Sr. Director, Corporate Communications
TeliaSonera International Carrier (TSIC) joined us today in announcing the deployment of the Infinera DTN-X into TSIC’s European network. The Infinera Intelligent Transport Network, featuring the DTN-X packet optical transport networking platform, has already enabled TSIC to deliver 10 Gigabit Ethernet (GbE) and 100 GbE services to wholesale customers in North America and now TSIC is expanding the network into Europe.
TSIC, a leading global IP carrier, operates a wholly owned fiber optic network connecting 200 Points of Presence (PoPs) across Europe, North America and Asia. Customers include content providers and network operators globally for IP, DWDM, Ethernet, voice and specialized mobile services.
TSIC first selected the Infinera DTN in 2009. In November 2011, TSIC and Infinera performed the world’s first terabit trial in California based on DTN-X equipment. In June 2012, TSIC was named Best Global NGN Optical Network at the DWDM Awards in Monaco. In November 2012 TSIC announced the deployment of the DTN-X platform with Infinera Instant Bandwidth™ across its North American network.
TSIC is deploying an Intelligent Transport Network to scale network capacity, automate operations and reduce complexity by converging multi-layer switching with the industry’s only commercially available 500 Gigabit per second coherent super-channel transmission. The Infinera DTN-X is designed to scale without compromise to enable future upgrades to terabit super-channels and Terabit Ethernet.
“Stockholm to Hamburg is an important network route for us both for delivering services to our Nordic and Baltic end-users and for being part of the global long haul routes Europe-Russia and Europe-overland to Asia,” said Erik Hallberg, President of TSIC. “Our experience in North America is that Infinera gives us a competitive edge by enabling us to use time as a weapon to deliver 10 GbE and 100 GbE services faster than the competition. We look forward to expanding our Intelligent Transport Network into Europe with Infinera.”
We also announced today the Infinera Intelligent Transport Network™, an architecture for carriers to exploit the increasing demand for cloud-based services and data center connectivity as they advance into the Terabit Era. The Terabit Era envisions a highly connected global community sharing experiences and doing business at light speed, delivered by an infinite pool of intelligent bandwidth. The Intelligent Transport Network helps carriers use time as a weapon to increase revenues with reliable, differentiated services while reducing operating costs through scale, multi-layer convergence and automation.
In 2005 Infinera introduced the Digital Optical Network, an architecture based on the company’s breakthrough 100 Gigabit per second (Gb/s) Photonic Integrated Circuits (PICs). By integrating PICs into the DTN platform Infinera was the first to converge Optical Transport Network (OTN) switching and wavelength division multiplexing (WDM) in the same chassis. Within 18 months of the DTN’s initial deployment Infinera led the highly competitive long-haul optical transport market in North America. The Intelligent Transport Network builds on the foundation of the Digital Optical Network, expanding the original vision as service providers and internet content providers prepare for the Terabit Era.
Infinera Product Evangelist
It’s great when a customer sets you a genuinely interesting technical challenge, and it happened recently with DANTE, which is the organization responsible for planning, building and operating the advanced backbone network for the European Research and Education community. While Infinera was competing to win last year’s upgrade to the GÉANT backbone, we explained to the DANTE folks how Infinera is an innovative technology leader, offering the best solution on the market today to meet our capacity, flexibility and service lead time requirements.
As good scientists, once the network was being deployed, DANTE asked if we would be able to back our claims up with a real demonstration. We like challenges, so we decided to make a formal attempt to turn up 2 Tb/s of long haul super-channel capacity in the shortest time possible, over a key route in the GÉANT network from Amsterdam to Frankfurt (a fiber distance of 671 kilometers). One of the conditions was to make the test as “real world” as possible, and so we had to use DANTE’s own line equipment – which meant that every piece of hardware and software in this test is truly “production quality.”
To be clear about why this is important, this kind of test has never been tried before on this scale, and using genuine production equipment. This is because these sorts of “hero experiments” are often carried out with prototype gear that may have been altered to (amongst other things) switch off certain “safety protocols.” These protocols are designed to make sure that analog DWDM wavelengths do not disrupt existing waves on a long haul fiber and its associated amplifier chain. We knew that production-grade technologies such as Infinera’s DAMP (Digital Amplifier Protocol) would help us stabilize the lit capacity while still allowing us to turn up capacity really quickly.
