Author: Annie Savage

Recently on The Connectivity Matters Podcast, we spoke to Raz Kivelevich-Carmi, the Chief Marketing and Business Development Officer at Cellium Technologies. With over 20 years experience from within the chipset ecosystem, Raz has invaluable insights into the tech that’s currently revolutionising the industry. We talked about the innovations happening at Cellium and how they’ll impact consumers. Read on to find out what’s coming next!

Tell us about Cellium’s Edge Air solution and how it differs from small cell solutions?

There are multiple technologies or topologies out there, ranging from full genome to small cell and all the way up to the oran types of solution, which is a very big hype these days. The databases have been around for some time, whereas the psyllium solution is, in essence, a desk-like solution. However, opposed to what standard desk does, we do it much better performance-wise as well as at a much lower cost than those existing solutions. The reason that the technology is based on an analogue-only type of deployment is that we do not do any analogue to digital conversions of any sort. We don’t utilise special cabling, most other deployments are going to be fibre which is an expensive means to transfer data from one end to the other, but we use standard copper cables which you have in your walls. If you don’t have them in your walls, we can put them into the walls at a much lower cost both for the equipment and the deployment cost itself. It’s very easy to put up copper wire, it’s much harder to put up a fibre wire. 

When you take all of this together, you’re getting an active DAS that works over copper wire in an analogue domain, with no need to do those conversions. The cherry on top of the cake is that we do it based on the SOC chipset that we have developed in house. The companies that do active DAS do all of those conversions. They take the RF signal, either they push it over coax cable, which is again very expensive, or they take it and they transfer it from analogue domain into the digital domain. In many cases, they then take the digital signal and then they push it to fibre and then they push it from one point of the building to another point of the building, then they do it again fibre to digital, digital to analogue and then they retransmit. The way that we do it is we take the RF signals and we drop them to a frequency which is referred to as an intermediate frequency, which is usually in the low hundreds of megahertz, and that can be pushed over a copper cable easily. I’m simplifying things that are chipped and are actually doing a lot in order to make it happen, but it’s much easier to keep the single analogue and push it across the wire. I make sure that I push it in such a way that I don’t degrade the performance. I do not add latency which all those digital technologies do. They modify the signal in order to have it passed from one point to the other in a robust way, which is very time consuming and expensive. That’s what we do, but we do it using our own chips. 

Some companies have chosen to take this path of analogue conversion only. However they do it with discrete components which isn’t very cost effective. This is why Cellium took the approach about two and a half or three years ago to, to take a system and to develop a chip that replaces those discrete components into a single chip. When they did it with discrete component, it was very expensive, which is the reason that they went for an SOC. We are the only company in the world today that has an SOC that does this RF live conversion. Again, this further reduces the overall cost of the system. When I compare it to the good performance active DAS, it’s where I have the merits of providing an active DAS at a very low cost. That’s the huge merit against those active DAS types of solutions.

So is the Edge Air solution suitable for all indoor vertical markets?

When we look at the market today, what we see is two segments. There is the private network segment, and then there is the bigger one, which is the eminent or the or the public domain. If you look at analysts, they would probably say that the public sector is going to account for about 70% of the entire 5G market, whereas 30%, will go into private networks. Now, there is a big difference between private networks and public networks from the deployment perspective, so I’ll touch each one of those and explain how the Edge Air solution does it. 

Let’s start with the public, which is the bigger segment, and probably the one that’s going to happen earlier in the game. One of the things that all the all the cell vendors and DAS vendors have seen is the fact that the majority of the MNO market today is what is called NSA or non-stand-alone, which means that legacy 4G is working alongside the 5G. Only about 10 to 20% of the seminars these days are using 5G standalone. That doesn’t mean that companies will immediately migrate from NSA into SA, but they will start to, and for the next three to five years they will continue using their legacy 4G cores and add 5G capabilities on top of it and have the cells work together. Now, this makes it harder for operators who now need to bring a box into the building that does NSA. They can do one of two things: they can rely on the 4G macro to penetrate in and to be synchronised with the 5G devices in the building, or they could bring 4G data into the building. In some countries it’s very common to have 4G data in the building, but in the majority of the countries that we’ve seen, what you see is the need to go into a building and put in a 4G and 5G transmitting and receiving element in. Now, that is a problem for many of the small sales companies because if you look at most of the small cells today, they are built around the Qualcomm chipset, there is a 4G Qualcomm chipset and there is a 5G Qualcomm chipset and you need to put both of them into the box and to have some collaboration between the two of them. In many cases, the 4G is coming from a different vendor altogether like Ericsson and Nokia, who are very common to be used as the macro vendor. But, going back to the box, you need to have 4G and 5G in the box in a single location and to get better coverage for both of them working together. 

