Industry Spotlight: Mattias Fridström On Arelion’s Network Evolution

It has now been more than a year since Arelion began its new life, leaving behind the Telia Carrier brand and facing the market as an independent entity. Over that year we have seen the company start to make more aggressive network investments, moving into new parts of the world, such as Mexico. With us today to give his perspective on Arelion’s network infrastructure, its future, and the state of the broader network infrastructure marketplace is Mattias Fridström, Arelion’s Chief Evangelist.

TR: What was your journey to your current role as Chief Evangelist, and what does that title mean to you?

MF: I started as a submarine cable engineer at Telia after earning my university degree in 1996. It was right before the SupOptic conference in San Francisco in 1997. During the event, four individuals announced they would build a private cable across the Atlantic called Atlantic Crossing. We went home from that conference thinking we should do something similar and quickly announced that we would build some new land cables in Europe. This, in combination with the traditional voice business and an early IP network, was the start of Telia International Carrier between 1998-1999. I was here when we established Telia Carrier and decided to build networks in both Europe and the U.S. I then took over the responsibility for building our network until the summer of 2002, when the dot-com bubble crashed. While many of our competitors disappeared or got bought, we were saved by a big and strong owner. We were essentially given a year (2003) to prove that we could take control of our costs while keeping most of the revenues. We managed to do that.

Since then, we have built a carrier business from the ground up based in part on the assets we built from 1998-2002. During 2003, I joined the business side of the company, moving from engineering to boost the business as we really needed to start selling services instead of just building networks. In 2010, I went back and became the CTO of Telia Carrier. Then in 2016, when I felt I had done my time again on the technical side, I wanted return to the business side. We needed someone in the management team who could spend time on what was actually going on around us: what are others doing, what are we doing, and how do we spread that word and our brand to even more customers? So we invented this Chief Evangelist role. In this role, I can meet everyone, not to sell but just to talk and find out what they are doing or missing and how we might help. I spend a lot of time trying to figure out where we need to go next but also how to let others understand what Arelion is doing and why we may be a good partner for them.

TR: How different has it been since the company split off from Telia as Arelion?

MF:  It’s quite different. From the outside, Telia Carrier looks very much like Arelion because we’re the same people operating the same network. But we have new owners who only care about us. Telia was a good owner. But for them, mobile services in the local Nordic and Baltic market will always be #1, and they should be. Sometimes, when we proposed IP network expansion, the Telia Management team preferred to spend the money on 5G instead. Now, with new owners, when we have a great opportunity to build something, they will look at the business case and make a decision based on the individual merits of that actual case. That’s the biggest difference. Obviously, there are still limits to what we can do, but we can at least ask and get a fair assessment. They are pension funds; they want long-term investments. But if we can prove to them that we can make money on something long term, then we have a good chance of them saying yes.

The other major difference is that when you are part of a big group, they want you to use one HR system, one financial system, etc. It’s hard to get those systems to work for every individual company across a large organization. Everyone needs to compromise a bit. The same is true when it comes to agreements with larger suppliers. The group will always focus on getting the best price on the boxes they buy the most. That may not necessarily be the boxes we would buy, and we would thereby not have the “perfect” price for us. Now, as a standalone company, all of that is up to us. We can build the IT systems the way we want the IT systems, which means they can be optimized for us. Systems are very different if you have 20 million private customers compared to if you have 2500 global customers, so our systems can be much more efficient, automated, and tailored to our needs.

TR: How has that flexibility translated into the way your network has evolved?

MF: The network itself is almost identical. We’ve never really been held back on the network apart from a more aggressive expansion. The way we build and operate the network will be the same going forward. We can make decisions a bit quicker and obviously focus better on the most critical supplier agreements for only us.

TR: What is currently driving changes in Arelion’s network infrastructure?

MF: With power prices going up, especially in Europe, suddenly it sometimes makes sense to re-engineer the network by removing and replacing power hungry gear. Five years ago, carriers expected to use a box as long as you could because it was considered already paid for. You wanted to sweat the assets, as we say. Now, the power consumption of that box might make it very inefficient. Therefore, we are focusing a lot on efficiency. The industry has historically also been very manual since you could serve your customers really well anyway. It was also very hard to automate things and the savings did not really justify the effort. But that has changed, and more automation is now coming. Networks are practically pouring out data and, if correctly used, you may find faults before they happen. That is the kind of automation we absolutely need to do.

Then, as real estate costs go up, the price for colocation goes up. Companies like Equinix and CoreSite are raising their prices. Therefore, we need to be more careful selecting the right locations where we will find our customers. The cost of adding another PoP to our network is higher now than it was before and, at the same time, we really need to be in more locations to find more customers. A complex challenge, really.

Being a large carrier, we always need to adapt because of new technologies. The development of 400ZR and ZR+ is extremely interesting for us, making it possible to run an IP network in a completely different way. That is going to dramatically change how operators run their business.

TR: How will 400ZR and ZR+ change the way you approach your network?

