Tony Poulos, TelecomTV (00:07):
Hi, I am Tony Poulos at Mobile World Congress in Barcelona. Today I have a very esteemed panel of people that are going to help me understand the adoption of O Ran or open RAN and its benefits. Gentlemen, welcome. Let me introduce you first. Sandeep Sharma, who is the Vice President and global head of portfolio 5G and O RAN networks at Tech Mahindra. Thank you for being here. Next we have Wei Yeang Toh, who is the general manager Ecosystem Development Organization Network and ED Solutions Group at Intel Corporation. You must be very important with such a long title I'm sure. And last but certainly not least, we have Harkirit Singh who is the CEO of Ascend Digital Solutions Limited. Gentlemen, thanks for being here and thanks for putting up with my very long introduction. But let's get right onto the subject. So Sandeep, let me start with you if I can. We're seeing strong momentum and progress with carrier's adoption of open RAN and it's taken a while to get there. I know, but what are the key takeaways and what are the next steps to continue enabling service providers to get this ecosystem running overall?

Sandeep Sharma, Tech Mahindra (01:09):
Yeah, right, you're right. Tony O-ran is now shaping a well really well and you know that Oran is actually providing an ecosystem to the operators or to the carriers who deploy a flexible solution the way they want to be. It's an inevitable change that will changing, that will change us changing the transforming the RAN industry as a whole. We as a techman involved in almost all the large scale deployments in Open ran, including small scale trial as well and the POCs of the new age solutions around it. At the same time, when we evolve ourselves as a system indicator in this domain, we realized their set of pillars or learnings that can be leveraged or that can be replicated for a successful or journey and few of them includes importance of interoperability. You need to make sure the different systems components interop. At the same time you need to look at the lab ecosystem place so that you can create a pre integrated pre-production sort of in a stack ready for the operators. At the same time, importance of automation as well because you are bringing layers of AI and cloudification virtualization, making it work for the customers, you need to put a lot of automation around it. And finally the importance of security. Since you are moving towards ization virtualization open interface, you need to make this networks more and more secure additionally of security from the IT world and telco world together. That's how I summarize it.

Tony Poulos, TelecomTV (02:29):
Wow. That was a good summary though. Wei Yeang, I'm coming to you now. How do you see the progression of O ran and how is Intel helping accelerate O ran adoption? Because a lot of people are wondering where Intel fits in this.

Wei Yeang Toh, Intel Corporation (02:41):
Alright, alright, of course. Yeah. Hey, so we see tremendous progress year by year. This is very encouraging, right? Just last year of different announcement happening at t with Ericssons, a Telus with Samsung and so on. Few more to mention, but Intel being the forefront of vRAN for over a decade and we've been in networking business for more than two decades as well. And thanks to all the key learning coming out from network function virtualization, this industrial knowledge accumulation helped us to win the trust from the partners that all the partners is leaning in, trusting the technology that evolve from network function virtualization into the V ran and V is the most open. RAN V ran by far is the most complex decoupling that required to go and address. So what Intel is doing, right, Intel is doing a couple of things right? First, we are providing a fundamental base V ran platform that could help the partners to protect the investment.

(03:39)
Therefore they started right at the beginning with the opportunity to add on more capability that built into the platform itself is just progressively activated. Let's put that way in the 11 terms. We're also working with the ecosystem partners to make sure that they consume the additional capability that we keep on adding on to make it perfect, right? We have VRE Bruce that come with the fourth generation Zon cable processor that we are now say last year able to deliver to the TC expectation and now this year itself we demonstrate how things can go one step further by using a vRAN AI development kits to activate the AI features and AI adoption will need to go by faces. So this is not a simple task, but thanks to all the partner trust we've been able to continuously help our partner to develop the solution on track. At the same time helping the telco to protect the desire. TCO

Tony Poulos, TelecomTV (04:35):
I'm very interested firstly to find out a little bit more about your company Ascend Digital Solutions. Let me know about that.

Harkirit Singh, Ascend Digital Solutions (04:41):
First, Tony Ascend Digital Solutions, a company which is based in Ghana. We are actually looking at doing large scale digital transformations. So over the last few years we have taken up a lot of digital infrastructure and right from the backbone infrastructure, fiber, the data centers, wireless infrastructure and we are transforming the entire paradigm to bring more and more digital connectivity to reduce digital divide and increase financial inclusion. So our mission is to actually create ubiquitous seamless networks to provide universal access.

Tony Poulos, TelecomTV (05:22):
Well this question is aimed for you then. The rate of new technology development and innovation has significantly increased over the last few years. Can you shed some light on how open RAN plays a crucial role in driving digital connectivity for places like say Africa and even India and Asia? How is it happening? How's it rolling out?

