‘The reality is that the cloud is actually under the ocean’


13 Feb 2019

Image: © norimoto/Stock.adobe.com

Nicole Starosielski from New York University explains how, even in a Wi-Fi world, the internet still depends on undersea cables.

A version of this article was originally published by The Conversation (CC BY-ND 4.0)

Recently, a digital blackout in Tonga caused by the severing of the country’s only undersea cable generated widespread recognition of the submerged systems our connected world depends upon.

Not many people realise that undersea cables transport nearly 100pc of transoceanic data traffic. These lines are laid on the very bottom of the ocean floor. They’re about as thick as a garden hose and carry the world’s internet, phone calls and even TV transmissions between continents at the speed of light. A single cable can carry tens of terabits of information per second.

While researching my book, The Undersea Network, I realised that the cables we all rely on to send everything from email to banking information across the seas remain largely unregulated and undefended. Although they are laid by only a few companies – including the American company SubCom and the French company Alcatel-Lucent – and often funnelled along narrow paths, the ocean’s vastness has often provided them protection. When one is broken, as the Tonga cable was, data traffic comes to a halt.

Far from wireless

The fact that we route internet traffic through the ocean – amidst deep-sea creatures and hydrothermal vents – runs counter to most people’s imaginings of the internet. Didn’t we develop satellites and Wi-Fi to transmit signals through the air? Haven’t we moved to the cloud? Undersea cable systems sound like a thing of the past.

The reality is that the cloud is actually under the ocean. Even though they might seem behind the times, fibre optic cables are actually state-of-the-art global communications technologies. Since they use light to encode information and remain unfettered by weather, cables carry data faster and cheaper than satellites. They criss-cross the continents, too – a message from New York to California also travels by fibre optic cable. These systems are not going to be replaced by aerial communications any time soon.

A vulnerable system?

The biggest problem with cable systems is not technological – it’s human. Because they run underground, underwater and between telephone poles, cable systems populate the same spaces people do. As a result, they’re accidentally broken all the time. Local construction projects dig up terrestrial lines. Boaters drop anchors on cables. And submarines can pinpoint systems under the sea.

Most media coverage about these systems has been dominated by the question of vulnerability. Are global communications networks really at risk of disruption? What would happen if these cables were cut? Should we all be worrying about a digital blackout – whether caused by accident or terrorists?

The answer to this is not black and white. Any individual cable is always at risk, but likely far more so from boaters and fishermen than any saboteur. Over history, the single largest cause of disruption has been people unintentionally dropping anchors and nets. The International Cable Protection Committee has been working for years to prevent such breaks.

As a result, cables today are covered in steel armour and buried beneath the seafloor at their shore-ends, where the human threat is most concentrated. This provides some level of protection. In the deep sea, the ocean’s inaccessibility largely safeguards cables – they need only be covered with a thin polyethylene sheath. It’s not that it’s much more difficult to sever cables in the deep ocean, it’s just that the primary forms of interference are less likely to happen. The sea is so big and the cables are so narrow, the probability isn’t that high that you’d run across one.

Sabotage has actually been rare in the history of undersea cables. There are certainly occurrences (though none recently), but these are disproportionately publicised. The World War I German raid of the Fanning Island cable station in the Pacific Ocean gets a lot of attention. And there was speculation about sabotage in the cable disruptions outside Alexandria, Egypt, in 2008, which cut 70pc of the country’s internet, affecting millions. Yet you hear little about the regular faults that occur, on average, about 200 times each year.

Redundancy provides some protection

The fact is it’s incredibly difficult to monitor these lines. Cable companies have been trying to do so for more than a century, since the first telegraph lines were laid in the 1800s. But the ocean is too vast and the lines simply too long. It would be impossible to stop every vessel that came anywhere near critical communications cables. Nations would need to create extremely long, ‘no go’ zones across the ocean, which itself would profoundly disrupt the economy. Even then, the cables could still be at risk from undersea landslides.

There are only several hundred cable systems that transport almost all transoceanic traffic around the world. And these often run through narrow pressure points where small disruptions can have massive impacts. Since each cable can carry an extraordinary amount of information, it’s not uncommon for an entire country to rely on only a handful of systems. In many places, like Tonga, it takes only a single cable cut to take out large swathes of the internet. If the right cables were disrupted at the right time, it could disrupt global internet traffic for weeks or even months.

The thing that protects global information traffic is the fact that there’s some redundancy built into the system. Since there is more cable capacity than there is traffic, when there is a break, information is automatically rerouted along other cables. Because there are many systems linking to the United States, and a lot of internet infrastructure is located here, a single cable outage is unlikely to cause any noticeable effect for Americans.

Any single cable line has been and will continue to be susceptible to disruption. And the only way around this is to build a more diverse system. But, as things are, even though individual companies each look out for their own network, there is no economic incentive or supervisory body to ensure the global system as a whole is resilient. If there’s a vulnerability to worry about, this is it.

The Conversation

By Nicole Starosielski

Nicole Starosielski is assistant professor of media, culture and communication at New York University. Her research focuses on the global distribution of digital media, and the relationship between technology, society and the aquatic environment.

This is an updated version of an article originally published by The Conversation on 3 November 2015.