Most of our customers would like to have a single box that has the capability to work on both systems. Now, this is where Cellium excels. Why am I saying that? Because the way that our system is built, it has two major components. One is called the CPU, which is the Cellium base unit, and one is called the CRU, which is the same remote unit or the radio unit. Our base unit interacts with third party equipment, we don’t compete with Ericsson, we’re not here to compete with Airspan. As I’ve stated before, we look at them as a signal source. What we do is distribute their signal across the building. We build our boxes in a way that interfaces with more than one vendor, so we can put in an Ericsson 5G and a Nokia or Huawei or a Samsung 4G and both of them will connect to our box. That CPU will then spread it to units across the building, and each one of those units will transmit both the 4G and 5G together from a single location. This is a great value to the customers because they need the 4G and 5G because they’re doing NSA. However, in the future, when they migrate away from NSA and reform their frequencies, they might be able to take out the 4G source signal source and put in another 5G signal source. Now it’s 5G, the cost of replacement or reforming for indoor types of deployment this way will be dramatically lower than replacing multiple small cells around the building. This is a future proof solution that allows the operators to build a system now and then move on to SA at a later stage with a very low cost of replacement. We wouldn’t make any money out of it at that stage, but customers will save a lot of money then. 

Let me just touch on the private network for a while. The private network is a very complex market these days, and there is a huge competition between 5G and WiFi. These two technologies want to play in the game of private networks. In the last two years or so everybody was saying private networks are 5G. This is because the 5G industry has been pushing this. In the recent year or so I’ve started to see more and more of the Wi Fi coming back and saying ‘No guys, we can do private networks’. You’re talking about quality of service, we know how to do quality of service, you’re talking about latency, we know how to do proper latency, you’re talking about security, we know how to do proper security. In my opinion, they will complement each other, in some cases it will be just WiFi, in some cases it’s just going to be 5G. In many cases, it’s going to be a mutual solution that serves the facility that it’s in. 

Now, the solution for private networks is going to look a bit different from the MNO that I’ve just described. Nobody cares about the NSA because these are private networks, so it’s going to be easy. From the 5G perspective, however, we need to facilitate the Wi Fi to go through the same infrastructure because at the end of the day, you don’t want to have one or two boxes that are in a single location and transmit WiFi and 5G over the same corporate wires. This is exactly what we’re doing in the private network. We’re building boxes that are doing 5G only. Our first products are 5G with what we call cascading capabilities. If there is a Cisco, Juniper or HP Aruba that is already in the facility, I don’t want to disturb that unit. The IT manager already bought that unit, he doesn’t want to take it off and replace it with a new one, but he does want to add 5G into the facility. So what we’re bringing in is our CRU remote units that are capable of 5G only. I facilitate the same infrastructure to work for 5G and WiFi. So, instead of having 4G coming in and 5G coming in, as I was explaining before, now only 5G will come in alongside the Wi Fi or Ethernet. I will facilitate both of them to go through the corporate wires into my box, and then I will spread the 5G and let the expert on WiFi spread the WiFi. 

To hear more of Raz’s insights into the industry and find out what advice he’d offer someone joining the sector, tune into the full episode of The Connectivity Matters Podcast. 

We sit down regularly with some of the biggest names in our industry, we dedicate our podcast to the stories of leaders in the technologies industries that bring us closer together. Follow the link here to see some of our latest episodes and don’t forget to subscribe.     

On The Connectivity Matters Podcast we were joined by James Morgan, SVP of sales at DriveNets. James is currently responsible for DriveNets’ growth strategy and leading the company’s European activity. Previously working at Juniper Networks as Head of Global MSP and NAS sales, as well as founding several successful startups, James has over 25 years experience across connectivity. He spoke to us about his perspectives on developments in communication technology.

So what’s your take on the current state of telecommunications?

Well, it’s interesting and dynamic. If you look at telcos and service providers, and where they make their money, there’s a huge ecosystem that’s eating into their profitability. Bandwidth is increasing dramatically, price per bit is decreasing dramatically, and the telcos and the service providers are in the middle of all that. There’s an opportunity for any business today to take advantage of cloud and the almost infinite bandwidth that’s available to really expand their businesses as enterprises. There are definitely challenges that a lot of CEOs and CTOs have, such as hyper scalers moving into their territory. They have to have a frenemy type relationship there, because cloud is the biggest macro trend that we’ve seen in my career. They’ve got over the top players coming in making serious amounts of money using their infrastructure. Whether it’s a consumer or business, if there’s something wrong with the network, it’s seen as the service providers fault, rather than anything else from the myriad of things that affect it. It’s increasingly complex.

