ECOC '15 Reflections: Part 2
Martin Zirngibl, head of network enabling components and technologies at Bell Labs.
Silicon Photonics is seeming to gain traction, but traditional component suppliers are still betting on indium phosphide.
There are many new start-ups in silicon photonics, most seem to be going after the 100 gigabit QSFP28 market. However, silicon photonics still needs a ubiquitous high-volume application for the foundry model to be sustainable.
There is a battle between 4x25 Gig CWDM and 100 Gig PAM-4 56 gigabaud, with most people believing that 400 Gig PAM-4 or discrete multi-tone with 100 Gig per lambda will win.
Will coherent make it into black and white applications - up to 80 km - or is there a role for a low-cost wavelength-division multiplexing (WDM) system with direct detection?
One highlight at ECOC was the 3D integrated 100 Gig silicon photonics by Kaiam.
In coherent, the analogue coherent optics (ACO) model seems to be winning over the digital coherent one, and people are now talking about 400 Gig single carrier for metro and data centre interconnect applications.
As for what I’ll track in the coming year: will coherent make it into black and white applications - up to 80 km - or is there a role for a low-cost wavelength-division multiplexing (WDM) system with direct detection?
Yukiharu Fuse, director, marketing department at Fujitsu Optical Components
There were no real surprises as such at ECOC this year. The products and demonstrations on show were within expectations but perhaps were more realistic than last year’s show.
Most of the optical component suppliers demonstrated support to meet the increasing demand of data centres for optical interfaces.
The CFP2 Analogue Coherent Optics (CFP2-ACO) form factor’s ability to support multiple modulation formats configurable by the user makes it a popular choice for data centre interconnect applications. In particular, by supporting 16-QAM, the CFP2-ACO can double the link capacity using the same optics.
Lithium niobate and indium-phosphide modulators will continue to be needed for coherent optical transmission for years to come
Recent developments in indium phosphide designs has helped realise the compact packaging needed to fit within the CFP2 form factor.
I saw the level of integration and optical engine configurations within the CFP2-ACO differ from vendor to vendor. I’m interested to see which approach ends up being the most economical once volume production starts.
Oclaro introduced a high-bandwidth lithium niobate modulator for single wavelength 400 gigabit optical transmission. Lithium niobate continues to play an important role in enabling future higher baud rate applications with its excellent bandwidth performance. My belief is that both lithium niobate and indium-phosphide modulators will continue to be needed for coherent optical transmission for years to come.
Chris Cole, senior director, transceiver engineering at Finisar
ECOC technical sessions and exhibition used to be dominated by telecom and long haul transport technology. There is a shift to a much greater percentage focused on datacom and data centre technology.
What I learned at the show is that cost pressures are increasing
There were no major surprises at the show. It was interesting to see about half of the exhibition floor occupied by Chinese optics suppliers funded by several Chinese government entities like municipalities jump-starting industrial development.
What I learned at the show is that cost pressures are increasing.
New datacom optics technologies including optical packaging, thermal management, indium phosphide and silicon integration are all on the agenda to track in the coming year.
ECOC 2015 Review - Part 1
- Several companies announced components for 400 gigabit optical transmission
- NEL announced a 200 gigabit coherent DSP-ASIC
- Lumentum ramps production of its ROADM blades while extending the operating temperature of its tunable SFP+
400 gigabit
Oclaro, Teraxion and NeoPhotonics detailed their latest optical components for 400 gigabit optical transmission using coherent detection.
Oclaro and Teraxion announced 400 gigabit modulators for line-side transmission; Oclaro’s based on lithium niobate and Teraxion’s an indium phosphide one.
NeoPhotonics outlined other components that will be required for higher-speed transmission: indium phosphide-based waveguide photo-detectors for coherent receivers, and ultra-narrow line-width lasers suited for higher order modulation schemes such as dual-polarisation 16-quadrature amplitude modulation (DP-16-QAM) and DP-64-QAM.
There are two common approaches to achieve higher line rates: higher-order modulation schemes such as 16-QAM and 64-QAM, and optics capable of operating at higher signalling rates.
Using 16-QAM doubles the data rate compared to quadrature phase-shift keying (QPSK) modulation that is used at 100 Gig, while 64-QAM doubles the data rate again to 400 gigabit.
Higher-order modulation can use 100 gigabit optics but requires additional signal processing to recover the received data that is inherently closer together. “What this translates to is shorter reaches,” says Ferris Lipscomb, vice president of marketing at NeoPhotonics.
These shorter distances can serve data centre interconnect and metro applications where distances range from sub-100 kilometers to several hundred kilometers. But such schemes do not work for long haul where sensitivity to noise is too great, says Lipscomb.