Harkirit Singh, Ascend Digital Solutions (05:42):
So I think technology in the markets or usually separate, so I won't say technology brings the right level of transformation and the country's agenda for digital, but what technology does it actually accelerate the pace of that particular digital transformation? But when we look at open RAN and the traditional ran, we are going through the whole process of launching an open ran network first in Africa. There are a lot of trials going on, but we are really looking at something that comes out this year. The challenges that we are seeing in the traditional RAN vendors and the open ran vendors, the open ran ecosystem is still small and it is also fragmented into how the different components are built up. So what is happening is that if I am a module A, I put a lot of my r and d money in module A and if I get an order, I'm not able to leverage the scale because scale does not exist.

(06:40)
So when you actually pack up all the modules, the price of open revenue is higher. But the question comes in for us when you look into the technology to bring the level of innovation that is needed to drive that level of connectivity to the edge, to bring that open innovation ecosystem that actually creates that hyperscale infrastructure. That's where open land differentiate itself with a traditional brand. Other key thing that we are looking at is the speed of bringing speed and acceleration of the digital transformation. Now if you take a traditional RAN ecosystem and look into the different processes to actually go in terms of zero touch provisioning, full automation configuration of the infrastructure, we said that we look at open ran, that they have done that there is a lot of possibility to bring the infrastructure faster in time and deliver on demand scalability that is needed as part of the cloud and virtualized infrastructure. So that one of one key component of having a technology which is cloud native and bringing that level of automation gives us an advantage or let's say open ran operators in advantage to go faster as compared to a traditional ran

Tony Poulos, TelecomTV (07:54):
Very well explained. For somebody like me, I sometimes grasp with all the layers that are involved with open ran rollouts and I think that was very well put, but I'm going to come back to you at a moment about new networks. New networks are also getting more complex, right? What role do you see AI taking in furthering network transformation? We have to talk about AI because that's the subject that's been hitting us here constantly, but what role do you see AI taking in furthering network transformation journey by addressing things like automation, manageability, and of course that awful word sustainability. It's coming up every day now.

Sandeep Sharma, Tech Mahindra (08:34):
Yeah, honestly speaking, the values that Oren is bringing in in terms of virtualization and having a common data lake across the different layers of network functions brings a lot of applicability of ai. Now you need not to worry about someone else looking into multiple domains. You can have your AI engines running, you can take observability end to end across right from hardware infrastructure layer to the network function functions layer and bringing a meaningful observability out of it so that network can take a meaningful decisions at the same time. There are two versions of AI is like traditional AI and the Jan ai and you can use Jan AI for more operational efficiencies as well. Taking all the customer complaints and all faster resolution. All those things can be better handled and most important thing, the sustainability part of it, sustainability is not new. It was there in the traditional network. There was need for that and it'll remain as long as we are operating on natural resources. But if you look at the AI part of it, AI once helps you in taking meaningful decisions out of how the network is operating, where the power consumption is happening in a higher level where you can optimize it. AI brings a lot of flexibility. Having a software virtualization done with open ran interfaces, you can bring different models and achieve efficiencies out of it in terms of lower power consumption or optimize power consumption across the networks.

Tony Poulos, TelecomTV (09:53):
I can see Wei Yeang going. "Oh, oh it's also important for us". He wants to let us know.

Wei Yeang Toh, Intel Corporation (09:58):
Go ahead. Yeah, I'll just power on top there. Absolutely right. AI is a topic set. Every single meeting in this trip towards the end of meeting we ask what's your AI strategy? Well, if I put AI into O ran context very specifically or vRAN context, it will be a progressive improvement, let's put it that way. So let's go by three layer. I talk about the Intel vRAN platform early on within the platform itself, it do come along with the AI capability that waiting to unlock is the same TCO Intel flex ran software reference software is the glue that's so far by far in the marketplace most of the ecosystem partner will refer to. So it is a glue to help the ecosystem to integrate between the upper layer with the platform itself and within the flex ran software. Now we are bringing the V ran AI development kits on top of it and that helps to unlock the capability.

(10:56)
I'll throw three example out the energy saving, helping to able to analyze the real time, say powers their usage and collect all those metrics and able to analyze the metrics and provide the best recommendation to put the CPU call into either stage and wake it up at and t showcase in the booth that we're just 16 one 6% of the power saving and translate to 4.2 million of the saving just basic things calculation energy, that's one second part of it. The beam management. By implementing ai, it could help to manage the beam forming beam management in a more effective way and can be user defined as well. And then the third one that we showcase is the resources management through the slicing. But the key point here is the platform and the software will provide the capability we would like to help together within the ecosystem.