This was a genuine test of our rapid provisioning capability, using real production equipment and software. If we had used conventional 100G transponders, we would need a total of 40 of them – 20 at each end. But the Infinera 500G solution allows an engineer to provision up to five times as much capacity in a single operational cycle. Enabling our customers to use time as a weapon is a key value of coherent super-channels.
In the true spirit of adventure, we went into this test without really knowing just how long the process might take to carry out. Since we had to have a goal to aim for, we decided that 20 minutes should be enough, and as a joke with DANTE we compared the duration of the process to cooking mashed potatoes. This is why you’ll hear DANTE CTO, Michael Enrico, refer to this comparison on the video we made with them.
By the time I arrived in the Amsterdam PoP, most of the hard work had already been done. The gear had been installed at both ends, and the amp chain was in place thanks to Infinera’s Professional Services team. Our Project Manager had somehow found a two-day window in the deployment schedule for us to attempt the test. We also had Craig Volp and Darren Clarke from DANTE on-site to make sure we weren’t cheating, and to act as official time-keepers.
So, while our on-site team of Steve Pegg and Kevin Elliot figured out the last minute details, all that was left for me to do was set up a few video cameras to record the test in detail, and to agree to just how aggressive we were going to get with the test conditions. For example, we could have quite legitimately chosen to plug the line cards in, and simply disable them in software. This would have avoided the delay caused by a “cold boot,” and would arguably have been closer to what might occur in certain types of real network maintenance operations. But, filled with confidence, we chose to make the test as difficult as possible, by starting the stopwatch when the first of four 500 Gb/s line cards was inserted by Craig. In this “cold boot” scenario, the line card must go through a full boot process, as well as use the DAMP protocol to bring up the super-channel power levels in a controlled way. As an aside, this tentative power increase, which is clearly visible on the OSA screen in the video, is designed to ensure that any existing wavelengths in the fiber are not disrupted with the addition of the new capacity. In the video you’ll see the following sequence:
- Craig inserts the first of the four line cards.
- Darren starts the stopwatch (an Android phone).
- The camera stays on the Optical Spectrum Analyzer (OSA) screen, but we fast-forward through the ten minute boot process.
- We see the four sets of waves of the super-channels using the DAMP protocol to ease the power levels up. This is complete by about 10 minutes 45 seconds on the stopwatch.
- To prove that this capacity really is available for service, we provision a 100GbE service over the link, which is why the camera (and stopwatch) moves to the test set screen.
- The entire process of lighting the super-channels and provisioning the 100GbE service is complete in about 11 minutes 45 seconds – more than eight minutes ahead of our target estimate!
Infinera Product Evangelist
Flexibility is a virtue in many areas, and transport networks are no exception. Last week, at the Light Reading Packet Optical Transport Evolution show in New York City, Shamim Akhtar, who is Senior Director, Network Architecture & Technology, at Comcast, called for a flexible Media Access Control (MAC) technology for the next generation of Ethernet beyond 100Gb/s.
In simple terms, Shamim’s idea is extremely pragmatic. The electronics industry is now facing serious challenges when it comes to increasing serial processing rates in silicon components. This is why, for example, your computer CPUs aren’t getting that much faster in terms of clock speeds, but you can have a whole bunch of cores running on your CPU chip. A flexible Ethernet MAC would allow the data rate of Ethernet to be increased in (for example) 100Gb/s chunks as advances in electronics allow, and without waiting for a drawn-out standards approval process (typically around five years for Ethernet standardization).
As I pointed out in a blog last year, it’s the same reason that the next “speed” of transport after 100Gb/s isn’t a fixed speed at all, but it’s a coherent super-channel that’s made up of multiple optical carriers. Each carrier can run as fast as the “state of the shelf” electronics will allow, while offering the same spectral efficiency and optical reach as an individual 100 Gb/s transponder. Technology such as Infinera’s FlexCoherent™ modulation allows the service provider to “dial in” the reach they need, and this will change the capacity of the super-channel.
I believe, and the market may agree, this solution is perfect for service providers who are facing the need to deploy more capacity without hiring an army of extra engineers to do it.
And flexibility has to go right up and down the stack. Since super-channels may operate at different data rates, and client services (like Shamim Akhtar’s flexible Ethernet MAC) will operate at various data rates, it’s essential to be able to “right-size” the OTN transport container in order to allow the service to be carried over a long haul carrier network efficiently. In ITU-T SG15, there is now an active proposal for a flexible transport container. The proposal – initially put forward by Infinera, Finisar and Verizon – is dubbed “OTUadapt” (the final version of the standard will be given an official ITU-T numerical designation as part of the OTN standards family). This will allow a standards-based approach for flexible OTN containers with dimensions of N x 100 Gb/s.
