Windstream to add ICE6 as it stirs its optical network
Windstream has sent an 800-gigabit optical signal between the US cities of Phoenix and San Diego. The operator used Infinera’s Groove modular chassis fitted with its latest ICE6 infinite capacity engine for the trial.
Infinera reported in March sending an 800-gigabit signal 950km with another operator but this is the first time a customer, Windstream, is openly discussing a trial and the technology.
The bulk of Windstream’s traffic is sent using 100-gigabit wavelengths. Moving to 800-gigabit will reduce its optical transport costs.
Windstream will also be able to cram more digital traffic down its fibre. It sends 12 terabits and that could grow to 40 terabits.
Motivation
Windstream provides residential broadband, business and wholesale services in the US.
“We operate a national footprint for wholesale and enterprise services,” says Art Nichols, vice president of architecture and technology at Windstream. “The optical focus is for wholesale and enterprise.”
Art Nichols
The communications service provider has 160,000 miles of fibre, 3,700 points-of-presence (PoPs) and operates in 840 cities. “We are continually looking to expand that,” says Nichols. “Picking up new PoPs, on-ramps and landing spots to jump onto the long-haul network.”
If Windstream’s traffic is predominantly at 100-gigabit, it also has 200-gigabit wavelengths and introduced recently 400-gigabit signals. In April Windstream and Infinera trialled Gigabit Ethernet (GbE) client-side services using LR8 modules.
Windstream is interested in adopting 800-gigabit wavelengths to reduce transport costs. “To try to draw as much efficiency as you can, using as few lasers as you can, to push down the cost-per-bit,” says Nichols.
The operator is experiencing traffic growth at a 20-30 per cent compound annual growth rate that is eroding its revenue-per-bit.
Weekly traffic has also jumped a further 20 per cent during the COVID-19 pandemic. Video traffic is the main driver, with peak traffic hours starting earlier in the day and continuing into the evenings.
Sending more data on a wavelength reduces cost-per-bit and improves revenue-per-bit figures.
In addition to sending a 800-gigabit signal over 730km, the operator sent a 700-gigabit signal 1,460km. The two spans are representative of Windstream’s network.
“Eight hundred gigabits is an easier multiple - better to fit two 400GbE clients - but 700 gigabits has tons of applications,” says Nichols. “We are predominantly filling 100-gigabit orders today so being able to multiplex them is advantageous.”
Another reason to embrace the new technology is to fulfill wholesale orders in days not months. “The ability to turn around multi-terabit orders from webscale customers,” says Nichols. “That is increasingly expected of us.”
One reason order fulfilment is faster is that the programming interfaces of the equipment are exposed, allowing Windstream to connect its management software. “We instantiate services in a short turnaround,” says Nichols.
ICE6 technology
Infinera’s ICE6 uses a 1.6-terabit photonics integrated circuit (PIC) and its 7nm CMOS FlexCoherent 6 digital signal processor (DSP). The 1.6 terabits is achieved using two wavelengths, each able to carry up to 800 gigabits of traffic.
The ICE6 uses several techniques to achieve its optical performance. One is Nyquist sub-carriers where data is encoded onto several sub-carriers rather than modulating all the data onto a single carrier.
The benefit of sub-carriers is that high data rates are achieved despite the symbol rate of each sub-carrier being much lower. The lower symbol rate means the optical transmission is more tolerant to non-linear channel impairments. Sub-carriers also have sharper edges so can be squeezed together enabling more data in a given slice of spectrum.
Infinera also applies probabilistic constellation shaping to each sub-carrier, enabling just the right amount of data to be placed on each one.
The FlexCoherent 6 DSP also uses soft-decision forward-error correction (SD-FEC) gain sharing. The chip can redistribute processing to the optical channel that needs it the most.
Some of the strength of the stronger signal can be cashed in to strengthen the weaker one, extending its reach or potentially allowing more bits to be sent by enabling a higher modulation scheme to be used.
Windstream cannot quantify the cost-per-bit advantage using the ICE6. “We don’t have finalised pricing,” says Nichols. But he says the latest coherent technology has significantly better spectral efficiency.
Spectral efficiency can be increased in two ways, says Rob Shore, Infinera’s senior vice president of marketing.
One is to increase the modulation scheme and the other is to close the link and maintain the high modulation over longer spans. If the link can’t be closed, lowering the modulation scheme is required which reduces the bits carried and the spectral efficiency.
Windstream will be able to increase capacity per fibre by as much as 70 per cent compared to the earlier generation 400-gigabit coherent technology and by as much as 35 per cent compared to 600-gigabit coherent.
A total of 42.4 terabits can be sent over a fibre using 800-gigabit wavelengths, says Shore, but the symbol rate needs to be reduced to 84 gigabaud shortening the overall reach.
Trial learnings
The rate-reach performance of the ICE6 was central to the trial but what Windstream sought to answer was how the ICE6 would perform across its network.
“We paid really close attention to margins and noise isolation as indicators as to how it would work across the network,” says Nichols. “The exciting thing is that it is extremely applicable.”
Windstream is also upbeat about the technology’s optical performance.
“We have a fair amount of information as to what the latest optical engines are capable of,” says Nichols. “This trial gave us a good view of how the ICE6 performs and it turns out it has advantages in terms of rate-reach performance.”
Ciena, Huawei and Infinera all have 800-gigabit coherent technology. Nokia recently unveiled its PSE-V family of coherent devices that does not implement 800-gigabit wavelengths.
Michael Genovese, a financial analyst at MKM Partners, puts the ICE6 on a par with Ciena’s WaveLogic 5 that is already shipping to over 12 customers.
“We expect 800 gigabit to be a large and long cycle," says Genovese in a recent research note. “We think most of the important internet content providers, telcos and subsea consortia will adopt a duel-vendor strategy, benefitting Ciena and Infinera over time.”
Windstream will adopt Infinera’s ICE6 technology in the first half of 2021. First customers to adopt the ICE6 will be the internet content providers later this year.
Infinera’s ICE6 sends 800 gigabits over a 950km link
Infinera has demonstrated the coherent transmission of an 800-gigabit signal across a 950km span of an operational network.
Infinera used its Infinite Capacity Engine 6 (ICE6), comprising an indium-phosphide photonic integrated circuit (PIC) and its FlexCoherent 6 coherent digital signal processor (DSP).
The ICE6 supports 1.6 terabits of traffic: two channels, each supporting up to 800-gigabit of data.
The trial, conducted over an unnamed operator’s network in North America, sent the 800-gigabit signal as an alien wavelength over a third-party line-system carrying live traffic.
“We have proved not only the state of our 800-gigabit with ICE6 but also the distances it can achieve,” says Robert Shore, senior vice president of marketing at Infinera.
800G trials
Several systems vendors have undertaken 800-gigabit optical trials.
Ciena detailed two demonstrations using its WaveLogic 5 Extreme (WL5e). One was an interoperability trial involving Verizon and Juniper Networks while the second connected two data centres belonging to the operator, Southern Cross Cable, to confirm the deployment of the WL5e cards in a live network environment.
Neither Ciena trial was designed to demonstrated WL5e’s limit of optical performance. Accordingly, no distances were quoted although both links were sub-100km, according to Ciena.
Meanwhile, Huawei has trialled its 800-gigabit technology in the networks of operators Turkcell and China Mobile.
The motivation for vendors to increase the speed of line-side optical transceivers is to reduce the cost of data transport. “One laser generating more data,” says Shore. “But it is not just high-speed transmissions, it is high-speed transmissions over distance.”
Infinera’s first 800-gigabit demonstration involved the ICE6 sending the signal over 800km of Corning’s TXF low-loss fibre.
“We did the demo on that fibre and we realised we had a ton of margin left over after completing the 800-gigabit circuit,” says Shore. The company then looked for a suitable network trial using standard optical fibre.
Infinera used a third-party’s optical line system to highlight that the 950km reach wasn’t due to a combination of the ICE6 module and the company’s own line system.
“What we have shown is that you can take any link anywhere, use anyone’s line system, carrying any kind of traffic, drop in the ICE6 and get 800-gigabit connections over 950km,” says Shore.
ICE 6
Infinera attributes the ICE6’s optical performance to its advanced coherent toolkit and the fact that the company has both photonics and coherent DSP technology, enabling their co-design to optimise the system’s performance.
One toolkit technique is Nyquist sub-carriers. Here, data is sent using several Nyquist sub-carriers across the channel instead of modulating the data onto a single carrier. The ICE6 is Infinera’s second-generation design to use sub-carriers, the first being ICE4, that doubles the number from four to eight.
The benefit of using sub-carriers is that high data rates can be achieved while the baud rate used for each one is much lower. And a lower baud rate is more tolerant to non-linear channel impairments during optical transmission.
Sub-carriers also improve spectral efficiency as the channels have sharper edges and can be packed tightly.