What we are seeing from our customers and from carriers looking at next-generation wavelength-division multiplexing systems for long haul is that they are starting to design their systems and are getting ready for 400 Gig
Lipscomb highlights the company’s dual integrable tunable laser assembly (iTLAs) with its 50kHz narrow line-width. “That becomes very important for higher-order modulation because the different states are closer together; any phase noise can really hurt the optical signal-to-noise ratio,” he says
The second approach to boost transmission speed is to increase the signalling rate. “Instead of each stream at 32 gigabaud, the next phase will be 42 or 64 gigabaud and we have receivers that can handle those speeds,” says Lipscomb. The use of 42 gigabaud can be seen as an intermediate step to a higher line rate - 300 gigabit – while being less demanding on the optics and electronics than a doubling to 64 gigabaud.
Oclaro’s lithium niobate modulator supports 64 gigabaud. “We have increased the bandwidth beyond 35 GHz with a good spectral response – we don’t have ripples – and we have increased the modulator’s extinction ratio which is important at 16-QAM,” says Robert Blum, Oclaro’s director of strategic marketing.
We have already demonstrated a 400 Gig single-wavelength transmission over 500km using DP-16-QAM and 56 gigabaud
Indium phosphide is now coming to market and will eventually replace lithium niobate because of the advantages of cost and size, says Blum, but lithium niobate continues to lead the way for highest speed, long-reach applications. Oclaro has been delivering its lithium niobate modulator since the third quarter of the year.
Teraxion offers an indium phosphide modulator suited to 400 gigabit. “One of the key differentiators of our modulator is that we have a very high bandwidth such that single-wavelength transmission at 400 Gig is possible,” says Martin Guy, CTO and strategic marketing at Teraxion. “We have already demonstrated a 400 Gig single-wavelength transmission over 500km using DP-16-QAM and 56 gigabaud.”
“What we are seeing from our customers and from carriers looking at next-generation wavelength-division multiplexing systems for long haul is that they are starting to design their systems and are getting ready for 400 Gig,” says Blum.
Teraxion says it is seeing a lot of activity regarding single-wavelength 400 Gig transmission. “We have sampled product to many customers,” says Guy.
NeoPhotonics says the move to higher baud rates is still some way off with regard systems shipments, but that is what people are pursuing for long haul and metro regional.
200 Gig DSP-ASIC
Another key component that will be needed for systems operating at higher transmission speeds is more powerful coherent digital signal processors (DSPs). NTT Electronics (NEL) announced at ECOC that it is now shipping samples of its 200 gigabit DSP-ASIC, implemented using a 20nm CMOS process.
Dubbed the NLD0660, the DSP features a new core that uses soft-decision forward error correction (SD-FEC) that achieves a 12dB net coding gain. Improving the coding gain allows greater spans before optical regeneration or longer overall reach, says NEL. The DSP-ASIC supports several modulation formats: DP-QPSK, DP-8-QAM and DP-16-QAM, for 100 Gig, 150 Gig and 200 Gig rates, respectively. Using two NLD0660s, 400 gigabit coherent transmission is achieved.
NEL announced its first 20nm DSP-ASIC, the lower-power 100 gigabit NLD0640 at OFC 2015 in March. At the same event, ClariPhy demonstrated its own merchant 200 gigabit DSP-ASIC.
Reconfigurable optical add/ drop multiplexers
Lumentum gave an update on its TrueFlex route & select architecture Super Transport Blade, saying it has now been qualified, with custom versions of the line card being manufactured for equipment makers. The Super Transport Blades will be used in next-generation ROADMs for 100 gigabit metro deployments. The Super Transport Blade supports flexible grid, colourless, directionless and contentionless ROADM designs.
“This is the release of the full ROADM degree for next-generation networks, all in a one-slot line card,” says Brandon Collings, CTO of Lumentum. “It is a pretty big milestone; we have been talking about it for years.”
Collings says that the cards are customised to meet an equipment maker’s particular requirements. “But they are generally similar in their core configuration; they all use twin wavelength-selective switches (WSSes), those sort of building blocks.”
This is the release of the full ROADM degree for next-generation networks, all in a one-slot line card. It is a pretty big milestone; we have been talking about it for years
Lumentum also announced 4x4 and 6x6 integrated isolator arrays. “If you look at those ROADMs, there is a huge number of connections inside,” says Collings. The WSSes can be 1x20 and two can be used - a large number of fibres - and at certain points isolators are required. “Using discrete isolators and needing a large number, it becomes quite cumbersome and costly, so we developed a way to connect four or six isolators in a single package,” he says.
A 6x6 isolator array is a six-lane device with six hardwired input/ output pairs, with each input/ output pair having an isolator between them. “It sounds trivial but when you get to that scale, it is truly enabling,” says Collings.