This development effectively completes the stack, as you can see here from the diagram. At the client level we have the proposed “flexible Ethernet MAC” (which could be a multiple of 100 Gb/s in capacity). At the OTN container level we find a flexible transport unit in the form of OTUadapt, and on the fiber we have flexible super-channels that can be sized according to the optical reach needed by the service provider. Note that the super-channel does not have to be sized to fit the OTUadapt container, because it would be quite normal to have multiple OTN transport containers used to fill the super-channel capacity. A flexible OTN container allows that to happen much more efficiently.
Sr. Director, Corporate Marketing
Last night we were delighted to see the good news out from the Dell’Oro Group as they published their Optical Transport Network Report with results for Q1’13. In the May 20, 2013 edition of the report the Dell’Oro Group ranked Infinera number one for the first quarter of 2013 in the global long-haul 100G wavelength division multiplexing (WDM) market as measured by the number of long-haul 100G ports sold. As a result, Infinera retains the number one market share position in long-haul 100G since the DTN-X entered the market in third quarter of 2012.
The report shows Infinera accounts for 34% of the long-haul 100G WDM ports sold in the first quarter of 2013, and also shows that Infinera accounts for 29% of all long-haul 100G ports sold since the long-haul 100G market emerged in 2010. This is remarkable considering the DTN-X just started showing up in market research results in Q3’12.
Also remarkable is the projected growth rate of this market segment. The Dell’Oro Group forecast projects 100G revenue to grow at a 47% CAGR between 2012 and 2017, reaching $5.6 B and contributing 60% of WDM capacity shipments by 2017.
Vice President, Corporate Marketing
“How do you deal with multiple simultaneous fiber cuts in what is now a mission critical system?” That was a question I posed to attendees of an Infinera Technology Briefing, “FastSMP as the Next-Generation Approach to Protection.” The occasion was the Optical Fiber Communication Conference and Exposition/National Fiber Optic Engineers Conference (OFC/NFOEC) held in Anaheim, California in March 2013.
Readers of this blog may recall a post around the same time from Infinera’s Renée Glowacki in which she highlighted the often bizarre reasons for network failures, including squirrels munching their way through cables. (See the post for the offender’s mugshot.)
Renée was referencing a Level 3 Communications blog post in 2011, “The 10 Most Bizarre and Annoying Causes of Fiber Cuts.” Having grown up in Pennsylvania, I was not surprised to learn that squirrels in my home state and in Florida were the cause of 17% of the fiber cuts in the US reported by Level 3 between January and August of 2013. Of course, it’s not just furry critters that cause problems. We humans routinely chop cables and wield backhoes, albeit accidentally in most cases. Add in hurricanes, earthquakes, ship anchors, tsunamis, and other natural disasters, and it’s easy to see how mitigating risks associated with multiple concurrent failures are major headaches for carriers and service providers.
Solutions do exist but at a price that’s becoming prohibitive to service providers struggling to meet customer demands for both 24×7 network availability and almost limitless bandwidth. Faced with a future where transmitting thousands of services over long haul fiber using coherent 100 Gb/s and 500 Gb/s super-channel technology will demand five or even six times current network reliability and availability, multiple fiber cuts will challenge MPLS Fast Re-Route and other approaches from both a performance and cost perspective. Just like Pieter Poll, SVP, Network Planning at CenturyLink said in the 2013 OSA Executive Forum, “55% of my costs go to protection.”
This is the backdrop behind the vision of Infinera’s Fast Shared Mesh Protection (FastSMP) technology, which leverages an increasingly intelligent transport network to enhance network reliability and delivers deterministic sub-50ms recovery for large-scale networks with thousands of services. In fact, Infinera, along with a consortium of industry vendors, are heavily involved in this standardization process (see ITU-Q9/SG15 and also IETF drafts on ‘Requirements for SMP’ and ‘Supporting SMP in MPLS-TP’).
The challenge, of course, is that in a world where we can carry 8 Tb/s on a single fiber, the ability to protect thousands of services with 50ms performance end-to-end requires more than just intelligent software. Thus, the DTN-X supports FastSMP using a dedicated hardware acceleration chip installed in every DTN-X board that has shipped and that will ship. I believe that Infinera is the only vendor in the industry that has hardware acceleration required to deliver FastSMP.