Infinera applies probabilistic constellation shaping to each sub-carrier, allowing fine-tuning of the data each carries. As a result, more data can be sent on the inner sub-carriers and less on the outer two outer sub-carrier where signal recovering is harder.
The sweet spot for sub-carriers is a symbol rate of 8-11 gigabaud (GBd). For the Infinera trial, eight sub-carriers were used, each at 12GBd, for an overall symbol rate of 96GBd.
“While it is best to stay as close to 8-11GBd, the coding gain you get as you go from 11GBd to 12GBd per sub-carrier is greater than the increased non-linear penalties,” says Shore.
Another feature of the coherent DSP is its use of soft-decision forward-error correction (SD-FEC) gain sharing. By sharing the FEC codes, processing resources can be shifted to one of the PIC’s two optical channels that needs it the most.
The result is that some of the strength of the stronger signal can be traded to bolster the weaker one, extending its reach or potentially allowing a higher modulation scheme to be used.
Applications
Linking data centres is one application where the ICE6 will be used. Another is sub-sea optical transmission involving spans that can be thousands of kilometres long, requiring lower modulation schemes and lower data rates.
“It’s not just cost-per-bit and power-per-bit, it is also spectral efficiency,” says Shore. “And a higher-performing optical signal can maintain a higher modulation rate over longer distances as well.”
Infinera says that at 600 gigabits-per-second (Gbps), link distances will be “significantly better” than 1,600km. The company is exploring suitable links to quantify ICE6’s reach at 600Gbps.
The ICE6 is packaged in a 5×7-inch optical module. Infinera’s Groove series will first adopt the ICE6 followed by the XTC platforms, part of the DTN-X series. First network deployments will occur in the second half of this year.
Infinera is also selling the ICE6 5×7-inch module to interested parties.
XR Optics
Infinera is not addressing the 400ZR coherent pluggable module market. The 400ZR is the OIF-defined 400-gigabit coherent standard developed to connect equipment in data centres up to 120km apart.
Infinera is, however, eyeing the emerging ZR+ opportunity using XR Optics. ZR+ is not a standard but it extends the features of 400ZR.
XR Optics is the brainchild of Infinera that is based on coherent sub-carriers. All the sub-carriers can be sent to the same destination for point-to-point links, but they can also be sent to different locations to allow for point-to-multipoint communications. Such an arrangement allows for traffic aggregation.
“You can steer all the sub-carriers coming out of an XR transceiver to the same destination to get a 400-gigabit point-to-point link to compete with ZR+,” says Shore. “And because we are using sub-carriers instead of a single carrier, we expect to get significantly better performance.”
Infinera is developing the coherent DSPs for XR Optics and has teamed up with optical module makers, Lumentum and II-VI.
Other unnamed partners have joined Infinera to bring the technology to market. Shore says that the partners include network operators that have contributed to the technology’s development.
Infinera planned to showcase XR Optics at the OFC conference and exhibition held recently in San Diego.
Shore says to expect XR Optics announcements in late summer, from Infinera and perhaps others. These will detail the XR Optics form factors and how they function as well as the products’ schedules.
Infinera rethinks aggregation with slices of light
An optical architecture for traffic aggregation that promises to deliver networking benefits and cost savings was unveiled by Infinera at this week’s ECOC show, held in Dublin.
Traffic aggregation is used widely in the network for applications such as fixed broadband, cellular networks, fibre-deep cable networks and business services.
Infinera has developed a class of optics, dubbed XR optics, that fits into pluggable modules for traffic aggregation. And while the company is focussing on the network edge for applications such as 5G, the technology could also be used in the data centre.
Optics is inherently a point-to-point communications technology, says Infinera. Yet optics is applied to traffic aggregation, a point-to-multipoint architecture, and that results in inefficiencies.
“The breakthrough here is that, for the first time in optics’ history, we have been able to make optics work to match the needs of an aggregation network,” says Dave Welch, founder and chief innovation officer at Infinera.
Infinera proposes coherent sub-carriers for a new class of problem
XR Optics
Infinera came up with the ‘XR’ label after borrowing from the naming scheme used for 400ZR, the 400-gigabit pluggable optics coherent standard.
“XR can do point-to-point like ZR optics,” says Welch. “But XR allows you to go beyond, to point-to-multipoint; ‘X’ being an ill-defined variable as to exactly how you want to set up your network.”
XR optics uses coherent technology and Nyquist sub-carriers. Instead of using a laser to generate a single carrier, pulse-shaping is used at the transmitter to generate multiple carriers, referred to as Nyquist sub-carriers.
The sub-carriers convey the same information as a single carrier but by using several sub-carriers, a lower symbol rate can be used for each. The lower symbol rate improves the tolerance to non-linear effects in a fibre and enables the use of lower-speed electronics.
Infinera first detailed Nyquist sub-carriers as part of its advanced coherent toolkit, and implemented the technology with its Infinite Capacity Engine 4 (ICE4) used for optical transport.
The company is bringing to market its second-generation Nyquist sub-carrier design with its ICE6 technology that supports 800-gigabit wavelengths.
Now Infinera is proposing coherent sub-carriers for a new class of problem: traffic aggregation. But XR optics will need backing and be multi-sourced if it is to be adopted widely.
Network operators will also need to be convinced of the technology’s merits. Infinera claims XR optics will halve the pluggable modules needed for aggregation and remove the need for intermediate digital aggregation platforms, reducing networking costs by 70 percent.
Aggregation optics
XR optics will be required at both ends of a link. The modules will need to understand a protocol that tells them the nature of the sub-carriers to use: their baud rate (and resulting spectral width) and modulation scheme.
Infinera cites as the example a 4GHz-wide sub-carrier modulated using 16-ary quadrature amplitude modulation (16-QAM) that can transmit 25-gigabit of data.
A larger capacity XR coherent module will be used at the aggregation hub and will talk directly with XR modules at the network edge, “casting out” its sub-carriers to the various pluggable modules at the network edge.
For example, the module at the hub may be a 400-gigabit QSFP-DD supporting 16, 25-gigabit sub-carriers, or an 800-gigabit QSFP-DD or OSFP module delivering 32 sub-carriers. A mix of lower-speed XR modules will be used at the edge: 100-gigabit QSFP28 XR modules based on four sub-carriers and single sub-carrier 25-gigabit SFP28s.
“As soon as you have defined that each one of these transceivers is some multiple of that 25-gigabit sub-carrier, they can all talk to each other,” says Welch.
The hub XR module and network-edge modules are linked using optical splitters such that all the sub-channels sent by the hub XR module are seen by each of the edge modules. The hub in effect broadcasts its sub-carriers to all the edge devices, says Welch.
A coding scheme is used such that each edge module’s coherent receiver can pick off its assigned sub-channel(s). In turn, an edge module will send its data using the same frequencies on a separate fibre.
Basing the communications on multiples of sub-carriers means any XR module can talk to any other, irrespective of their overall speeds.
Sub-carriers can also be reassigned.
“In that fashion, today you are a 25-gigabit client module and tomorrow you are 100-gigabit,” says Welch. Reassigning edge-module capacities will not happen often but when undertaken, no truck roll will be needed.
System benefits
In a conventional aggregation network, the edge transceivers send traffic to an intermediate electrical aggregation switch. The switch’s line-side-facing transceivers then send on the aggregated traffic to the hub.
Using XR optics, the intermediate aggregation switch becomes redundant since the higher-capacity XR coherent module aggregates the traffic from the edge. Removing the switch and its one-to-one edge-facing transceivers account for the halving of the overall transceiver count and the overall 70 percent network cost saving (see diagram below).
The disadvantage of getting rid of the intermediate aggregation switch is minor in comparison to the plusses, says Infinera.
“In a network where all the traffic is going left to right, there is always an economic gain,” says Welch. And while a layer-2 aggregation switch enables statistical multiplexing to be applied to the traffic, it is insignificant when compared to the cost-savings XR optics brings, he says.
Challenges
XR transceivers will need to support sub-carriers and coherent signal processing as well as the language that defines the sub-carriers and their assignment codes. Accordingly, module makers will need to make a new class of XR pluggable modules.
“We are working with others,” says Welch. “The object is to bring the technology and a broad-base supply chain to the market.” The fastest way to achieve this, says Welch, is through a series of multi-source agreements (MSAs). Arista Networks and Lumentum were both quoted as part of Infinera’s XR Optics press release.
Another challenge is that a family of coherent digital signal processors (DSPs) will need to be designed that fit within the power constraints of the various slim client-side pluggable form factors.
Infinera stresses it is unveiling a technological development and not a product announcement. That will come later.
However, Welch says that XR optics will support a reach of hundreds of kilometres and even metro-regional distances of over 1,000km.