MF: In the old world, the port you have within the router is always connected to a transponder on the optical layer. Now, you can put the laser that used to be in the transponder in the plug itself within the router, which means that you don’t need a transponder anymore in the optical layer. You shoot the signal from the router directly over the optical network. Since we run an enormous IP network around Europe, North America, and the world, this gives us a big opportunity to reduce CAPEX spend on building the IP network. The traditional optical wavelength network will still be there for the companies we will serve with our wavelength service. That coherent optical footprint will be exactly like it was before, but the way you can run your IP network is going to dramatically change with 400ZR and ZR+. ZR+ is for longer distances while ZR has a very short range, and both are targeting maximum standardization. The target is to be able to buy your plug from any vendor and put it into any router in your network regardless of brand, which is obviously something we in the industry have looked forward to for a long, long time.  

TR: You mentioned opportunities for geographical expansion. Where do you foresee building out to at the fiber level?

MF: Right now, we are building into the Mexican market, leasing fiber. That is a big step for us as we do not often enter new markets with fiber. It’s going pretty well, but projects like this take a lot of time, and you need to learn new procedures for how to operate in countries you have never been in before. Going forward, there are a few countries in Europe where I could see us leasing fiber if the traffic increase is what we believe it may be. Beyond that, it’s still too early to go with our own fiber network into South America, Africa, the Middle East, or even Asia. I think we’re still better off leasing capacity from already built cables than trying to build into those countries ourselves. It is too expensive and too risky to take that step right now.

TR: Are there opportunities to build in the subsea markets currently?     

MF: We have traditionally built subsea cables ourselves in the Baltic Sea and the North Sea; I can see that happening again. But for transatlantic or transpacific cables, you, as an operator, are still better off leasing capacity from the companies that have built the new cables. Prices are still falling in these markets. It’s a risky business unless you are a hyperscaler and have an enormous amount of traffic yourself. For an operator to build a cable across the Atlantic, you would have to hope to sell capacity to the hyperscalers who are typically building the cables themselves. This may change though, as there are new technologies coming out, such as new fiber types that are slightly less efficient but cheaper, allowing you to have many more fiber pairs in one cable. New cables across the Atlantic will most likely host at least 24 fiber pairs. If you can lay a 24-fiber pair cable for the same price as you lay an 8-fiber pair cable today, then maybe the price of one fiber pair will be interesting enough for us to actually buy our own fiber. This may be especially true for fiber across the Atlantic and maybe even the Pacific. But that remains to be seen.

TR: Will you be looking to address more of the SD-WAN marketplace?

MF: We rarely say no to good opportunities, but we primarily focus on pure connectivity services. I think we are happily selling IP, Ethernet, and wavelength services and try to stay away from the more complex managed services.  We can sell SD-WAN to anyone, but that’s not our core strength, which is in higher order capacities. If an enterprise from the financial industry needs a lot of traffic between their SD-WAN nodes, then we would be perfect for that need. But if someone needs help connecting 4,000 fast food restaurants in the US, that’s not us. 

TR: Are there any verticals you see opportunities in this year?

MF: We’re very well known among the operators and hyperscalers around the world. But I think this is the year we will focus a lot more on trying to find the enterprise customers that have wholesale needs, for example, banks, manufacturers, or pharmaceutical companies in America who need to connect their facilities in Europe or their data centers in the U.S. They typically have the skillset to run their own internal network. We don’t need to do that for them, but we can sell them the connectivity.

TR: What challenges do you foresee in reaching those enterprise markets?

MF: The biggest challenge is still for us to be known. I think that very few companies in the enterprise sector have ever heard about Telia Carrier. Then we changed our name, and now absolutely no one has heard about Arelion. So, one of our biggest challenges is to become known. After that, we need to find these customers. What type of shows do they go to? Usually, when we find them and explain who we are and what we can do, the interest is there.

TR: In your talks with hyperscalers, other carriers, and everyone else, what comes up as the biggest challenges ahead for the network industry as a whole?

MF: I think one of the biggest issues in the whole industry is that Shannon’s limit is around the corner. There is a physical limit on how much traffic you can send down a fiber and this limit is referred to as Shannon’s limit. Five years ago, you could always replace your old DWDM system and get twice as much capacity, and it was fine to have one fiber pair everywhere. That is not the case anymore. Now, you need to find more fiber pairs. That’s a huge challenge for everyone right now, especially in the U.S., where there is less fiber in the ground than in Europe. Then, when traffic is growing, you need to add more fibers in parallel to the fibers you already have, and obviously a new fiber pair comes with new costs.

TR: How have recent global events impacted the way people think about network infrastructure?

MF: Until recently, security discussions were focused on logical security – what’s going on in the traffic and so on. I think this is the first year where the big topic has become the physical security of networks, both in the U.S. and Europe. The likelihood of sabotage is much higher today than ever before. Pre-2022, I’d never heard of any sabotage ever, but then we had some in Europe last year. There was the Nord Stream gas pipeline in the Baltic Sea, but there was also some sabotage around the Paris and Marseille area, where someone cut cables last year. Now, network owners and customers are afraid that this is the next step for terrorists. Instead of DDoS attacks, attackers might actually go out to physically disconnect or even destroy cables. That is something that all of us must consider this year: how do we protect our networks better? How do we limit the understanding of where they are physically installed? Five years ago, this type of network knowledge was regarded as “super” public information. And, of course, we want all the fishermen to know where subsea cables are, so they don’t cut them with an anchor or their trawl. But at the same time, we do not really want anyone to understand where they are – a difficult contradiction. 

TR: Thank you for talking with Telecom Ramblings!

If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!

Categories: Fiber Networks · Industry Spotlight · Internet Backbones