Earlier this month, Ciena held a webinar entitled State of the Global Submarine Cable Network.  At this webinar, Brian Lavallée. Senior Director, Submarine Network Solutions at Ciena, gave a presentation on protecting submarine cables.  The following is our coverage of his talk.

Brian pointed out that there is a lot of new capacity being added on submarine cable networks, in large part by the COVID-19 situation.  “How much of that is directed attributable to COVID-19 versus normal growth is hard to determine, but we have seen strong growth to date.  Ciena introduced its WaveLogic5 earlier this year and a significant portion of those shipments went to submarine cable operators around the world,” Brian noted. 

“Protecting submarine cable networks is important because there is no Plan B,” he continued.  “This is what we have.  This is how we connect continents.  Some people may say that satellite communications networks is a Plan B, but I would say that while satellite can complement submarine, it can’t replace it.  We’re talking about economies of scale and levels of capacity that the satellite constellations just can’t maintain pace with.  So there is a place for satellites, but they don’t replace submarine.  Since there is no Plan B for submarine cable networks, we have to protect what we have and the additional capacity that we continually turn up on new and existing submarine cable networks.” 

To protect the submarine cable networks, we first have to understand the threats.  To show this, Brian reviewed data from SubCom, previously presented at SubOptic, that between 2016 and 2018, external aggression made up 87-92% of submarine cable faults. Next was manufacturing issues (defects in the equipment, for example) at 5-10% and “other” causes makes up the rest.  He noted that data from SubCom, ASN and Global Marine vary only slightly. 

Within the category of “external aggression,” Brian further broke down the data.  Marine activity is the predominant cause of cable faults with 84% of the total (fishing activities accounting for 49% and anchoring 35%).  Geological activities, abrasion and “other” make up the rest.  Brian also noted that 71-80% of all faults occur in less than 100 meters of water. 

The SubCom data show that there is a downward trend in submarine cable faults.  In shallow water, at under 1,000 meters of water depth, the trend line has gone down to about 0.26 faults per year per thousand kilometers, as compared to 0.4 (the 8-year average from 2010-2017).  This is from SubCom’s database only and covers from 2010 to 2018.  In deep water, greater than 1,000 meters, the fault rate has remained stable at about 0.04 per year per 1,000 kilometers. 

“But while the frequency of cable cuts is going down, the impact of each cut is increasing,” Brian said.  “This is because the new cable systems have so much more capacity.  They’re carrying orders of magnitude more capacity across these cables.  And with the advent of spatial division multiplexing (SDM) cables, we’re talking about incredible capacities.  So if you cut a newer cable, the amount of capacity that is being cut is far higher than an older cable.” 

Brian noted the efforts of the International Cable Protection Committee (ICPC), which has several proactive initiatives for cable protection, including cable protection agreements, education, etc.  Many of these initiatives are applicable to new cables, including:

  • Wet Plant Material Improvements
  • Wet Plant Optical Improvements
  • Wet Plant Electronics Improvements
  • Improved Manufacturing techniques
  • Improved Cable Deployment
  • Improved Route Diversity
  • Deeper Submarine Cable Burial.

“This is all on a proactive basis for new cables.  Once a cable is buried, you don’t want to pull it up to make these improvements,” he added.  “The better way to do it is to try to avoid these issues before they occur.” 

Beyond these methods, the industry can protect both old and new cables by going to Mesh Network Protection.  This can include overland and undersea as you can seamlessly integrate the terrestrial and submarine networks for alternative paths.  This looks at the network holistically.  If there is a cable cut, you can reroute that traffic to another cable.  In a more extreme case, you can reroute that traffic over a terrestrial route and then hop onto another submarine cable.  This is a way that you can actually prevent these issues from occurring. 

“The reason why you want to look at mesh protection is because faults still will occur,” Brian said.  “The trend is going downward and a lot of that has to do with the ICPC, but stuff still happens.  By putting Mesh Protection, you’re actually going beyond the proactive work and having a reactive solution to cable outages.  Mesh protection has been and is being deployed in many areas around the world.” 

A newer type of proactive protection is analytics-based preventative maintenance.  This is addressing issues before they occur.  To do this, you will listen to your network.  “The networks of today are highly instrumented, so you can extract data to see how the network is doing,” Brian continued.  “An example is a transmitter that is starting to show signs of impending failure.  It’s still working, but the error rate is going up and it’s corrected by forward error correction, so that the end user doesn’t see anything, but the operator can see that the card is correcting more errors than it should, because the laser is about to die.  So you can do preventative maintenance, reroute traffic around that laser, replace it, then reroute traffic back and the customer just sees an available network and all is good.  The more data you can collect along that entire surface path, the better insights are provided by the analytics.  This is a critical area because you are actually addressing issues before they occur, specifically on the equipment side. 

“So you want to protect the submarine cable network from a variety of perspectives.  From our perspective, we have a vision called the adaptive network, which covers both overland and undersea.  We make use of three key tenets.  First, programmable infrastructure.  We call this GeoMesh Extreme.  You can move traffic around, you can upshift wavelengths, but you can also access telemetry data.  That is real-time data that is fed into analytics.  By using algorithms you can come into actual insights.  By using the data that are continually streaming from the transmitters, you can begin to see when one of them is approaching failure and predict those failures.  On the software control and automation side, you can decide exactly what to do.  You can automatically reroute around the transmitter or have a human make the final call and go out and replace the card.  It doesn’t have to be a fully closed loop.  It can be up to the operator how much control to give to the network.  This is the way we see operators moving in both terrestrial and undersea networks,” Brian concluded