“We are comfortable we are working with partners to get this out,” says Welch. “We are comfortable we have some key technologies that will enhance these capabilities as well.”
Other applications
Infinera’s is focussing its XR optics on applications such as 5G. But it says the technology will benefit many network applications.
“If you look at the architecture in the data centre or look are core networks, they are all aggregation networks of one flavour or another,” says Welch. “Any type of power, cost, and operational savings of this magnitude should be evaluated across the board on all networks.”
Access drives a need for 10G compact aggregation boxes
Infinera has unveiled a platform to aggregate multiple 10-gigabit traffic streams originating in the access network.
The 1.6-terabit HDEA 1600G platform is designed to aggregate 80, 10-gigabit wavelengths. The use of ten-gigabit wavelengths in access continues to grow with the advent of 5G mobile backhaul and developments in cable and passive optical networking (PON).
A distributed access architecture being embraced by cable operators. Shown are the remote PHY devices (RPD) or remote MAC-PHY devices (RMD), functionality moved out of the secondary hub and closer to the end user. Also shown is how DWDM technology is moved closer to the edge of the network. Source: Infinera.
A distributed access architecture being embraced by cable operators. Shown are the remote PHY devices (RPD) or remote MAC-PHY devices (RMD), functionality moved out of the secondary hub and closer to the end user. Also shown is how DWDM technology is moved closer to the edge of the network. Source: Infinera.
Infinera has adopted a novel mechanical design for its 1 rack unit (1RU) HDEA 1600G that uses the sides of the platform to fit 80 SFP+ optical modules.
The platform also features a 1.6-terabit Ethernet switch chip that aggregates the traffic from the 10-gigabit streams to fill 100-gigabit wavelengths that are passed to other switching or transport platforms for transmission into the network.
Distributed access architecture
Jon Baldry, metro marketing director at Infinera, cites the adoption of a distributed access architecture (DAA) by cable operators as an example of 10-gigabit links that are set to proliferate in the access network.
DAA is being adopted by cable operators to compete with the telecom operators’ rollout of fibre-to-the-home (FTTH) broadband access technology.
A recent report by market research firm, Ovum, addressing DAA in the North American market, discusses how the architectural approach will free up space in cable headends, reduce the operators’ operational costs, and allow the delivery of greater bandwidth to subscribers.
Implementing DAA involves bringing fibre as well as cable network functionality closer to the user. Such functionality includes remote PHY devices and remote MAC-PHY devices. It is these devices that will use a 10-gigabit interface, says Baldry: “The traffic they will be running at first will be two or three gigabits over that 10-gigabit link.”
Julie Kunstler, principal analyst at Ovum’s Network Infrastructure and Software group, says the choice whether to deploy a remote PHY or a remote MAC-PHY architecture is a issue of an operator's ‘religion’. What is important, she says, is that both options exploit the existing hybrid fibre coax (HFC) architecture to boost the speed tiers delivered to users.
The current, pre-DAA, cable network architecture. Source: Infinera.
In the current pre-DAA architecture, the cable network comprises cable headends and secondary distribution hubs (see diagram above). It is at the secondary hub that the dense wavelength-division multiplexing (DWDM) network terminates. From there, RF over fibre is carried over the hybrid fibre-coax (HFC) plant. The HFC plant also requires amplifier chains to overcome cable attenuation and the losses resulting from the cable splits that deliver the RF signals to the homes.
Typically, an HFC node in the cable network serves up to 500 homes. With the adoption of DAA and the use of remote PHYs, the amplifier chains can be removed with each PHY serving 50 homes (see diagram top).
“Basically DWDM is being pushed out to the remote PHY devices,” says Baldry. The remote PHYs can be as far as 60km from the secondary hub.
“DAA is a classic example where you will have dense 10-gigabit links all coming together at one location,” says Baldry. “Worst case, you can have 600-700 remote PHY devices terminating at a secondary hub.”
The same applies to cellular.
At present 4G networks use 1-gigabit links for mobile backhaul but 5G will use 10-gigabit and 25-gigabit links in a year or two. “So the edge of the WDM network has really jumped from 1 gigabit to 10 gigabit,” says Baldry.
It is the aggregation of large numbers of 10-gigabit links that the HDEA 1600G platform is designed to address.
HDEA 1600G
Only a certain number of pluggable interfaces can fit on the front panel of a 1RH box. To accommodate 80, 10-gigabit streams, the two sides of the platform are used for the interfaces. Using the HDEA’s sides creates much more space for the 1RU’s input-output (I/O) compared to traditional transport kit, says Baldry.
The 40 SFP+ modules on each side of the platform are accessed by pulling the shelf out and this can be done while it is operational (see photo below). Such an approach is used for supercomputing but Baldry believes Infinera is the first to adopt it for a transport product.
Infinera has also adopted MPO connectors to simplify the fibre management involved in connected 80 SFP+, each module requiring a fibre pair.
The HDEA 1600 has two groups of four MPO connectors on the front panel. Each MPO cluster connects 40 modules on each side, with each MPO cable having 20 fibres to connect 10 SFP+ modules.
A site terminating 400 remote PHYs, for example, requires the connection of 40 MPO cables instead of 800 individual fibres, says Baldry, simplifying installation greatly.
>
“DAA is a classic example where you will have dense 10-gigabit links all coming together at one location. Worst case, you can have 600-700 remote PHY devices terminating at a secondary hub.”
The other end of the MPO cable connects to a dense multiplexer-demultiplexer (mux-demux) unit that separates the individual 10-gigabit access wavelengths received over the DWDM link.
Each mux-demux unit uses an arrayed waveguide grating (AWG) that is tailored to the cable operators’ wavelengths needs. The 24-channel mux-demux design supports 20, 100GHz-wide channels for the 10-gigabit wavelengths and four wavelengths reserved for business services. Business services have become an important part of the cable operators’ revenues.
Infinera says the HDEA platform supports the extended C-band for a total of 96 wavelengths.
The company says it will develop different AWG configurations tailored for the wavelengths and channel count required for the different access applications.
In the rack, the HDEA aggregation platform takes up one shelf, while eight mux-demux units take up another 1RU. Space is left in between to house the cabling between the two.
The HDEA 1600G pulled out of the rack, showing the MPO connectors and the space to house the cabling between the HDEA and the rack of compact AWGs. Source: Infinera.
Baldry points out that the four business service wavelengths are not touched by the HDEA platform, Rather, these are routed to separate Ethernet switches dedicated to business customers. "We break those wavelengths out and hand them over to whatever system the operator is using," he says.
The HDEA 1600G also features eight 100-gigabit line-side interfaces that carry the aggregated cable access streams. Infinera is not revealing the supplier of the 1.6 terabit switch silicon - 800-gigabit for client-side capacity and 800-gigabit for line-side capacity - it is using for the HDEA platform.
The platform supports all the software Infinera uses for its EMXP, a packet-optical switch tailored for access and aggregation that is part of Infinera’s XTM family of products. Features include multi-chassis link aggregation group (MC-LAG), ring protection, all the Metro Ethernet Forum services, and synchronisation for mobile networks, says Baldry
Auto-Lambda
Infinera has developed what it calls its Auto-Lambda technology to simplify the wavelength management of the remote PHY devices.
Here, the optics set up the connection instead of a field engineer using a spreadsheet to determine which wavelength to use for a particular remote PHY. Tunable SFP+ modules can be used at the remote PHY devices only with fixed-wavelength (grey) SFP+ modules used by the HDEA platform to save on costs, or both ends can use tunable optics. Using tunable SFP+ modules at each end may be more expensive but the operator gains flexibility and sparing benefits.
Jon Baldry
Establishing a link when using fixed optics within the HDEA platform, the SFP+ is operated in a listening mode only. When a tunable SFP+ transceiver is plugged in at a remote PHY, which could be days later, it cycles through each wavelength. The blocking nature of the AWG means that such cycling does not disturb other wavelengths already in use.
Once the tunable SFP+ reaches the required wavelength, the transmitted signal is passed through the AWG to reach the listening transceiver at the switch. On receipt of the signal, the switch SFP+ turns on its transmitter and talks to the remote transceiver to establish the link.
For the four business wavelengths, both ends of the link use auto-tunable SFP+ modules, what is referred to a duel-ended solution. That is because both end-point systems may not be Infinera platforms and may have no knowledge as to how to manage WDM wavelengths, says Baldry.
In this more complex scenario, the time taken to establish a link is theoretically much longer. The remote end module has to cycle through all the wavelengths and if no connection is made, the near end transceiver changes its transmit wavelength and the remote end’s wavelength cycling is repeated.
Given that a sweep can take two minutes or more, an 80-wavelength system could take close to three hours in the worst case to establish the link; an unacceptable delay.
Infinera is not detailing how its duel-ended scheme works but a combination of scanning and communications is used between the two ends. Infinera had shown such a duel-ended scheme set up a link in 4 minutes and believes it can halve that time.