The growth rates are not anything like some of those other players that I’ve just touched on. Companies like Juniper and Cisco and Fortinet, who are providing services through service providers in some instances are also partnering with the cloud players and, and so on. It’s all pushing the value of the service provider or telco down. We’re trying to help push back against those forces that are against them. Telcos need to solve these big strategic challenges as they try to digitally transform. Companies that have been around for a long time seem antiquated, even mediaeval in some instances. They’re hanging on to systems, processes and infrastructure that they really need to change and transform.

Moving into the future, are there any particular technologies that you are particularly excited about that are on the horizon?

Yeah, there’s a lot of artificial intelligence, which is a big word. And there’s a lot of artificial intelligence washing that goes on in everything. Everybody’s got an artificial intelligence capability all of a sudden, but I think that when that’s harnessed in collaboration with human beings it has the potential to really transform a lot of areas and do a lot of good. Whether that’s in healthcare, medicine, research or education, there’s some real positives from that perspective. Everything is hyper-connected nowadays. We have infinite access to compute and store and network that creates a combination of all of those aspects, realistically. I still get buzzed off that I can set my alarm on my phone, or my alarm system, or I can see someone ringing the doorbell when I’m down the shops. I think everything’s got a little role to play, but fundamentally the interconnectedness of everything is what excites me. As long as it’s used for the right things, in the right manner.

What impact would you say virtualisation and cloud native software has had in the telecommunications industry?

When you look at telcos and service providers, they’ve got different domains of the network and different areas that I think there’s a lot of benefit to be had from virtualization. I mean, one of the things that I did when I left Intact was actually set up a small startup that was focused on virtualization and using IT infrastructure to pool resources. Fundamentally what that brings, if you look at what’s happened in data centres, the same thing starts to happen in networking and security. Virtualization brings an awful lot of flexibility and an awful lot of power. What we’re seeing is that desire to have a hyperscaling operational model. That’s part of the transition that needs to be happening in the search provider world, if they’re going to compete with hyper scalars, we’ve got to be agile and flexible and have speed when it comes to delivering services and what your customers want. Fundamentally it’s all about business outcomes, and what technology can bring to an organisation with the right vision and desire. It really revolutionises and transforms the way that they can operate.

But again, it’s a journey, right? It’s an evolution rather than a revolution. We’re certainly seeing a lot of virtualization in the open RAM space. There’s definitely virtualization in cloud hyper scalers with regards to value added services, whether that be security services or other types of applications. And again, the telcos are kind of caught in the middle a little bit and need to really harness and define their roles as to exactly what it is that they’re gonna be providing and offering to businesses. Because it’s a hyper connected world, but it’s also hyper competitive.

How do you think that virtualization will affect how we’re building and designing these networks?

I’m right in the heart of it with DriveNets. What virtualization brings operationally is it’s just a completely different transformation from a kind of a very fixed, modular legacy, monolithic type approach. It’s very much a chassis where you’re always limited by the capacity of that particular box. You can knit them together and you can get more scalable, etc. but you’re still dealing with everything box-by-box. What we see with virtualization and the advent of white boxes that are carrier class is the ability to really disaggregate and virtualize the entire network. The entire network effectively acts and as if it’s one router, for example, or one switch will be at a scale of distributed different countries, it doesn’t matter.

There’s a whole raft of operational benefits in terms of the way you manage that; everything down to the logistics runs on one or two different bits of white box software as opposed to five or six different routing boxes to do different functions. When you combine functions to offer a multi service approach the network becomes like a cloud, and that’s the bit that virtualization really brings. Then you can just drop containers into those networks, whether it’s a third party firewall, an edge router, a core router… It’s all interoperable, and it’s truly scalable. You just add more white boxes as you want more scale. The huge demand for capacity, has been driven by the ability of the technology to enable that.

To find out more about James Morgan’s insights on connectivity media, listen to the full episode of The Connectivity Matters Podcast.

We sit down regularly with some of the biggest names in our industry, we dedicate our podcast to the stories of leaders in the technologies industries that bring us closer together. Follow the link here to see some of our latest episodes and don’t forget to subscribe.     