Isolators are needed to keep light from going in the wrong direction. “These things can start to accumulate and can be disruptive just because of the sheer volume of connections that are present,” says Collings.
Tunable transceivers
Lumentum offers a tunable SFP+ module that consumes less than 1.5W while operating over a temperature range of -5C to +70C. At ECOC, the company announced that in early 2016 it will release a tunable SFP+ with an extended temperature range of -5C to +85C.
Further information
Heading off the capacity crunch, click here
For the ECOC Review, Part 1, click here
u2t Photonics pushes balanced detectors to 70GHz
- u2t's 70GHz balanced detector supports 64Gbaud for test and measurement and R&D
- The company's gallium arsenide modulator and next-generation receiver will enable 100 Gigabit long-haul in a CFP2
"The performance [of gallium arsenide] is very similar to the lithium niobate modulator"
Jens Fiedler, u2t Photonics
u2t Photonics has announced a balanced detector that operates at 70GHz. Such a bandwidth supports 64 Gigabaud (Gbaud), twice the symbol rate of existing 100 Gigabit coherent optical transmission systems.
The German company announced a coherent photo-detector capable of 64Gbaud in 2012 but that had an operating bandwidth of 40GHz. The latest product uses two 70GHz photo-detectors and different packaging to meet the higher bandwidth requirements.
"The achieved performance is a result of R&D work using our experience with 100GHz single photo-detectors and balanced detector technology at a lower speed,” says Jens Fiedler, executive vice president sales and marketing at u2t Photonics.
The monolithically-integrated balanced detector has been sampling since March. The markets for the device are test and measurement systems and research and development (R&D). "It will enable engineers to work on higher-speed interface rates for system development," says Fiedler.
The balanced detector could be used in next-generation transmission systems operating at 64 Gbaud, doubling the current 100 Gigabit-per-second (Gbps) data rate while using the same dual-polarisation, quadrature phase-shift keying (DP-QPSK) architecture.
A 64Gbaud DP-QPSK coherent system would halve the number of super-channels needed for 400Gbps and 1 Terabit transmissions. In turn, using 16-QAM instead of QPSK would further halve the channel count - a single dual-polarisation, 16-QAM at 64Gbaud would deliver 400Gbps, while three channels would deliver 1.2Tbps.
However, for such a system to be deployed commercially the remaining components - the modulator, device drivers and the DSP-ASIC - would need to be able to operate at twice the 32Gbaud rate; something that is still several years out. That said, Fiedler points out that the industry is also investigating baud rates in between 32 Gig and 64 Gig.
Gallium arsenide modulator
u2t acquired gallium arsenide modulator technology in June 2009, enabling the company to offer coherent transmitter as well as receiver components.
At OFC/NFOEC 2013, u2t Photonics published a paper on its high-speed gallium arsenide coherent modulator. The company's design is based on the Mach-Zehnder modulator specification of the Optical Internetworking Forum (OIF) for 100 Gigabit DP-QPSK applications.
The DP-QPSK optical modulation includes a rotator on one arm and a polarisation beam combiner at the output. u2t has decided to support an OIF compatible design with a passive polarisation rotator and combiner which could also be integrated on chip. The resulting coherent modulator is now being tested before being integrated with the free space optics to create a working design.
"The performance [of gallium arsenide] is very similar to the lithium niobate modulator," says Fiedler. "Major system vendors have considered the technology for their use and that is still ongoing."
The gallium arsenide modulator is considerably smaller than the equivalent lithium niobate design. Indeed u2t expects the technology's power and size requirements, along with the company's coherent receiver, to fit within the CFP2 optical module. Such a pluggable 100 Gigabit coherent module would meet long-haul requirements, says Fiedler.
The gallium arsenide modulator can also be used within the existing line-side 100 Gigabit 5x7-inch MSA coherent transponder. Fiedler points out that by meeting the OIF specification, there is no space saving benefit using gallium arsenide since both modulator technologies fit within the same dimensioned package. However, the more integrated gallium arsenide modulator may deliver a cost advantage, he says.
Another benefit of using a gallium arsenide modulator is its optical performance stability with temperature. "It requires some [temperature] control but it is stable," says Fiedler.
Coherent receiver
u2t's current 100Gbps coherent receiver product uses two chips, each comprising the 90-degree hybrid and a balanced detector. "That is our current design and it is selling in volume," says Fiedler. "We are now working on the next version, according to the OIF specification, which is size-reduced."
The resulting single-chip design will cost less and fit within a CFP2 pluggable module.
The receiver might be small enough to fit within the even smaller CFP4 module, concludes Fiedler.