Finisar detailed its own Flextune fast-tuning technology at ECOC 2018. However, Infinera stresses its technology is different.
Infinera says it is talking to several pluggable optical module makers. “They are working on 25-gigabit optics which we are going to need for 5G,” says Baldry. “As soon as they come along, with the same firmware, we then have auto-tunable for 5G.”
System benefits
Infinera says its HDEA design delivers several benefits. Using the sides of the box means that the platform supports 80 SFP+ interfaces, twice the capacity of competing designs. In turn, using MPO connectors simplifies the fibre management, benefiting operational costs.
Infinera also believes that the platform’s overall power consumption has a competitive edge. Baldry says Infinera incorporates only the features and hardware needed. “We have deliberately not done a lot of stuff in Layer 2 to get better transport performance,” he says. The result is a more power-efficient and lower latency design. The lower latency is achieved using ‘thin buffers’ as part of the switch’s output-buffered queueing architecture, he says.
The platform supports open application programming interfaces (APIs) such that cable operators can make use of such open framework developments as the Cloud-Optimised Remote Datacentre (CORD) initiative being developed by the Open Networking Foundation. CORD uses open-source software-defined networking (SDN) technology such as ONOS and the OpenFlow protocol to control the box.
An operator can also choose to use Infinera’s Digital Network Administrator (DNA) management software, SDN controller, and orchestration software that it has gained following the Coriant acquisition.
The HDEA 1600G is generally available and in the hands of several customers.
Infinera buying Coriant will bring welcome consolidation
Infinera is to purchase privately-held Coriant for $430 million. The deal will effectively double Infinera’s revenues, add 100 new customers and expand the systems vendor’s product portfolio.
Infinera's CEO, Tom FallonBut industry analysts, while welcoming the consolidation among optical systems suppliers, highlight the challenges Infinera faces making the Coriant acquisition a success.
“The low price reflects that this isn't the best asset on the market,” says Sterling Perrin, principal analyst, optical networking and transport at Heavy Reading. “They are buying $1 of revenue for 50 cents; the price reflects the challenges.”
Benefits
According to Perrin, there are still too many vendors facing "brutal price pressures" despite the optical industry being mature. Removing one vendor that has been cutting prices to win business is good news for the rest.
For Infinera, the acquisition of Coriant promises three main benefits, as outlined by its CEO, Tom Fallon, during a briefing addressing the acquisition.
The first is expanding its vertically-integrated business model across a wider portfolio of products. Infinera develops its own optical technology: its indium-phosphide photonic integrated circuits (PICs) and accompanying coherent DSPs that power its platforms. Having its own technology differentiates the optical performance of its platforms and helps it achieve leading gross margins of over 40 percent, said Fallon.
Exploiting the vertical integration model will be a central part of the Coriant acquisition. Indeed, the company mentioned vertical integration 21 times in as many minutes during its briefing outlining the deal. Infinera expects to deliver industry-leading growth and operating margins once it exploits the benefits of vertical integration across an expanded portfolio of platforms, said Fallon.
Having a seat at the table with the largest global service providers to strategise about where their business is going will be invaluable
Buying Coriant also gives Infinera much-needed scale. Not only will Infinera double its revenues - Coriant’s revenues were about $750 million in 2017 while Infinera’s were $741 million for the same period - but it will expand its customer base including key tier-one service providers and webscale players. According to Fallon, the newly combined company will include nine of the top 10 global tier-one service providers and the six leading global internet content providers.
Infinera admits it has struggled to break into the tier-one operators and points out that trying to enter is an expensive and time-consuming process, estimated at between $10 million to $20 million each time. “[Now, with Coriant,] having a seat at the table with the largest global service providers to strategise about where their business is going will be invaluable,” said Fallon.
Sterling Perrin of Heavy Reading The third benefit Infinera gains is an expanded product portfolio. Coriant has expertise in layer 3 networking, in the metro core with its mTera universal transport platform as well as SDN orchestration and white box technologies. Heavy Reading’s Perrin says Coriant has started development of a layer-3 router white box for edge applications.
Combining the two companies also results in a leading player in data centre interconnect.
“Coriant expands our portfolio, particularly in packet and automation where significant network investment is expected over the next decade,” said Fallon. The deal is happening at the right time, he said, as operators ramp spending as they undertake network transformation.
Infinera will pay $230 million in cash - $150 million up front and the rest in increments - and a further $200 million in shares for Coriant. The company expects to achieve cost savings of $250 million between 2019 and 2021 by combining the two firms, $100 million in 2019 alone. The deal is expected to close in the third quarter of 2018.
If a company is going to put that integrated product into their network, it’s a full-blown RFP process which Infinera may or may not win
Challenges
Industry analysts, while seeing positives for Infinera, have concerns regarding the deal.
A much-needed consolidation of weaker vendors is how George Notter, an analyst at the investment bank, Jefferies, describes the deal. For Infinera, however, continuing as before was not an option. Heavy Reading’s Perrin agrees: ”Infinera has been under a lot of pressure; their core business of long-haul has slowed.”
The deal brings benefits to Infinera: scale, complementary product sets, and the promise of being able to invest more in R&D to benefit its PIC technology, says Notter in a research note.
Gaining customers is also a key positive. “Infinera is really excited about getting the new set of customers and that is what they are paying for,” says Vladimir Kozlov, CEO of LightCounting Market Research. “However, these customers were gained by pricing products at steep discounts.”
What is vital for Infinera is that it delivers its upcoming 2.4-terabit Infinite Capacity Engine 5 (ICE5) optical engine on time. The ICE5 is expected to ship in early 2019. In parallel, Infinera is developing its ICE6 due two years later. Infinera is developing two generations of ICE designs in parallel after being late to market with its current 1.2-terabit optical engine.
Infinera is really excited about getting the new set of customers and that is what they are paying for
But even if the ICE5 is delivered on time, upgrading Coriant's platforms will be a major undertaking. “It sounds like they are going to fit their optical engines in all of Coriant’s gear; I don’t see how that is going to happen anytime quickly,” says Perrin.
Customers bought Coriant's equipment for a reason. Once upgraded with Infinera’s PICs, these will be new products that have to undergo extensive lab testing and full evaluations.
Perrin questions how moving customers off legacy platforms to the new will not result in the service providers triggering a new request-for-proposal (RFP). “If a company is going to put that integrated product into their network, it’s a full-blown RFP process which Infinera may or may not win,” says Perrin. “Infinera talked a lot about the benefits of vertical integration but they didn’t really address the challenges and the specific steps they would take to make that work.”
LightCounting's Vladimir KozlovLightCounting’s Kozlov also questions how this will work.
“The story about vertical integration and scaling up PIC production is compelling, but how will they support Coriant products with the PIC?” he says. “Will they start making pluggable modules internally? Will Coriant’s customers be willing to move away from the pluggables and get locked into Infinera’s PICs? Do they know something that we don’t?”
While Infinera is a top five optical platform supplier globally it hasn’t dominated the market with its PIC technologies, says Perrin. “Even if they technically pull off the vertical integration with the Coriant products, how much is that going to win business for them?” he says. “It is one architecture in a mix that has largely gone to pluggables.”
Transmode
Infinera already has experience acquiring a systems vendor when it bought in 2015 metro-access player, Transmode. Strategically, this was a very solid acquisition, says Perrin, but the jury is still out as to its success.
“The integration, making it work, how Transmode has performed within Infinera hasn’t gone as well as they wanted,” says Perrin. “That said, there are some good opportunities going forward for the Transmode group.”
Infinera also had planned to integrate its PIC technology within Transmode’s products but it didn't make economic sense for the metro market. There may also have been pushback from customers that liked the Transmode products, says Perrin: “With Coriant it looks like they really are going to force the vertical integration.”
Infinera acknowledges the challenges ahead and the importance of overcoming them if it is to secure its future.
“Given the comparable sizes of each company’s revenues and workforce, we recognise that integration will be challenging and is vital for our ultimate success,” said Fallon.
Infinera’s ICE flow
Infinera’s newest Infinite Capacity Engine 5 (ICE5) doubles capacity to 2.4 terabits. The ICE, which comprises a coherent DSP and a photonic integrated circuit (PIC), is being demonstrated this week at the OFC show being held in San Diego.
Infinera has also detailed its ICE6, being developed in tandem with the ICE5. The two designs represent a fork in Infinera’s coherent engine roadmap in terms of the end markets they will address.
Geoff BennettThe ICE5 is targeted at data centre interconnect and applications where fibre in being added towards the network edge. The next-generation access network of cable operators is one such example. Another is mobile operators deploying fibre in preparation for 5G.
First platforms using the ICE5 will be unveiled later this year and will ship early next year.