On the first episode of our newly rebranded The Connectivity Matters Podcast, we were joined by Chris Lennartz, the Vice President of Product Management for mobile services iBASIS. He’s had an incredibly successful career within the telecommunications industry, spanning over the last 20 years with a focus on strategic planning, business development and product management. Chris has also recently been voted number 18 in the top 100 most influential figures in the telco industry. We asked him for his insights on technology in the industry. 

What technologies do you think have had the biggest impact on the industry?

That’s a very simple one, it’s IP, because when I started, nothing was IP. Everything was TDM, everything was circuit switched. When IP came it completely revolutionised not only the cost of the network but the nature of the network and the way that the telecoms network interacted with a lot of different players, because suddenly the internet and telecommunications became one and a lot of new players came in. All the cloud stuff that we’re discussing right now started back in the 90s when the internet was opened up, and that completely revolutionised the telecoms industry. I know it’s been a long time since then, but if you ask me, that has been the biggest disruption and the biggest impact to the industry.

What impact do you think 5G has had and where do you see that progressing?

That’s a comparably disruptive impact. 5G is not only taking IP to a whole new level, it’s an evolution and revolution at the same time. If you look at evolution it’s more bandwidth. 4G already has more bandwidth than 3G, and 5G has more bandwidth than 4G, and that keeps going on and on. The most important thing about 5G in particular is that it revolutionises the way we look at networks. 4G was still a point network, then the whole virtualization thing came around. Now with 5G, you can make virtual networks from end to end spanning multiple networks. Network slicing will become very interesting if you look at specific IoT service providers or enterprises that need specific end-to-end connectivity with the specific quality of service or other parameters that you would normally do with a dedicated network. Now we can do it over the 5G network and just reserve a specific price for that, which makes it very interesting to have really one network that does it all. That will change the way we use the telecoms networks. 

The fourth industrial wave is really building on what 5G can give us. With the advent of private 5G networks, if you look at the predictions for them, it could be hundreds and even thousands of private networks that are being built. For example, it’s logical for airports to have their own network, given the fact that you have so many tourists or travellers in general to transport, all their suitcases to transport, on top of all the logistics and fleet management stuff that you need to do. How great would it be to have a network for yourself to do that? This proliferation of private 5G networks could really give a new boost to the way we automate the industry. 

Is there any tech on the horizons of either 5G or moving even further into the next generations that you’re most excited about in particular?

From a roaming perspective, it’s very exciting. It’s also challenging because on one hand, there will be even more bandwidth, but on the other hand there’s yet another new technology which is coming in. What we did right when LTE came along was go from SS7 signalling to diameter signalling. It was really good for us because at that time, we didn’t have market share, and disruption is always good for the challenger and not good for the leader. At that time, we were challenging the way things were so we took the opportunity to disrupt IP committee signalling, and we made a name for ourselves, and we built a solution earlier than anybody else. That’s why we became number three. 

Now, with 5G, this has happened again. This time we are leaders, which means that it can be a threat and opportunity at the same time. We’re going from diameter signalling to something that is called HTTP to signal, so yet another disruption. There will be a lot of companies that will think ‘hey, wait a minute, what I basically did 10 years ago, we can do now. So let’s do to them what they did to others.’ That’s challenging, but we need to challenge ourselves, it’s also an opportunity for ourselves, right? Because we are still number one, we can still get a lot of market share by doing this game right again, but there’s a lot more disruption this time to 5G than just the signalling changing its name. 

There’s also a lot more interesting use cases in roaming that will make you rethink the way we think about roaming. Today roaming basically means the traffic goes home from the visitor network, and is being handled there by the home operator. That takes 100 milliseconds in some cases, if you have to go from Singapore to Amsterdam, for example. However, if you have a self-driving car, that is simply not acceptable – the traffic can’t go halfway around the globe and then come back because that 100 millisecond is far too long. The latency of 5G for some specific use cases that all have to do with machine to machine or IoT should go down to less than 10 milliseconds. That means that traffic will need to stay in that specific region or in some cases, very close to the base station. That means that you have to start working with MEC on local breakout and applications being run, and very close to the base station. That means a reevaluation of the way we think about roaming. It’s going to be very interesting to see how all of these varieties of uses will need a variety of solutions. The way roaming works will be very different than 10 years ago when there were no devices, but just people going in on a day and using their phone.

To hear more about Chris Lennox’s insights into developments in the connectivity industry, listen to the full Connectivity Matters Podcast episode here. 

We sit down regularly with some of the biggest names in our industry, we dedicate our podcast to the stories of leaders in the technologies industries that bring us closer together. Follow the link here to see some of our latest episodes and don’t forget to subscribe.