By Geoff Bennett, Director of Solutions & Technology for Infinera

Next Gen Submarine Cable Architectures

The demand for submarine cable capacity is incredibly high.  In fact, from 2015 to 2019, submarine fiber design capacity on major routes has increased at a Compound Annual Growth Rate (CAGR) of 32 percent, including upgrades and new system builds (sources: Subtel Forum, Telegeography).

With that level of growth set to continue, can we build new cables fast enough?  Should we try to squeeze more capacity out of existing cables?  And if we do build new cables, what can we do to increase the capacity beyond what is possible today?

Those are big questions but let me start with a transatlantic cable that most people agree represents the state of the art today - the MAREA cable, owned jointly by Telxius, Facebook and Microsoft according to TeleGeography.

MAREA is so highly regarded because it consists of an optical fiber that is highly optimized for coherent transmission at high modulation order -16QAM (quadrature amplitude modulation) modulation is used on at least one of the fiber pairs of MAREA, and this is a first for production grade transatlantic transmission.  In addition, MAREA's optical amplifiers operate at very high-power levels and are more closely spaced than normal at around 55 kilometres.

By using extremely high performance transponders, such as Infinera's ICE4, the commercial capacity on a single fiber pair on MAREA is at least 24 terabits per second (Tb/s). Note that up to 26.2 Tb/s of capacity was achieved in trials of ICE4, but operators will decide the optical budget safety margin required.

MAREA was deployed with eight fiber pairs and it's interesting to note that each one of these pairs delivers as much capacity as all the operational transatlantic cables in service at the time.  Even so, the demand for subsea capacity is so great that we must plan now to meet the needs of the near future.

You may be wondering why not deploy more fiber pairs in a cable such as MAREA.  The answer is that the amplifier chains on submarine cables must be powered by setting up huge voltages at each end of the cable - positive at one end and negative at the other.  Based on the power level used in MAREA's amps, and the fact that the amps are so closely spaced, there simply isn't a way to get enough power into the cable to supply more fiber pairs.

When MAREA was designed, the goal was to maximize the capacity per fiber pair - and this was a great success.  For the next generation of transoceanic cables, the focus is shifting to maximize the total cable capacity, not necessarily the capacity per fiber pair.  So how do we do this?

The limiting factor is total electrical power.  One step would be to lower the power level of the amplifiers and increase the spacing.  This would mean that conventional modulations like 16QAM would not be able to close the link.  In the past we would need to dial down the modulation to 8QAM or QPSK (quadrature phase shift keying), but modern coherent technology introduces a capability called probabilistic constellation shaping (PCS) that can smoothly optimize the modulation efficiency to the optical budget of the link. 

By operating in a lower power regime, a new cable architecture emerges, known as Space Division Multiplexing (SDM).  The rationale is that, while there is a small reduction in the capacity per fiber pair with SDM, it's possible to increase the number of pairs by 50-100 percent and achieve a much higher total cable capacity in future submarine deployments.

Google recently announced plans to bring the first such cable into service next year - the Dunant cable between the East coast of the USA and a landing point in Western France.  The total capacity for Dunant is forecast to be 250 Tb/s over 12 fiber pairs, compared to a potential 192 Tb/s for MAREA's 8 fiber pairs.

In a separate webinar with speakers from Facebook, Infinera and Corning, Steve Grubb, Global Optical Architect for Facebook, described some of the advances needed to achieve a transatlantic cable capacity of 1 petabit per second using SDM techniques!  A recording of this webinar is available here.

So, it does appear that the next wave of subsea cables will operate with an SDM architecture, but what about current generations of subsea cables?

For systems like MAREA, with large area, positive dispersion fibers, there may still be some room to increase capacity - perhaps by 25 to 50 percent with Infinera's next generation of optical engines.

Before cables like MAREA were deployed there were tens of "dispersion managed" submarine cables in operation - including a very early transatlantic cable that first went into service in 2003.  Recently Infinera's ICE4 technology was used to double the existing capacity on that cable with the result that the commercial life will be extended yet again.

Similarly, there are hundreds of small cables that are laid without amplifiers to cover short underwater distances across straits, between islands, or laid in festoon routes along coastlines between cities.  The capacity of one such cable, operated by OTEGLOBE, was also doubled recently using ICE4 technology.

The demand for subsea capacity is set to continue, and one way or another we must meet that demand.  SDM represents a great way to increase cable capacity for future long-distance cables, but high-performance submarine transponder technology allows more capacity to be extracted from existing cables of all types.

##

About the Author

Geoff Bennett is the Director of Solutions & Technology for Infinera, a leading manufacturer of Intelligent Transport Network solutions.  He has over 25 years of experience in the data communications industry, including IP routing with Proteon and Wellfleet; ATM and MPLS experience with FORE Systems; and optical transmission and switching experience with Marconi, where he held the position of Distinguished Engineer in the CTO Office.  Geoff is a frequent conference speaker and is the author of "Designing TCP/IP Internetworks", published by VNR.