Infinera’s ICE6 is set to appear two years after the ICE5. Like the ICE4, Infinera’s current Infinite Capacity Engine, the ICE6 will be used across all of Infinera’s product portfolio.
Meanwhile, the 1.2 terabit ICE4 will now be extended to work in the L-band of optical wavelengths alongside the existing C-band, effectively doubling a fibre’s capacity available for service providers.
Infinera’s decision to develop two generations of coherent designs follows the delay in bringing the ICE4 to market.
“The fundamental truth about the industry today is that coherent algorithms are really hard,” says Geoff Bennett, director, solutions and technology at Infinera.
By designing two generations in parallel, Infinera seeks to speed up the introduction of its coherent engines. “With ICE5 and ICE6, we have learnt our lesson,” says Bennett. “We recognise that there is an increased cadence demanded by certain parts of the industry, predominately the internet content providers.”
ICE5
The ICE5 uses a four-wavelength indium-phosphide PIC that, combined with the FlexCoherent DSP, supports a maximum symbol rate of 66Gbaud and a modulation rate of up to 64-ary quadrature amplitude modulation (64-QAM).
Infinera says that the FlexCoherent DSP used for ICE5 is a co-development but is not naming its partners.
Using 64-QAM and 66Gbaud enables 600-gigabit wavelengths for a total PIC capacity of 2.4 terabits. Each PIC is also ‘sliceable’, allowing each of the four wavelengths to be sent to a different location.
Infinera is not detailing the ICE5’s rates but says the design will support lower rates, as low as 200 gigabit-per-second (Gbps) or possibly 100Gbps per wavelength.
Bennett highlights 400Gbps as one speed of market interest. Infinera believes its ICE5 design will deliver 400 gigabits over 1,300km. The 600Gbps wavelength implemented using 64-QAM and 66Gbaud will have a relatively short reach of 200-250km.
“A six hundred gigabit wavelength is going to be very short haul but is ideal for data centre interconnect,” says Bennett, who points out that the extended reach of 400-gigabit wavelengths is attractive and will align with the market emergence of 400 Gigabit Ethernet client signals.
Probabilistic shaping squeezes the last bits of capacity-reach out of the spectrum
Hybrid Modulation
The 400-gigabit will be implemented using a hybrid modulation scheme. While Infinera is not detailing the particular scheme, Bennett cites several ways hybrid modulation can be implemented.
One hybrid modulation technique is to use a different modulation scheme on each of the two light polarisations as a way of offsetting non-linearities. The two modulation schemes can be repeatedly switched between the two polarisation arms. “It turns out that the non-linear penalty takes time to build up,” says Bennett.
Another approach is using blocks of symbols, varying the modulation used for each block. “The coherent receiver has to know how many symbols you are going to send with 64-QAM and how many with 32-QAM, for example,” he says
A third hybrid modulation approach is to use sub-carriers. In a traditional coherent system, a carrier is the output of the transmit laser. To generate sub-carriers, the coherent DSP’s digital-to-analogue converter (DAC) applies a signal to the modulator which causes the carrier to split into multiple sub-carriers.
To transmit at 32Gbaud, four sub-carriers can be used, each modulated at 8Gbaud, says Bennett. Nyquist shaping is used to pack the sub-carriers to ensure there is no spectral efficiency penalty.
“You now have four parallel streams and you can deal with them independently,” says Bennett, who points out that 8Gbaud turns out to be an optimal rate in terms of minimising non-linearities made up of cross-phase and self-phase modulation components.
Sub-carriers can be described as a hybrid modulation approach in that each sub-carrier can be operated at a different baud rate and use a different modulation scheme. This is how probabilistic constellation shaping - a technique that improves spectral efficiency and which allows the data rate used on a carrier to be fine-tuned - will be used with the ICE6, says Infinera.
For the ICE5, sub-carriers are not included. “For the applications we will be using ICE5 for, the sub-carrier technology is not as important,” says Bennett. “Where it is really important is in areas such as sub-sea.”
Silicon photonics has a lower carrier mobility. It is going to be harder and harder to build such parts of the optics in silicon.
Probabilistic constellation shaping
Infinera is not detailing the longer-term ICE6 beyond highlights two papers that were presented at the ECOC show last September that involved a working 100Gbaud sub-carrier-driven wavelength and probabilistic shaping applied to a 1024-QAM signal.
The 100Gbaud rate will enable higher capacity transponders while the use of probabilistic shaping will enable greater spectral efficiency. “Probabilistic shaping squeezes the last bits of capacity-reach out of the spectrum,” says Bennett.
“In ICE6 we will be doing different modulation on each sub-carrier,” says Bennett. “That will be part of probabilistic constellation shaping.” And assuming Infinera adheres to 8Gbaud sub-carriers, 16 will be used for a 100Gbaud symbol rate.
Infinera argues that the interface between the optics and the DSP becomes key at such high baud rates and it argues that its ability to develop both components will give it a system design advantage.
The company also argues that its use of indium phosphide for its PICs will be a crucial advantage at such high baud rates when compared to silicon photonics-based solutions. “Silicon photonics has a lower carrier mobility,” says Bennett. “It is going to be harder and harder to build such parts of the optics in silicon.”
ICE4 embraces the L-band
Infinera’s 1.2 terabit six-wavelength ICE4 was the first design to use Nyquist sub-carriers and SD-FEC gain sharing, part of what Infinera calls its advanced coherent toolkit.
At OFC, Infinera announced that the ICE4 will add the L-band in addition to the C-band. It also announced that the ICE4 has now been adopted across Infinera’s platform portfolio.
The first platforms to use the ICE4 were the Cloud Xpress 2, the compact modular platform used for data centre interconnect, and the XT-3300, a 1 rack-unit (1RU) modular platform targeted at long-haul applications.
A variant of the platform tailored for submarine applications, the XTS-3300, achieved a submarine reach of 10,500km in a trial last year. The modulation format used was 8-QAM coupled with SD-FEC gain-sharing and Nyquist sub-carriers. The resulting spectral efficiency achieved was 4.5bits/s/Hz. In comparison, standard 100-gigabit coherent transmission has a spectral efficiency of 2bits/s/Hz. The total capacity supported in the trial was 18.2 terabits.
Since then, the ICE4 has been added the various DTN-X chassis including the XT-3600 2.4 terabit 4RU platform.
Books in 2017: Part 2
Dave Welch, founder and chief strategy and technology officer at Infinera
One favourite book I read this year was Alexander Hamilton by Ron Chernow. Great history about the makings of the US government and financial systems as well as a great biography. Another is The Gene: An Intimate History by Siddhartha Mukherjee, a wonderful discussion about the science and history of genetics.
Yuriy Babenko, senior expert NGN, Deutsche Telekom
As part of my reading in 2017 I selected two technical books, one general life-philosophy title and one strategy book.
Today’s internet infrastructure design is hardly possible without what we refer to as the cloud. Cloud is a very general term but I really like the definition of NIST: Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources that can be rapidly provisioned and released with minimal management effort or service provider interaction.
Cloud Native Infrastructure: Patterns for scalable infrastructure and applications in a dynamic environment by Kris Nova and Justin Garrison helps you understand the necessary characteristics of such cloud infrastructure and defines the capabilities of the service architecture that fits this model. The Cloud Native architecture is not just about ‘lift and shift’ into the cloud, it is the redesign of your services focusing on cloud elasticity, scalability, and security as well as operational models including but not limited to infrastructure as code. If you already heard about Kubernetes, Terraform and Cloud Native Foundation but want to understand how various technologies and frameworks fit together, this is a great and easy read.
High Performnce Browser Networking: What Every Web Developer Should Know About Networking and Web Performance by Ilya Grigorik provides a thorough look into the peculiarities of modern browser networking protocols, their foundation, methods and tools that help to optimise and increase the performance of internet sites.
Every serious business today has a web presence. Many services and processes are consumed through the browser, so a look behind the curtains of these infrastructure is informative and useful.
Probably one of the more interesting conclusions is that is not always the bandwidth which is necessary for a site’s successful operation but rather the end-to-end latency. The book discusses HTTP, HTTP2 and SPDY and will be of great interest to anyone who wants to refresh their knowledge of the history of the internet as well as to understand the peculiarities of performance optimisation of (big) internet sites.
Principles: Life and Work by Ray Dalio is probably one of the most discussed books of 2017. Mr Dalio is one of the most successful hedge-fund investors of our generation. In this book, he decided to share the main life and business principles that have guided his decisions during the course of his life. The main message which Dalio shares is not to copy or use his particular principles, although you are likely to adopt several of them, but to have your own.
