Celona New Edgeless Architecture – Streamline On-Premises Footprint for Scale

Hello everyone. So I wanna continue on the tradition that Mammoth and Rajiv started in 2019. So we always use MFD as a platform to come and talk about something new and exciting that CEL is gonna deliver to our end customers.

And continuing in that tradition. And back then we, I think it was said earlier, back then, we actually launched CEL at this event. Uh, and today we want to give you all a sneak peek of the edgeless private 5G architecture, which we believe is truly going to transform how enterprises start consuming private 5G going forward.

So without any further ado, let me show you something that all of you're very, very familiar with. Uh, this is how the WLAN wireless LAN architecture started, uh, back in the early two thousands, right? You had a wireless land controller, you had your access points that were being managed by it.

You probably had a local management system and then eventually moved to a cloud-based management system where the wireless land controller still remained. This transformed into this where you actually had autonomous access points being deployed on sites which were completely cloud managed. And today enterprises, this has become commonplace.

This is what enterprises now expect from their wireless solutions. And, uh, this is how most private 5G solutions still look like, right? It's a rack full of appliances and servers to run your ran software, to run your core software and all of the additional things that are required to run a cellular network.

So we have always been on a mission to create something that enterprises can consume, and this is what our architecture looks like today, right? You have an access point, which is here, that plugs into your existing access switch. Uh, it's usually POE or p oe plus plus powered.

You have your CEL Edge appliance, which is typically running the Cel Edge os, the Cel Edge software, which comprises of your mobile core, your spectrum management, and some additional services that are required to manage the private network. And then you have your cloud-based management and monitoring platform. This is what all of our deployments look like today, okay?

And this is obviously a significant improvement from what you just saw in the previous slide. Uh, but again, as I said earlier, we are always on a mission to find ways to make this easier to consume for our enterprise customers. And as we kept looking at this architecture, we felt the CEL Edge appliance or the need to deploy a CEL Edge op operating system on premise is a friction point that probably we can remove.

And that's what we are calling as the edgeless architecture, which we believe is the future of private 5G enter, uh, for enterprises. And what we are doing in this case is we are actually getting rid of the CEL Edge appliance. And to achieve this, we are not just lifting and shifting the entire software and moving into the cloud.

We are very carefully and surgically picking specific microservices that need to actually run on the access point, and then those services that actually need to run on the cloud. We understand that for customers, the data security and integrity is very important. Their application data needs to stay on premise.

So we have taken all of the data parts services, all of the services that require you to integrate with your existing L two L three infrastructure, your existing firewall and NAC infrastructure. All of those services will now actually be running on the access point. And then the services that are not related to data path, the control part, the spectrum management will actually move the cloud.

And with that, we've actually removed, which we believe was one big friction point in salon's private 5G deployment. And, uh, MJ is actually gonna demonstrate that to you in just a little bit. You actually see the access point here and nothing else.

This has just been plugged into, uh, ethernet. You have internet connectivity and boom, you have private 5G network in the room right now without a rack full of hardware. How much additional hardware did you have to put in the access point?

And, and by the way, thank you for now calling it an access point. So Keith, Keith, Keith likes that. I hate it.

No. Well, no. I, I kind of do, but it's Enterprise people you're trying to sell to.

Exactly. Sell. It's an easier Absolutely Absolute.

Except for you have to explain to that it's not wifi mm-hmm. Which is now 80% of the conversation. Yeah.

5 G access point. Yeah. Wifi access point.

Yeah. Uh, but yeah, so I think from a hardware perspective, and that's a really good question. What we have done when we were designing our, our 5G access point, we actually thought about this because we realized that eventually we'll have to get to this architecture.

So we put in enough horsepower in our access point to be able to run these additional services. But from a form factor perspective, this access point has a 4G radio as a 5G radio and is currently running all of the services that I talked about earlier. And there's local traffic breakout happening on this access point P OE plus plus, sorry, P OE plus plus.

That's right. Yeah. Just a curious, uh, please, I kind of perceiving this as a hybrid edge architecture, but I hear edgeless, it's almost likey gravity, the networking.

