An Interview with Ciena’s Brian Lavallée

prx author 194 Lavallee

Editor’s Note: For most of the first two decades of the submarine fiber optic cable era, technological development followed a fairly steady process during which bandwidth capacity gradually increased to meet the modest demand growth of the pre-Internet era.  That ended in the last decade when the Internet really got going.  Demand for video streaming and other high-bandwidth services rushed in, pushing submarine network vendors to accelerate to the development of new technologies to rapidly increase capacity.  While system capacities moved from Gigabits to Terabits, new ways of thinking about the management of the cable systems emerged as well. 

SCW recently had the pleasure to speak with Brian Lavallée, Senior Director of Portfolio Marketing with Ciena.  The following are his views on how submarine fiber optic systems technology has reached the point we are at now and what the future might hold. 

Brian Lavallée: Coherent technology started about 10 years ago.  It is ironic that it came out of the wireless space.  Ciena converted that technology into the optical space and the industry never looked back. 

At that time, the industry used “on/off keying,” which was basically just shuttering the laser light on and off.  It worked fine up to 10G, but then we found that at 40G and beyond, shuttering the light too fast gave us a whole set of problems.  So on/off keying was set aside. 

Then we had something called “coherent detection,” which enables you to send information in a completely different way.  That modem technology was first deployed over terrestrial networks.  Some of the first long routes to be turned up were from the East to West Coasts of the United States.  We then were invited to test it on submarine cables, which we did. 

We tested our terrestrial technology on a submarine cable and it actually worked really, really well.  Then the proverbial lightbulb went off and since then we’ve been optimizing our coherent technology for submarine networks.  It’s based on the same core technology as terrestrial but obviously we adapt it to the unique performance attributes of submarine cables, with their much greater distances, and to accommodate the differences in submarine versus terrestrial cables. 

Each submarine cable has a unique personality.  That is why when you use the same technology on different cables, you get different ultimate capacities on each cable. 

So coherent technology came into the submarine cable market about 10 years ago and now it is the de facto standard to upgrade any cable, old or new or even future cables.  As the bandwidth grows about 40 percent per year in all regions, the technology of choice for use in these cables is coherent modems.

Then there is Shannon’s Limit, which effects every type of communications, whether you’re a wireless carrier or a submarine cable operator.  There is a limit to how much information you can squeeze into a communications medium, in our case being the core of a submarine optical fiber. 

We’re getting close to Shannon’s Limit in the submarine cable industry.  We’re getting close to how much information we can send down that individual optical fiber.  We keep developing a whole bunch of technologies that complement coherent detection as it pushes very close to Shannon’s Limit – forward error correction, linear and non-linear mitigation techniques, etc.  And once we hit that limit as an industry – unless somebody comes up with something radically new – it means more submarine cables will have to be deployed.  Once you hit that maximum limit on the fiber, that’s it.  So, the only way to increase capacity is to start laying more cables at some point in time. 

The way submarine cables are managed right now is very similar to the way terrestrial systems are managed.  One thing that Ciena has been the pioneer in is that we have taken terrestrial technology and adapted it in the submarine space.  We started with coherent detection and intelligent mesh protection switching, which together we call GeoMesh, and the same thing is happening now in management. 

The traditional way of managing a network in the old school would be with manual offline tools – ledgers, spreadsheets, etc.  Most operators now use network management system software.  The big difference that has happened in the last couple of years is the introduction and increased adoption of automation and analytics.  These are very tightly coupled together.  The automation allows you to basically point and click to endpoints and the network will automatically turn up that service.  We’re talking seconds to minutes, when in the past it would have taken months.  It’s quicker and it’s also more secure.  It is machine-to-machine communications so the chances of an error being made are extremely small.  When it was done manually, there was always the possibility that something could go wrong.  Automation is becoming increasingly important part in submarine cable management and that came out of the terrestrial side of the network.  Also, you have to keep in mind that a lot of submarine cable operators own terrestrial networks as well. 

The other part is the analytics.  You must have a highly instrumented network and you have to be able to extract a lot of data from your network.  And that can be anything – it can be real-time capacity, bit error rate, or laser power.  It can be anything that you can measure and extract from the network.  You would feed that into an analytics engine and it would provide you with actionable insights. 

One example would be if transmitter number 35 is starting to show a trend to higher bit error rate, what do you do?  Chances are it is about to fail.  You can go out into the field and swap out that card before it has an outage, which will affect your customers.  You can actually address failures before they occur.

And that is where the analytics part comes in.  Analytics can show you trends in your network such as where it’s not balanced; where you are putting too much traffic on one submarine cable and not the other.  Analytics allows network operators to know “what is not normal” in their network and can help with a whole variety of things.  And because it is extracting, analyzing, and crunching that data in real-time, you don’t have to go looking for an insight or a problem before it happens – the network will actually tell you. 

We also have something we call “closed-loop automation.” The analytics engine listens to the streaming telemetry sensor data extracted from the network and the orchestration part of your network acts upon the provided insight based on secure and trusted policies developed by the network operators.  To use the example I mentioned where a transmitter is about to fail, you can take that information to automatically re-route around that particular transmitter, if possible, and send a message to the operations team to go investigate this transmitter.  When you replace it and the network can detect that the new transmitter is online, and after a certain amount of time that it determines that the transmitter is working well, it can re-configure the network and put traffic back on it.  This would be totally automated.  That’s where the industry is going.  We call that “The Adaptive Network”, which will be able to self-diagnose, self-optimize, and self-heal over time based on ever-changing network conditions. 

And that’s why I say, “Listen to your network.”  Because who better than the network itself to tell you what’s going on in your network?