Telia’s Fridström explains how fibre providers can avoid a subsea sinking feeling
Subsea cable capacity will provide the backbone capacity to enable the vast number of new business requirements that are arriving with 5G, artificial intelligence, machine learning and within IoT. However, price pressure on cable providers, the demise of the consortium model and continued transformational innovation in technology are radically altering the operational mode – and economics – for subsea cable owners. Mattias Fridström, the chief evangelist at Telia Carrier, tells George Malim what this new mode of operation looks like and the challenges subsea cable operators face in meeting the new needs of customers while maintaining profitability.
George Malim: Although the hype is all about 5G and IoT, the new services and apps enabled will be dependent on vast subsea capacity to connect people and things globally. How will this demand be satisfied?
Mattias Fridström: I’m confident that basically everything, in the end, will be connected. There are new cables being built. In the Atlantic now, for example, there are new cables that are coming in this year and next. The demand is there are we are building to the right size. The increased demand from 5G and IoT is going to be satisfied with new cables.
The technological cycle is much faster today in terms of the advances being made in equipment. In the past you used to have a seven-year cycle with a dense wave division multiplexing (DWDM) system and that was good because you could invest and know that you had that time to make a return on your investment. The pressure now is that new technology is going to come in far sooner in that cycle that enables newer cables to operate at lower cost.
GM: Undersea cables are extremely costly to deploy and because of this, the traditional approach has seen consortia formed to enable the stakeholders to share the expenses of deployment. However, there now seems to be a move away from this model. How do you see this affecting subsea capacity?
MF: I think you’re right. Although there are obviously still a number of consortium-led cables being built, there are fewer now that in previous years and I think we will continue to see less consortia-led activity. In the Atlantic, the biggest demand for traffic we have is from the web giants and you have to ask the question of why, given their financial strength, they would buy from us rather than build cables themselves?
From our point of view, they’d demand pricing from us that would be very close to our costs so there’s no great advantage in use deploying capacity to meet their needs. I see the web giants financing cables because they are the ones that have the traffic growth.
GM: In what ways will subsea network architectures change in future?
MF: Ten years ago practically every cable between the US and Europe went through the UK but the trend now is for diversity with cable from Spain, France and even Norway to the US. As the countries have demand for capacity, the locations for cables to land change. On top of this, two or three years ago, everything we did was manual and whenever we had a fault we have to have people come in and fix it.
At that time we had ten cables but the new tools we have now mean we can simulate how to re-route traffic if there is an issue and it becomes apparent that you could probably use only five cables to support the same traffic.
The new architecture is more diverse and the land-based network covers the diversity. We have spread our traffic across fewer cables than before and that makes it easier to spend money on more targeted capacity.
GM: Do you see SDN, machine learning and artificial intelligence (AI) helping to maximise utilisation of subsea capacity?
MF: Absolutely. There’s a lot happening within AI and machine learning to help maximise utilisation. We’re testing technologies now and it’s clear that people want data. We have data about the performance of the network and with that you can start to predict demand and issues.
The SDN and AI capability is just being touched on in our world but it’s going to make a dramatic change. For example, in the past we never knew how fibre would degrade over time and couldn’t call the end of life of a cable accurately. Now, we’ll be able to follow the development of our fibre capacity in real-time and be able to squeeze out more from our fibre capacity and simply know more about it on land and on sea.
GM: How can you plan for the future and ensure subsea capacity is flexible enough to meet the unknown needs of the future?
We’re trying every day to see how much flexibility we can build into the network without it costing 50-60% more to build it. No one really knows how much traffic 5G will pour into the backbone – the first traffic will be local in situations such as factory networks so it doesn’t provide an indicator. We’re trying to be as flexible as we can.
Building flexible networks is not simple and we’re also looking at the possibility of providing self-service. However the amount of capex needed to enable that is enormous.
Price pressure in the market is really tough. You need the latest technology all the time to keep up.