OIF promotes uni-fabric switches & 100G transmitter
The OIF's OTN implementation agreement (IA) allows a packet fabric to also switch OTN traffic. The result is that operators can now use one switch for both traffic types, aiding IP/ Ethernet and OTN convergence. Source: OIF
Improving the switching capabilities of telecom platforms without redesigning the switch as well as tinier 100 Gigabit transmitters are just two recent initiatives of the Optical Internetworking Forum (OIF).
The OIF, the industry body tackling design issues not addressed by the IEEE and International Telecommunication Union (ITU) standards bodies, has just completed its OTN-over-Packet-Fabric protocol that enables optical transport network (OTN) traffic to be carried over a packet switch. The protocol works by modifying the line cards at the switch's input and output, leaving the switch itself untouched (see diagram above).
In contrast, the OIF is starting a 100 Gigabit-per-second (Gbps) transmitter design project dubbed the integrated dual-polarisation quadrature modulated transmitter assembly (ITXA). The Working Group aims to expand the 100Gbps applications with a transmitter design half the size of the OIF's existing 100Gbps transmitter module.
The Working Group also wants greater involvement from the system vendors to ensure the resulting 100 Gig design is not conservative. "We joke about three types of people that attend these [working group] meetings," says Karl Gass, the OIF’s Physical and Link Layer Working Group vice-chair. "The first group has something they want to get done, the second group has something already and they don't want something to get done, and the third group want to watch." Quite often it is the system vendors that fall into this third group, he says.
OTN-over-Packet-Fabric protocol
The OTN protocol enable a single switch fabric to be used for both traffic types - packets and time-division multiplexed (TDM) OTN - to save cost for the operators.
"OTN is out there while Ethernet is prevalent," says Winston Mok, technical author of the OTN implementation agreement. "What we would like to do is enable boxes to be built that can do both economically."
The existing arrangement where separate packet and OTN time-division multiplexing (TDM) switches are required. Source: OIF
Platforms using the protocol are coming to market. ECI Telecom says its recently announced Apollo family is one of the first OTN platforms to use the technique.
The protocol works by segmenting OTN traffic into a packet format that is then switched before being reconstructed at the output line card. To do this, the constant bit-rate OTN traffic is chopped up so that it can easily go through the switch as a packet. "We want to keep the [switch] fabric agnostic to this operation," says Mok. "Only the line cards need to do the adaptations."
The OTN traffic also has timing information which the protocol must convey as it passes through the switch. The OIF's solution is to vary the size of the chopped-up OTN packets. The packet is nominally 128-bytes long. But the size will occasionally be varied to 127 and 126 bytes as required. These sequences are interpreted at the output of the switch as rate information and used to control a phase-locked loop.
Mok says the implementation agreement document that describes the protocol is now available. The protocol does not define the physical layer interface connecting the line card to the switch, however. "Most people have their own physical layer," says Mok.
100 Gig ITXA
The ITXA project will add to the OIF's existing integrated transmitter document. The original document addresses the 100 Gigabit transmitter for dual-polarisation, quadrature phase-shift keying (DP-QPSK) for long-haul optical transmission. The OIF has also defined 100Gbps receiver assembly and tunable laser documents.
The latest ITXA Working Group has two goals: to shrink the size of the assembly to lower cost and increase the number of 100Gbps interfaces on a line card, and to expand the applications to include metro. The ITXA will still address 100Gbps coherent designs but will not be confined to DP-QPSK, says Gass.
"We started out with a 7x5-inch module and now there is interest from system vendors and module makers to go to smaller [optical module] form factors," says Gass. "There is also interest from other modulator vendors that want in on the game."
The reduce size, the ITXA will support other modulator technologies besides lithium niobate that is used for long-haul. These include indium phosphide, gallium arsenide and polymer-based modulators.
Gass stresses that the ITXA is not a replacement for the current transmitter implementation. "We are not going to get the 'quality' that we need for long-haul applications out of other modulator technologies," he says. "This is not a Gen II [design].
The Working Group's aim is to determine the 'greatest common denominator' for this component. "We are trying to get the smallest form factor possible that several vendors can agree on," says Gass. "To come out with a common pin out, common control, common RF (radio frequency) interface, things like that."
Gass says the work directions are still open for consideration. For example, adding the laser with the modulator. "We can come up with a higher level of integration if we consider adding the laser, to have a more integrated transmitter module," says Gass.
As for wanting great system-vendor input, the Working Group wants more of their system-requirement insights to avoid the design becoming too restrictive.
"You end up with component vendors that do all the work and they want to be conservative," says Gass. "The component vendors don't want to push the boundaries as they want to hit the widest possible customer base."
Gass expects the ITXA work to take a year, with system demonstrations starting around mid-2013.