One of Dalio’s key ideas is that everything works as a machine so if you define the general rules of how the machine (i.e. life in general) works, it will be significantly easier to follow the ups and downs and apply clear thinking in case of difficulties and challenges. He sums it up in an easy-to-comprehend approach which goes like the following: you try things out, reflect on them if something goes well or wrong, log all problems you face along the way, reflect on them and formulate and write down the principles. In due course, you will end up with your own version of Principles. Sounds easy but doing it is the key.
Edge Strategy: A New Mindset for Profitable Growth by Dan McKone and Alan Lewis is about the edges of a business, opportunities sitting comfortably in front of you and your business and which can be 'easily’ tackled and addressed.
Why would you go for a crazy new and risky business idea when there is a bunch of market opportunities just outside of the main door of your core business?
This sounds like “focus and expand” to me and makes a lot of sense. The authors identify three main “edges” which a business can address: product edge, journey edge and enterprise edge.
The book goes into detail about how product edge can be expanded (remember your shiny new iPhone leather case?); a firm can focus more on the complete customer journey (What are the jobs to be done? What problem is the customer really trying to solve? Airbnb service can be a great example); and finally leveraging the enterprise edge (like Amazon renting and selling unused server capacity via its AWS services).
Edge strategies are not new per se, but this book helps to formulate and structure the discussion in an understandable and comprehensive framework.
Has coherent optical transmission run its course?
Feature: Coherent's future
Three optical systems vendors share their thoughts about coherent technology and the scope for further improvement as they look two generations ahead to symbol rates approaching 100 gigabaud
Optical transmission using coherent detection has made huge strides in the last decade. The latest coherent technology with transmitter-based digital signal processing delivers 25x the capacity-reach of 10-gigabit wavelengths using direct-detection, according to Infinera.
Since early 2016, the optical systems vendors Infinera, Ciena and Nokia have all announced new coherent digital signal processor (DSP) designs. Each new generation of coherent DSP improves the capacity that can be transmitted over an optical link. But given the effectiveness of the latest coherent systems, has most of the benefits already been achieved?
Source: Infinera
“It is getting harder and harder,” admits Kim Roberts, vice president, WaveLogic science at Ciena. “Unlike 10 years ago, there are no factors of 10 available for improvement.”
Non-linear Shannon limit
It is the non-linear Shannon limit that defines how much information can be sent across a fibre, a function of the optical signal-to-noise ratio.
Kim Roberts of CienaThe limit is based on the work of famed mathematician and information theorist, Claude Shannon. Shannon's work was based on a linear communication channel with added Gaussian noise. Optical transport over a fibre is a more complex channel but the same Shannon bound applies, although assumptions for the non-linearities in the fibre must be made.
Roberts stresses that despite much work, the industry still hasn't figured out just what the upper limit is over a fibre for a given optical signal-to-noise ratio.
It is getting harder and harder. Unlike 10 years ago, there are no factors of 10 available for improvement.
"There are papers that show that with this method and this method, you can do this much," says Roberts. "And there are other papers that show that as the power goes up, there is no theoretical limit until you melt the fibre."
These are theoretical things, he says, but the key is that the headroom available remains unknown. What is known is that the theoretical limit remains well ahead of practical systems. Accordingly, systems performance can be improved using a combination of techniques and protocols coupled with advances in electro-optics.
Design goals
A key goal when designing a new optical transmission system is to increase the data sent for a given cost i.e. decrease the cost-per-bit. This is an ongoing requirement as the service providers contend with ever growing network traffic.
Another challenge facing engineers is meeting the demanding power, density and thermal constraints of their next-generation optical transport system designs.
One way to reduce the cost-per-bit is to up the symbol rate to increase the data sent over a wavelength. Traditional 100-gigabit and 200-gigabit dense wavelength-division multiplexing (DWDM) systems use 32-35 gigabaud (GBaud). The latest coherent DSPs already support more than one baud rate: Nokia’s PSE-2s coherent DSP supports 33Gbaud or 45Gbaud while Ciena’s WaveLogic Ai chipset supports 35Gbaud or 56Gbaud.
Having a choice of baud rates coupled with the various modulation scheme options means the same number of bits can be sent over a range of optical reaches. The more complex the modulation scheme, the closer the points are in a constellation and the harder it is to correctly detect the data at the receiver in the presence of noise. Accordingly, using the combination of a simpler modulation scheme and a higher baud rate allows the same data to be sent further.
Capacity-reach is what matters: how much capacity you can extract for a given reach
Nokia's 1.4-billion transistor PSE-2s supports two 200 gigabit-per-second (Gbps) formats: polarisation-multiplexing, 16-ary quadrature amplitude modulation (PM-16QAM) at 33Gbaud, or using PM-8QAM at 45Gbaud. The 200-gigabit wavelength has an optical reach of some 800km using 16-QAM at 33Gbaud but this rises to 1,600km when PM-8QAM at 45Gbaud is used. Alternatively, using 45Gbaud and PM-16QAM, more data can be sent: 250 gigabits-per-wavelength over 800km.
Nokia's Randy EisenachCoherent systems designers are not stopping there. “The next higher baud rate the industry is targeting is 61-68 Gbaud,” says Randy Eisenach, senior product marketing manager, optical networks at Nokia.
Operating at the higher gigabaud range - Infinera talks of 65-70Gbaud - a single transmitter-receiver pair sends twice the amount of data of traditional 32-35Gbaud systems using the same modulation format. But the higher-baud rates require the electro-optics to operate twice as fast. The analogue-to-digital and digital-to-analogue converters of the coherent DSP must sample at twice the baud rate - at least 130 billion samples-per-second. A 65-70Gbaud rate also requires silicon implemented using a more advanced and expensive CMOS process mode - 16nm instead of 28nm. In turn, the optical modulator and drivers need to work well at these higher rates.
“The optical networking industry is well on its way to solving these engineering and component issues in the next year or so,” says Eisenach.
The capacity-per-wavelength also goes up with baud rate. For shorter reach links, 400-600 gigabits-per-wavelength are possible at 65-70Gbaud and, according to Pravin Mahajan, Infinera’s director of product and corporate marketing, power consumption in terms of watts-per-gigabit will improve by some 2.5x.
Pravin Mahajan of InfineraAnd the system vendors are not stopping there: the next baud rate hike after 65-70Gbaud will be in the region of 80-100 Gbaud. The coherent DSPs that will support such data rates will need to be implemented using 7nm CMOS process (see table).
“Capacity-reach is what matters: how much capacity you can extract for a given reach,” says Mahajan. “These successive generations [of faster baud rates] all keep moving that curve upwards.”
DSP features
In addition to the particular baud rates chosen by the vendors for their DSP designs, each includes unique features.
Instead of modulating the data onto a single carrier, Infinera’s FlexCoherent DSP uses multiple Nyquist sub-carriers spread across a channel. The number of subs-carriers varies depending on the link. The benefit of the approach, says Infinera, is that it allows a lowering of the baud rate used which increases the tolerance to non-linear channel impairments experienced during optical transmission.
The FlexCoherent DSP also supports enhanced soft-decision forward-error correction (SD-FEC) including the processing of two channels that need not be contiguous. This is possible as the FlexCoherent DSP is dual-channel which particularly benefits long-haul and subsea applications, claims Infinera. By pairing two channels, the FEC codes can be shared. Pairing a strong channel with a weak one and sharing the codes allows some of the strength of the strong signal to be used to bolster the weaker one, extending its reach or even allowing a more advanced modulation scheme to be used.
Infinera has just announced that by using Nyquist sub-carriers and the FEC gain sharing technologies, its customer, Seaborn Networks, is able delivering 11.8 terabits of capacity over a 10,600km submarine link.
Nokia’s PSE-2s DSP has sufficient processing performance to support two coherent channels. Each channel can implement a different modulation format if desired, or the two can be tightly coupled to form a super-channel. Using 45Gbaud and PM-16QAM, two 250-gigabit channels can be implemented to enable a 500-gigabit muxponder card. The PSE-2s can also implement 400-gigabit wavelength but that is the only format where only one channel can be supported by the PSE-2s.
Ciena’s WaveLogic Ai, meanwhile, uses advanced coding schemes such that it no longer mentions particular modulation schemes but rather a range of line rates in 50-gigabit increments.
Coding schemes with names such as set-partition QPSK, matrix-enhanced PM-BPSK, and 8D-2QAM, have already started to appear in the vendors’ coherent DSPs.
“Vendors use a lot of different terms essentially for the same thing: applying some type of coding to symbols to improve performance,” says Eisenach.
There are two main coding approaches: constellation shaping, also known as probabilistic shaping, and multi-dimensional coding. Combining the two - probabilistic shaping and multi-dimensional coding - promises enhanced performance in the presence of linear and non-linear transmission impairments. These are now detailed.
Probabilistic shaping
The four constellation points of QPSK modulation are equidistant from the origin. With more advanced modulation schemes such as 16-QAM, the constellation points differ in their distance from the origin and hence have different energies. Points in the corners of the constellation, furthest from the origin, have the most energy since a point’s power is the square of the distance from the origin.