(11:50)
How the telco with the data center coming through the telco is the most important thing. The dataset will decide how accurate and how effective it is to help the telco to gain the benefit out from it. So therefore we are providing the pre-trained model based on flex and software and the vRAN kits, v AI kits with the data that come in dataset coming with telco to help to pre-train the model. The more model we pre-train, the better it is and it benefit all the ecosystem partner use it. You don't need to repeat again, you just take whatever we have it help the ecosystem partner to grow.

Tony Poulos, TelecomTV (12:21):
Harkirit, how is AI going to to affect your operations?

Harkirit Singh, Ascend Digital Solutions (12:24):
I think for ai, what we're looking at largely from a sustainable perspective, our goal is zero bit zero watt. So what it means is that if there is no traffic on the network, I don't want to power it up because see the challenge, what we are seeing in Orran is there is a lot of decoupling of interfaces and opportunity to bring course hardware, but it is also increasing what we have seen so far, the power consumption in the data centers because the amount of CU that you need to deliver the similar line of capacity and the amount of interface that you need from the fiber to the back hole is huge. So from a perspective of power consumption, it is a bit higher. But I think overall from an AI perspective, what we really want as Max said, is we need to have the models which are trained because we want something which is plugin. We want something that something comes up and says, guys, here it is, and we will have a 30% power saving. We don't have to wait for six months one year to have those train models. We might have a localization impacts like how much of the sites are running on the electricity, how much of sites are running on the solar and how the local operation is working. But those trained models, if they come pre-trained, then we actually save a lot of times to grade that efficiency gain. That efficiency

Tony Poulos, TelecomTV (13:51):
Makes a lot of sense. In summary, I'm going to ask you one question and I want all of you to give me a nice concise answer on this one. From an O ran perspective, how do you envisage the network evolving in the future? How you do that in a short time? I don't know, but have a go.

Sandeep Sharma, Tech Mahindra (14:09):
So network's going to be more and more autonomous day by day. As we evolve further, AI will play a more important role and the systems will be native AI driven. We need not to create additional layer of AI and how fast we can do or better we can do is as we evolve in the industry, we all will see it. But that's how I see it is happening. With the Rick evolution happening with the accepts and a reps coming in, evolution ultimately to six G, AI is going to drive how things should happen.

Tony Poulos, TelecomTV (14:36):
Yes, I knew someone was going to say six G and it had to be you. Wei Yeang. What do you think?

Wei Yeang Toh, Intel Corporation (14:42):
Well, few things. We were going to see more innovation coming up within the edge computing and 5G converge together become a bigger market. Right? Partner will come in and collaborate. We'll see the interfaces will get defined better and better, right? Interoperability will getting better and better as well. The partnership among the ecosystem will be better as well because they are all those key learning will get shared across the ecosystem. You encourage, we all encourage the more ecosystem to come and participate in this, grow this business, grow itself is very healthy and of course AI will come in a progressive way and any key learning will help the others here. Network networks function as well. So to simply put the Oran going to be pretty positive in the future the way I look at it. So if we make progress and the only things that we see is upside right now,

Tony Poulos, TelecomTV (15:40):
That's good to hear. Ah, Harkirit, what's your opinion of the Future Network?

Harkirit Singh, Ascend Digital Solutions (15:44):
I think we want Orran to be embraced with a larger ecosystem. The likes of Ericsson, Nokia's really, really need to actually embrace it because so far they've been resisting the whole thing. It has been driven by new age operators likes of Rakuten Mobile, which has done an excellent work on bringing all and live, but these other big vendors needs to embrace so that the cost of ownership and the TCO gets lower. From an innovation perspective, it's the best technology any telecom operators will have. And if you see from what one in one has done with about 500 of edge data centers and stuff, they're opening the interfaces and bringing the content closer to the user. So today, if I have to stream a very high quality video, I don't need to cash from the dollar center and I can do a local stuff. I can break out there. The user plan doesn't need to go back into my core. So that's the advantage of bringing us, as Max said. And Sandeep is saying is that the innovation power of open ran is humongous to transform any telecom operations world where it'll bring the benefits of a local content and distribution, mobile edge computing to the users.

Tony Poulos, TelecomTV (16:59):
Gentlemen, thank you so much. I now think that Ran has a big future. You've convinced me. Thanks very much, very much for your help. Sandeep, Wei Yeang and Harkirit, thanks for being with me today.

Wei Yeang Toh, Intel Corporation (17:10):
Thanks.

Sandeep Sharma, Tech Mahindra (17:10):
Thanks.

Harkirit Singh, Ascend Digital Solutions (17:11):
Thanks Tony.