This post is not the optimal place to delve into the details of how FastSMP works. However, a paraphrased quote from Infinera’s FastSMP whitepaper is apt:
“FastSMP Processor operations are completely independent of software and the GMPLS control plane. This provides not only guaranteed sub-50ms protection at enormous scale, but also provides high availability by decoupling protection activation from the control plane software. For example, FastSMP protection is available even when software or GMPLS is not fully functioning – such as during a software update event. In this way, innovative protection SLAs can be designed using the DTN-X platform.” (Page 6, Document Number: WP-SD-FEC-1-2013).
What really matters is how customers make use of this technology. That’s why we were very pleased to announce recently that Pacnet, a global carrier headquartered in both Hong Kong and Singapore, has selected the DTN-X platform to support its long haul submarine mesh architecture. Enabling the hardware-accelerated FastSMP will give Pacnet a competitive advantage in offering fully protected services and restoration capability to their customers.
If you’d like to know more, then allow me to recommend Infinera’s FastSMP page. In addition to the whitepaper mentioned earlier, there’s also a very informative 10-page document on the business case for SMP instead of 1+1 protection. Written by ACG Research, one of the key findings from their modeling is that SMP’s TCO is 27% less than 1+1 protection over a five-year period. That’s significant because many customers are looking at the DTN-X as a five to 10-year infrastructure investment.
Finally, here’s a video of my presentation at the FastSMP technology briefing I mentioned at the start of this post. It covers many of the points made here and may be of interest to those who enjoy the audio-visual learning experience.
Also viewable on our YouTube page are presentations for the same event by our co-founder and Chief Strategy Officer, Dave Welch, and analyst firm Ovum’s VP & Practice Leader for Network Infrastructure, Dana Cooperson.
Sr. Director, Corporate Communications
This week the news keeps flowing out of Europe for Infinera as we announce new customers in Europe for the Infinera DTN-X platform, as well as the Infinera DTN and ATN platforms. Earlier this week, BICS and Infinera announced the selection of the DTN-X for BICS’ pan-European network. This news was followed by our announcement that the Nordic cable operator Agder Breiband (Agder) has selected the Infinera DTN and ATN platforms for the flexibility, scalability and simplicity of Infinera’s Digital Optical Network solutions.
Agder, a leading broadband service provider in the south of Norway, offers customers high speed Broadband Internet, IPTV and VoIP services. Agder offers services via their partner Altibox. Agder is deploying the Infinera DTN and ATN platforms in its regional backbone to provide high capacity transport for voice, video and data services.
“We selected Infinera for our network based on the flexibility of their solutions, ease of use and the ability to provide us with a simplified end-to-end solution for our network,” said Terje Abusland, Technical Manager at Agder. “At Agder, we have a small team and it’s critical for us to deploy with solutions that are cost effective for our business. Infinera’s solutions were easy to install and our team was able to deploy the network in seven days.”
Infinera’s new success in the European cable operator market builds off our previous successes in the North American cable market. For more information on Infinera’s solution for cable operators, check out our cable solutions page.
By Mark Showalter
Sr. Director, Corporate Communications
Today we are pleased to add BICS to the list of customers who have selected Infinera’s Digital Optical Network solution, as well as to the list of announced DTN-X customers. This morning, BICS, a global provider of international wholesale carrier services, announced the selection of the Infinera DTN-X platform to upgrade the BICS’ Pan-European network.
BICS offers international wholesale solutions to any communication service provider through a network of 100 points of presence in 55 cities and 33 countries across the globe. The BICS’ Pan-European network stretches across 9,000 kilometers of fiber in Europe and linking through a cable landing station in Marseille into the EIG & SEA-ME-WE 4 submarine cable systems.
“The Infinera DTN-X allows BICS to provide flexible solutions and ensures a faster service implementation, translating into a shorter time to market for our customers,” said Johan Wouters, SVP Capacity Business Management at BICS. “This new platform will enable the aggregation of multiple high speed services on a single OTN interface. The advanced control plane offers the possibility of self-provisioning for high capacity services making BICS the perfect network outsourcing partner.”
“We are delighted to be working with leading carriers like BICS to upgrade their networks with the DTN-X,” said Chris Champion, Infinera VP, EMEA Sales. “We’re seeing increasing traction for the DTN-X, evidenced by the purchase commitments we have received from 27 customers around the world for this new platform, out of our 115 total customers.”