Here the origin is at the centre of the square 64-QAM constellation. With probabilistic shaping, more of the points closer to the origin are chosen with the resulting data rate going down. Source: Nokia
Probabilistic shaping uses the inner constellation points more than the outer points, thereby reducing the overall average energy and this improves the signal-to-noise ratio. To understand why, Ciena points out that the symbol error rate at the receiver is dominated by the distance between neighbouring points of the constellation. Reduced the average energy still keeps the distance between the points the same, but when gain is applied to restore the signal’s power levels, the effect is to increase the distance between points. “It means we have better separation between the points, we’ve expanded everything,” says Roberts.
Using probabilistic shaping delivers a maximum 1.53dB of improvement in a linear transmission channel. “That is the theoretical limit,” says Roberts. “In a non-linear world, we get a greater benefit from shaping beyond just shaping the noise.”
Probabilistic shaping also has another benefit: it allows the number of bits sent per symbol to be defined.
Using standard modulation schemes such as 64-QAM with no constellation shaping, 6 bits-per-symbol are sent. Using shaping and being selective in what points are used, fewer bits are sent and they don’t need to be integer values. “I can send 5.7, 5.6, 5.3, even 5.14 bits-per symbol,” says Roberts. “Until I get to 5 bits, and then I have a choice: do I use more shaping or do I start with 32-QAM, which is 5 bits-per-symbol.”
Technology A shows today's coherent DSPs: operating at 30-35Gbaud and delivering 100, 150 and 200Gbps capacities per wavelength. Technology B is Ciena's WaveLogic A. Operating at 56Gbaud, it delivers up to 400Gbps per wavelength in 50Gbps. Technology C will continue this trend. Operating around 70Gbaud, up to 600Gbps per wavelength will be possible in even finer speed increments of 25Gbps. Is this Ciena's next WaveLogic? Source: Ciena
This is very useful as it allows fine control of the data sent such that operators can squeeze just enough data to suit the margins available on a particular fibre link. “You don't have to choose between 100-gigabit and 200-gigabit wavelengths,” says Roberts. "You can use smaller jumps and that sometimes means sending more capacity.”
Three things are needed to fine-tune a link in this way. One is a coherent DSP that can deliver such variable increments on a wavelength using probabilistic shaping. Also needed is a flexible client signalling scheme such as the OIF’s Flexible Ethernet (FlexE) protocol, a protocol mechanism to vary the Ethernet payload for transmission. Lastly, intelligent networking software is required to determine what is happening in the network and the margins available to assess how much data can be squeezed down a link.
Ciena says it has not implemented probabilistic shaping in its latest WaveLogic Ai coherent DSP. But given the Ai will be a family of devices, the technique will feature in upcoming coherent DSPs.
Nokia published a paper at the OFC event held earlier this year showing the use of probabilistic shaping over a transatlantic link. Using probabilistic-shaped 64-QAM (PS-64QAM), a spectral efficiency of 7.46b/s/Hz was achieved over the 5,523km link. This equates to 32 terabits of capacity over the fibre, more than 2.5x the 12 terabits of the existing DWDM system that uses 100Gbps PM-QPSK.
Advanced coding
Multi-dimensional coding is another technique used to improve optical transmission. A 16-QAM constellation is a two-dimensional (2D) representation in one polarisation, says Roberts. But if both polarisations of light are considered as one signal then it becomes a 4D, 256-point (16x16) symbol. This can be further extended by including the symbols in adjacent time slots. This forms an 8D representation.
Non-linear compensation has been an interesting research topic. Nokia continues to investigate the topic and implementation methods but the benefits appear small for most real-world applications
The main two benefits of multi-dimensional coding are better noise performance and significantly better performance in the presence of non-linear impairments.
Nokia’s PSE-2s uses coding for its set-partition QPSK (SP-QPSK). Standard PM-QPSK uses amplitude and phase modulation, resulting in a 4-point constellation. With SP-QPSK, only three of the four constellation points are used for each symbol. A third fewer constellation points means less data is transported but the benefit of SP-QPSK is extended reach due to the greater Euclidean distance between the symbol points created by carefully mapping the sequence of symbols. This results in 2.5dB of extra gain compared to PM-QPSK, for a reach beyond 5,000km.
Using the PSE-2’s 45Gbaud symbol rate, the fewer constellation points of SP-QPSK can be compensated for to achieve the same overall 100Gbps capacity as PM-QPSK at 33Gbaud.
Infinera’s FlexCoherent uses what it calls matrix-enhanced PM-BPSK, a form of averaging that adds 1dB of gain. “Any innovation that adds gain to a link, the margin that you give to operators, is always welcome,” says Mahajan.
Ciena’s WaveLogic 3 Extreme coherent DSP supports the multi-dimension coding scheme 8D-2QAM to improve reach or capacity of long-reach spans.
Such techniques mean vendors have a wealth of available choices available. It is also why Ciena has stopped referring to modulation schemes and talks about its WaveLogic Ai at 35Gbaud supporting 100-250Gbps data rates in 50-gigabit increments while at 56Gbaud, the WaveLogic Ai delivers 100-400Gbps optical channels in 50-gigabit steps.
Probabilistic shaping and multi-dimensional coding are distinct techniques but combining the two means the shaping can be done across dimensions.
Design engineers thus have various techniques to keep improving performance and there are other directions too.
Forward-error correction is about 2dB from the theoretical limit and with improved design Ciena’s Roberts expects 1dB can be reclaimed.
In turn, signal processing techniques could be applied at the transmitter to compensate for expected non-linear effects. “Non-linear compensation has been an interesting research topic,” says Eisenach. “Nokia continues to investigate the topic and implementation methods but the benefits appear small for most real-world applications.”
So is there much scope for further overall improvement?
“There is still a lot more juice left," says Mahajan.
“It [coherent transmission improvement] is getting harder and harder,” adds Roberts. “It is taking more mathematics and more and more CMOS gates, but Moore’s law is providing lots of CMOS gates.”
This is an updated and extended version of an article that first appeared in Optical Connections magazine earlier this year.
Infinera unveils its next-gen packet-optical platforms
Source: Infinera
Infinera has unveiled its latest metro products that support up to 200-gigabit wavelengths using CFP2-DCO pluggable modules.
The XTM II platform family is designed to support growing metro traffic, low-latency services and the trend to move sophisticated equipment towards the network edge. Placing computing, storage and even switching near the network edge contrasts with the classical approach of backhauling traffic, sometimes deep within the network.
“If you backhaul everything, you really do not know if it belongs in that part of the network,” says Geoff Bennett, director, solutions and technology at Infinera. Backhauling inherently magnifies traffic whereas operators want greater efficiencies in dealing with bandwidth growth, he says: “This is where the more cloud-like architectures towards the network edge come in.”
But locating equipment at the network edge means it must fit within existing premises or in installed prefabricated huts where space and the power supplied are constrained.
“If you are asking service providers to put more complex equipment there, then you need low power utilisation,” says Bennett. “This has been a key piece of feedback from customers we have been asking as to how they want our existing products to evolve in the metro-access.”
Having a distributed switch fabric is a long-term advantage for Infinera
Infinera says its latest XTM II products are eight times denser in terms of tranmission capacity while setting a new power-consumption low of 20W-27W per 100 gigabits depending on the operating temperature (25oC to 55oC). Infinera claims its nearest metro equipment competitor achieves 47W per 100 gigabits.
Sterling Perrin, principal analyst, optical networking and transport at Heavy Reading, says Infinera has achieved the power-efficient design by using a distributed switch architecture rather that a central switch fabric and adopting the CFP2-DCO pluggable module with its low-power coherent DSP.
“If you have a centralised fabric and you put it into an edge application then for some cases it will be a perfect fit but for many applications, it will be overkill in terms of capacity and hence power,” says Perrin. “Infinera is able to do it in a modular fashion in terms of just how much capacity and power is put in an application.”
Having a distributed switch fabric is a long-term advantage for Infinera for these applications, says Perrin, whereas competitor vendors will also benefit from the CFP2-DCO for their next designs.
And even if a competitor uses a distributed design, they will not leapfrog Infinera, says Perrin, although he expects competitors’ designs to come down considerably in power with the adoption of the CFP2-DCO.
Infinera has chosen not to use its photonic integrated circuit (PIC) technology for its latest metro platform given the large installed base of XTM chassis that already use pluggable modules. “It would make sense that customers would give feedback that they want a product that has industry-leading performance but which is also backwards compatible,” says Bennett.
Infinera has said it will evaluate whether its PIC technology will be applied to each new generation of the product line. “So when you get to the XTM III they will have another round looking at it,” says Perrin. “If I were placing bets on the XTM III, I would say they are going to continue down this route [of using pluggables].”
