The meteoric rise of an Internet backbone

Thursday, June 9 2016


In many ways, it is possible to draw parallels between the evolution of the commercial Internet backbone at the beginning of the millennium and the heady days of space exploration back in the 1960s and 70s – one of these being that they were both essentially a game of ‘catch-up’ between superpowers. In an Internet context, this was initially played out between competing incumbent operators, all desperate for a slice of the global Internet ‘promised land’ and more recently, it has shifted towards a tug-of- war between the content providers and those facilitating access to it.

The fundamental difference though, is that while space exploration is for the most part a fairly clear-cut (albeit complex) path of development towards a series of well-defined and highly tangible goals (the first man in orbit or the first to reach the moon and perhaps soon even the first to inhabit Mars), it has never really been fully apparent quite where the Internet is headed or perhaps more importantly, where it will take us mere mortals. Indeed, its humble end-users have more often than not been left behind as the connectivity juggernaut rolls on, in a dust cloud of technical naivety and left to the mercy of modern day cyber pirates.

Closer to home, if we look at the TeliaNet backbone topology from 1997, the idea of our current pan-continental, market leading (#1 ranked at the time of writing) terabit backbone would have seemed almost inconceivable. Then again, the same can be reasonably said of Donald Trump as 2016 republican front-runner or Leicester City as Premier League champions. The fact is though, you can now get more bandwidth from a domestic fiber broadband connection than the aggregated capacity of our entire network core in 1997:

Image 1

Apart from the negligible bandwidth between core network hubs or PoPs, the very fact that Stockholm was connected to London directly and not via Copenhagen (across the bridge subsequently built between Sweden and Denmark), combined with a fundamental reliance on upstream transit reflects a changing and increasingly connected world, in the truest sense. Of course, a dig through Europe and acquisition of some significant fiber assets, together with razor sharp focus and a passionate organization have also played a leading role in our growth story.

Back in December 2008, Renesys (who have since been acquired by Dyn Internet) published their first annual ‘Baker’s Dozen’ report about the 13 backbone providers who frequent the top ten of their global rankings. A lot of water has passed under the bridge since then and many of the now defunct networks (in name at least) in the 2008 report allude to a dramatic and for some, sentimental history lesson about a global carrier industry undergoing constant and often brutal change. The old guard -Tiscali, Global Crossing and Qwest languished below a seemingly invincible pair of networks at the top of the list, Sprint and Level 3. Then way down at the bottom of the ‘middle cluster’ (at least in January 2008) sat TeliaSonera International Carrier at a modest #8. While the announcement of our new name, ‘Telia Carrier’ on April 13 th also confides our own identity in the 2008 report to the ‘where are they now’ files, we are still very much alive, well and looking back fondly upon nearly a decade of staggering growth:

Image 2

Source: Dyn

And this is something that was reinforced in the latest Dyn report. In fact it wasn’t just reinforced – the authors seemed unable to hold back their praise for the rapid growth of Telia Carrier’s customer base and ascent through the global Internet backbone rankings during the past two years: ‘The biggest story of 2015 was the meteoric ascent of TeliaSonera (#2) as it challenged Level 3 for the #1 global ranking, a position Level 3 has held without interruption since the end of 2008. TeliaSonera actually surpassed Level 3 briefly in 2016 and is a fraction of a percent below them as of this writing, an achievement we never could have imagined.’ Dyn Baker’s Dozen, 2015 Edition

Image 3

Source: Dyn

In the 2009 report, Renesys included a carefully formulated disclaimer that the rankings and their market share do not automatically imply profitability or even a well-run, resilient and stable network. With the benefit of hindsight though, history has taught us that these things are mutually inclusive – at least over the long term.

In the June 2009 release of the Cisco Visual Networking Index, it was predicted that Global IP traffic would grow 5 times during the period 2008-2013 and that overall, IP traffic would grow at a compound annual growth rate (CAGR) of 40 percent. Looking at our own network growth stats, customer traffic growth in AS1299 from Jan 2009 to December was in fact more than 10 times, or double that predicted by Cisco. This demonstrates that not only did the VNI predictions hold up (in our network at least) but it also alludes to an overall increase in IP transit market share for AS1299, as reflected in our rise through the Dyn rankings.

As our CEO, Brendan Ives, recently pointed out – this is a fantastic achievement that few, if any of us, could have really imagined until very recently. It is also testament to our unstinting focus on quality and customer experience, and the result of a lot of hard work by the company as a whole.

There is though, something else that I believe has played a pivotal role. One of the main ingredients of our success comes down to the fact that, unlike many others in the industry, once the race started, we decided NOT to simply follow the crowd and get caught up in a game of unsustainable ‘superpower catch-up’ with our peers.

Instead, we focused relentlessly on doing what we do best, wholesale. As the carriers’ carrier, we just kept on connecting businesses and societies as they matured into fully blown digital natives, with an insatiable appetite for connectivity. As it turns out, we have also established a fairly unique position amongst our peers in that that our growth is almost exclusively organic, without the inevitable crutch of consolidation that has so often punctuated our industry.

At the end of the day, it’s been a rollercoaster ride and for those of us fortunate to be part of it, a once-in- a-lifetime opportunity – indeed, people still refer fondly to the halcyon days when the carrier revolution started.

The big question for many now is, who will be the Internet superpowers of the future and what will they look like? Further consolidation and convergence is not a matter of ‘if’ but ‘when’. Bandwidth growth shows no signs of abating either, nor does disintermediation of the Internet transit cloud as we know it today. That said, if the lessons of Internet history are anything to go by there is certain to be more traffic and a fair number of surprises ahead.