Yeah, I think it's, it's a, I would say it's something that we have for, for example, if I go back to the previous slide, all of the services that I was talking about earlier was running on the CEL edge. And hence, as we have now removed that requirement from our end customer perspective, it's kind of coming from that perspective that now it's an edgeless deployment. And from a customer's perspective, all they now need to do is do an out of design for the room, for the site, deploy the access point where they need it for the use case they need based on capacity and coverage.

And they don't need to think about anything else with regards to, Hey, I have an appliance that I need to find a rack for, I need to find connectivity for. That's helpful. I mean, from a CEL portfolio perspective.

Yep. This is edgeless. Yep.

However, you know, uh, we're talking to folks out there and they hear edgeless. I think it might cause a, a bit of initial confusion, just kind of an FYI that's a fair point. Yeah.

Did I miss, um, size and scale as far as like how many access points, uh, deployment like this will support? Is it meant for one meant for small meant for large campuses? It's A, it's a great question.

So maybe you hold that thought. I'm gonna cover that in the next slide. Right.

So as I look at these components, CEL access point, as I said earlier, has been now architected both from a hardware and a software perspective to run these additional services. So the data path, the L two LT integration, the NAC and firewall integration, which is running on the CEL Edge appliance, is now running on our access points, the control path, the RAN management, the Spectrum management services are now actually moving to the cloud. And we dynamically instantiate these points of presence with affinity to the location of the site because obviously since the control path is now in the cloud, we wanna make sure the latency is as minimal as possible.

So these are dynamically instantiated and the CEL Orchestrator continues to help manage and operate these particular networks. Right. And going back to your question, from our perspective, I expect that this particular edgeless deployment is gonna be adopted at a pretty high scale.

Obviously the biggest advantage is gonna be for these large distributed enterprises where you have remote branches, one or two access points. Because like one outdoor access point or two outdoor access point can cover a pretty big area when it comes to cellular. So when you're deploying those and you're deploying those across hundreds of sites, not having to deploy a single edge appliance or a multi edge appliance for high availability reduces friction both from a cost and just from a manageability perspective.

So I expect that for large distributed enterprises, this is gonna be a no brainer. But even as the size of the site, uh, goes up, I think this is still gonna be a fairly attractive option. There might be some use cases where you have high speed mobility, some advanced use cases where latency comes into picture.

And that's why we are gonna continue to have our edge appliance based deployments. They're not going away. There are still going to be use cases where that is still the right recommended solutions.

Customers will deploy a single appliance or multiple appliances for high availability, and those would be advanced use cases like high speed mobility with AGVs and AMRs, and you just don't want to, uh, rely on the latency of the WAN link to run those critical applications. Customers can also deploy the operating system Edge os on a virtualized infrastructure. They have that option if they want to do that, and then very soon they will have the ability to deploy the edgeless architecture or the edgeless deployment.

And as I said earlier, it's going to be very, very attractive to large distributed enterprises, but more so I expect that this is going to be the mode. And we already seeing that as our, as I get pulled into customer discussions, prospect discussions, and I have to present roadmap. A lot of customers are, you know, what edges is coming, let me wait for a few months before I do the POC.

So I see that, uh, the attraction of this is already causing customers to get very, very excited. Please. Um, does that allow you to also be a little bit more competitive on the pricing since you're removing some of that hardware component?

Uh, to be honest, I mean from a pricing perspective, and obviously we can, uh, get into another discussion and happy to discuss later as well from a pricing discussion, the solution is already very, very competitive. Okay. Uh, I would say that when you have a large number of sites, particularly when you have hundreds and thousands of sites and you're deploying just two or three access points at that point, having one or two appliances that every site starts to add to cost.

But most deployments, edge appliance cost is really not a big consideration. To be honest. I think the, the biggest driver for this is the ease of deployment and not having to manage an additional overhead with additional hardware.

And, and as I said earlier, this is how they deploy their wifi today, right? It's the controller less architecture has become the norm. So why not on the private 5G as well?