Perrin expects line-side pluggable technology to continue to progress with companies such as Acacia Communications and the collaboration between Ciena with its WaveLogic DSP technology and several optical module makers.
“At what point is the PIC going to be better than what is available with the pluggables?” says Perrin. “For this application, I don’t see it.”
XTM II family
Infinera has already been shipping upgraded XTM chassis for the last 18 months in advance of the launch of its latest metro cards. The upgraded chassis - the one rack unit (1RU) TM-102/II, the 3RU TM-301/II and the 11RU TM-3000/II - all feature enhanced power management and cooling.
What Infinera is unveiling now are three cards that enhance the capacity and features of the enhanced chassis. The new cards will work with the older generation XTM chassis (without the ‘II’ suffix) as long as a vacant card slot is available and the chassis’ total power supply is not exceeded. This is important given over 30,000 XTM chassis have been deployed.
The Infinera cards announced are the 400-flexponder, a 200-gigabit muxponder, and the EMXP440 packet-optical transport switch. The distributed switch architecture is implemented using the EMXP440 card.
Operators will also be offered Infinera’s Instant Bandwidth feature as part of the XTM II whereby they can pay for the line side capacity they use: either 100-gigabit or 200-gigabit wavelengths using the CFP2-DCO. The Instant Bandwidth offered is not the superchannel format available for Infinera’s other platforms that use its PIC but it does offer operators the option of deploying a higher-speed wavelength when needed and paying later.
400G flexponder
The flexponder can operate as a transponder and as a muxponder. For a transponder, the client signal and line-side data rate operate at the same data rate. In contrast, a muxponder aggregates lower data-rate client signals for transport on a single wavelength.
Infinera’s 400-gigabit flexponder card uses four 100 Gigabit Ethernet QSFP28 client interfaces and two 200-gigabit CFP2-DCO pluggable line-side modules. Each CFP2-DCO can transport data at 100 gigabits using polarisation-multiplexing, quadrature phase-shift keying (PM-QPSK) modulation or at 200 gigabits using 16-ary quadrature amplitude modulation (PM-16QAM).
The 400-gigabit card can thus operate as a transponder when the CFP2-DCO transports at 100 gigabits and as a muxponder when it carries two 100-gigabit signals over a 200-gigabit lambda. Given the card has two CFP2 line-side modules, it can even operate as a transponder and muxponder simultaneously.
The flexponder card also supports OTN block encryption using the AES-256 symmetric key protocol.
The flexponder is an upgrade on Infinera’s existing 100-gigabit muxponder card. The eightfold increase in capacity is achieved by using two 200-gigabit ports instead of a single 100-gigabit module and halving the width of the line card.
Using the flexponder card, the TM-102/II chassis has a transport capacity of 400 gigabits, up to 1.6 terabits with the TM-301/II and a total of 4 terabits using the TM-3000/II platform.
We can dial back the FEC if you need low latency and don't need the reach
200G muxponder
The double-width 200G card includes all the electronics needed for multi-service multiplexing. The line-side optics is a single CFP2-DCO module whereas the client side can accommodate two QSFP28s and 12 SFP+ 10-gigabit modules. The card can multiplex a mix of services including 10GbE, 40GbE, and 100GbE; 8-, 16- and 32-gigabit Fibre Channel; OTN and legacy SONET/SDH traffic.
Other features include support for OTN block encryption using the AES-256 symmetric key protocol.
The card’s forward error correction performance can also be traded to reduce the traffic latency. “We can dial back the FEC if you need low latency and don't need the reach,” says Bennett.
OTN add-drop multiplexing can also be implemented by pairing two of the multiplexer cards.
EMXP440 switch and flexible open line system
The EMXP440 packet-optical transport switch card supports layer-two functionality such as Carrier Ethernet 2.0 and MPLS-TP. “Mobile backhaul and residential broadband, these are the cards the operators tend to use,” says Bennett.
The two-slot EMXP440 card has two CFP2-DCOs and 12 SFP+ client-side interfaces. The reason why the line side and client side interface capacity differ (400 gigabits versus 120 gigabits) is that the card can be used to build simple packet rings (see diagram, top).
The line-side interfaces can be used for ‘East’ and ‘West' traffic while the SFP+ modules can be used to add and drop signals. The EMXP440 card also has an MPO port such that up to 12 SFP+ further ports can be added using Infinera’s PTIO-10G card, part of its PT Fabric products.
A flexible grid open line system is also available for the XTM II. The XTM II’s 100-gigabit and 200-gigabit wavelengths fit within a 50GHz-wide fixed grid channel but Infinera is already anticipating future higher baud rates that will require channels wider than 50GHz. A flexible grid also improves the use of the fibre’s overall capacity. In turn, RAMAN amplification will also be needed to extend the reach using future higher order modulation schemes such as 32- and 64-QAM.
Infinera says the 400-gigabit flexponder card will be available in the next quarter while the 200-gigabit muxponder and the EMXP440 cards will ship in the final quarter of 2017.
Infinera inches closer to cognitive networking
The second and final part as to how optical networking is becoming smarter
Infinera says it has made it easier for operators to deploy optical links to accommodate traffic growth.
The system vendor says its latest capability, known as Instant Network, also paves the way for autonomous networks that will predict traffic trends and enable capacity as required.
The latest announcement builds on Infinera’s existing Instant Bandwidth feature, introduced in 2012, that uses its photonic integrated circuit (PIC) technology.
Instant Bandwidth exploits the fact that all five 100-gigabit wavelengths of a line card hosting Infinera’s 500-gigabit PIC are lit even though an operator may only need a subset of the 100-gigabit wavelengths. Using Instant Bandwidth, extra capacity can be added to a link - until all five wavelengths are used - in a matter of hours.
The technology allows 100-gigabit wavelengths to be activated in minutes, says Geoff Bennett, director, solutions and technology at Infinera (pictured). It takes several hours due to the processing time for the operator to raise a purchasing order for the new capacity and get it signed off.
Instant Bandwidth has been enhanced since its introduction. Infinera has introduced its latest generation 2.4 terabit PIC which is also sliceable. With a sliceable PIC, individual wavelengths can be sent to different locations using reconfigurable optical add-drop multiplexer (ROADM) technology within the network.
Another feature added is time-based Instant Bandwidth. This allows an operator to add extra capacity without first raising a purchase order. Paying for the extra capacity is dealt with at a later date. This feature has already benefited operators that have experienced a fibre cut and have used Instant Bandwidth to reroute traffic.
Infinera says over 70 of its customers use Instant Bandwidth. These include half of its long-haul customers, its top three submarine network customers and over 60 percent of its data centre interconnect players that use its Cloud Xpress and XTS products. Some of its data centre interconnect customers request boxes with all the licences already activated, says Bennett.
The internet content providers are banging the drum for cognitive networking
Instant Network
Now, with the Instant Network announcement, Infinera has added a licence pool and moveable licences. The result is that an operator can add capacity in minutes rather than hours by using its pool of prepaid licenses.
Equally, if an operator wants to reroute a 100-gigabit or 200-gigabit wavelength to another destination, it can transfer the same licence from the original end-point to the new one.
“They [operators] can activate capacity when the revenue-generating service asks for it,” says Bennett.
Another element of Instant Network still to be introduced is the Automated Capacity Engineering that is part of Infinera’s Xceed software.
Source: Infinera
“Automated Capacity Engineering will be an application that runs on Xceed,” says Bennett. The Automated Capacity Engineering is an application running on the OpenDaylight open source software-defined networking (SDN) controller that takes advantage of plug-ins that Infinera has added to the Xceed platform such as multi-layer path computation and traffic monitoring.
Using this feature, the SDN orchestrator can request a 100 Gigabit Ethernet private line, for example. If there is insufficient capacity, the Automated Capacity Engineering app will calculate the most cost-effective path and install the necessary licences at the required locations, says Bennett.
“We think this is leading the way to cognitive networking,” he says. “We have the software foundation and the hardware foundation for this.”
Networks that think
With a cognitive network, data from the network is monitored and fed to a machine learning algorithm to predict when capacity will be exhausted. New capacity can then be added in a timely accordingly.
Bennett says internet content providers, the likes of Google, Microsoft and Facebook, will all deploy such technology in their networks.
Being consumers of huge amounts of bandwidth, they will be the first adopters. Wholesale operators which also serve the internet content providers will likely follow. Traditional telecom operators with their more limited traffic growth will be the last to adopt such technology.
But cognitive networking is not yet ready. “The machine learning algorithms are still basic,” says Bennett. “But the biggest thing that is missing is the acceptance [of such technology] by network operations staff.”
However, this is not an issue with the internet content providers. “They are banging the drum for cognitive networking,” says Bennett.
Part 1: Ciena's Liquid Spectrum, click here