quality. So this work has already been underway. We're super excited to share that. We've got already an outdoor proof of concept that we've been collaborating on with SoftBank and it's already showing results. So the solution uses Fujitsu vRAN software, which will run on an Nvidia GH 200 Grace Hopper, super chipp, and as I mentioned, it's already producing results. We're able to support 20 cells simultaneously, all with running high quality communications. So it's not just showing that it's something that's in the future, it's real, it's here, it's working, and we're really excited to share that with people.

Ray Le Maistre, TelecomTV (05:11):
Okay, excellent. That sounds like a great development. And of course SoftBank and Nvidia are two of the companies really pushing things forward in AI and the RAN as we know.

Rob Hughes, Fujitsu (05:19):
Yes.

Ray Le Maistre, TelecomTV (05:20):
Now of course Fujitsu is also well known for its optical networking portfolio. What's new in that realm?

Rob Hughes, Fujitsu (05:28):
Sure. So we've been working on our open line system designer. So this uses machine learning to create a digital twin using real-time data from the optical network so it can be used to really speed up and optimize multivendor networks, sorry, speed up the network planning and testing of them and to optimize it for multivendor networks. And the challenge with some of the today is they're not really set up for multi-vendor networks. So that results in having to do a lot of manual field testing and field collection of data and that it slows things down, makes it a bit complex. And one of the challenges is it really ends up with just sort of a static view in time of what the network looks like. So with OLS designer, it uses machine learning to gather the data such as things like the transponder, configurations, the ROTEM topology, and then uses intelligent network planning to do things like transponder, reachability assessments, recommends modulation formats and data rates so that you really get peak performance and scalability. And the beauty is that it's always learning. So it's always training the model to improve the designs, to speed up your network planning, which speeds up service delivery so you get to revenue faster and really maximizes the amount of capacity in the network at any point in time. So lots of exciting things happening.

Ray Le Maistre, TelecomTV (07:10):
Okay. Well I mean obviously there's a lot of excitement about AI and the ram, but it's great to hear that machine learning tools are making a difference to the really important core transport networks as well. So that's a great development. Rob, thanks so much for joining us today. Thanks for having bringing us up to date with what Fujitsu is doing. Lots of exciting stuff there. Thank you.

Rob Hughes, Fujitsu (07:15): Yes, thanks for having me Ray.