BroadLight’s GPON ICs: from packets to apps
BroadLight has announced its Lilac family of customer premise equipment (CPE) chips that support the Gigabit Passive Optical Network (GPON) standard.
The company claims its GPON devices with be the first to be implemented using a 40nm CMOS process. The advanced CMOS process, coupled with architectural enhancements, will double the devices' processing performance while improving by five-fold the packet-processing capability. The devices also come with a hardware abstraction layer that will help system vendors tailor their equipment.
"Traffic models and service models are not stable, and there are a lot of differences from carrier to carrier"
Didi Ivancovsky, BroadLight
Lilac will also act as a testbed for technologies needed for XG-PON, the emerging next generation GPON standard. XG-PON will support a 10 Gigabit-per-second (Gbps) downstream and 2.5Gbps upstream rate, and is set for approval by the International Telecommunication Union (ITU) in September.
Why is this important?
GPON networks are finally being rolled out by carriers after a slow start. Yet the GPON chip market is already mature; Lilac is BroadLight’s third-generation CPE family.
Major chip vendors such as Broadcom and Marvell are also now competing with the established GPON chip suppliers such as BroadLight and PMC-Sierra. “The [big] gorillas are entering [the market],” says Didi Ivancovsky, vice president of marketing at BroadLight.
BroadLight claims it has 60% share of the GPON CPE chip market. For the central office, where the optical line terminal (OLT) GPON integrated circuits (ICs) reside, the market is split between 40% merchant ICs and 60% FPGA-based designs. BroadLight claims it has over 90% of the OLT merchant IC market.
The Lilac family is Broadlight’s response to increasing competition and its attempt to retain market share as deployments grow.
GPON market
US operator Verizon with its FiOS service remains the largest single market in terms of day-to-day GPON deployments. But now significant deployments are taking place in Asia.
“Verizon might still be the largest individual deployer, but China Telecom and China Mobile are catching up fast,” says Jeff Heynen, directing analyst, broadband and video at Infonetics Research. “In fact, the aggregate GPON market in China is now larger than what Verizon has been deploying, given that Verizon’s OLT numbers have really slowed while its optical network terminal (ONT) shipments remain high at some 200,000 per quarter.”
“Verizon might still be the largest individual deployer, but China Telecom and China Mobile are catching up fast”
Jeff Heynen, Infonetics Research
Chinese operators are deploying both GPON and Ethernet PON (EPON) technologies. According to BroadLight, Chinese carriers are moving from deploying multi-dwelling unit (MDU) systems to single family unit (SFU) ONTs.
An MDU deployment involves bringing PON to the basement of a building and using copper to distribute the service to individual apartments. However such deployments have proved less popular that expected such that operators are favouring an ONT-per-apartment.
“Through this transition, China Telecom and China Unicom are also making the transition to GPON,” says Ivancovsky. “End–to-end prices of EPON and GPON are practically the same,” GPON has a download speed of 2.5Gbps and an upload speed of 1.25Gbps (Gbps) while for EPON it is 1.25Gbps symmetrical.
China Telecom and China Unicom are deploying GPON is several provinces whereas in major population centres the PON technology is being left unspecified; vendors can propose either the use of EPON or GPON. “This is a big change, really a big change,” says Ivancovsky.
In India, BSNL and a handful of private developers have been the primary GPON deployers, though the Indian market is still in its infancy, says Heynen. Ericsson has also announced a GPON deployment with infrastructure provider Radius Infratel that will involve 600,000 homes and businesses.
“I expect there to be a follow-on tender for BSNL later this year or early next year that will be twice the size of the first tender of 700,000 total GPON lines,” says Heynen. He also expects MTNL to begin deploying GPON early next year.
In other markets, Taiwan’s Chunghwa Telecom has issued its first tender for GPON. Telekom Malaysia is deploying GPON as is Hong Kong Broadband Network (HKBN), while in Australia the National Broadband Network Company will roll-out a 100Mbps fibre-to-the-home network to 90% of Australian premises over eight years working with Alcatel-Lucent.
“Let’s not forget about Europe, which has been basically dormant from a GPON perspective,” adds Heynen. “We now have commitments from France Telecom, Deutsche Telekom, and British Telecom to roll out more FTTH using GPON, which should help increase the overall market, which really was being driven by Telefonica, Portugal Telecom, and Eitsalat.”
Infonetics expects 2010 to be the first year that GPON revenue will exceed EPON revenue: US $1.4 billion worldwide compared to $1.02 billion. By 2014, the market research firm expects GPON revenue to reach $2.5 billion with EPON revenue - 1.25Gbps and 10Gbps EPON - to be US $1.5 billion. “At this point, China, Japan, and South Korea will be the only major EPON markets with many MSOs also using EPON for FTTH and business services,” says Heynen.
What’s been done?
BroadLight offers a range of devices in the Lilac family. It has enhanced the control processing performance of the CPE devices using 40nm CMOS, and has also added more network processor unit (NPU) cores to enhance the ICs’ data plane processing performance.
“It’s been the same story for some time now,” says Heynen. “System-on-chip vendors differentiate themselves on four key aspects: footprint, power consumption, speed and, most importantly, cost”
A key driver for upping the Lilac family’s control processor’s performance is to support the Java programming language and the OSGi Framework, says Ivancovsky. The OSGi Framework used with embedded systems has yet to be deployed on gateways but is becoming mandatory. This will enable the CPE gateway to run downloaded applications much as applications stores now complement mobile handsets.
BroadLight has also doubled to four the on-chip RunnerGrid NPU cores. “Traffic models and service models are not stable, and there are a lot of differences from carrier to carrier” says Ivancovsky. “This [on-chip] flexibility helps us to have a single device that can support many different requirements.”
As an example, Broadlights cites South Korean operator, SK Broadband, which is deploying an ONT supporting two gigabit Ethernet (GbE) ports – one for laptops and the other for the home’s set-top box. Thus a single GPON 2.5Gbps stream is delivered down the fibre and shared between the PON’s 32 or 64 ONTs, with each ONT having two 1GbE links.“The carrier wants to limit the IPTV downstream rate according to the service level agreement,” says Ivancovsky. Having the network processor as part of the CPE, the carrier can avoid deploying more more expensive NPUs at the central office.
The overall result is a Lilac family rated at 2,000 Dhrystone MIPS (DMIPS) and a packet processing capability of 15 million packets per second (Mpps) compared to BroadLight’s current-generation CPE family of 650-900 DMIPs and 3Mpps.
“Broadlight understands that you have to have a range of chips that provide flexibility across a wide range of CPE and infrastructure types,” says Heynen. The CPE demands of a Verizon differ markedly from those of China Telecom, for example, primarily because average-revenue-per-user expectations are so different. Verizon wants to provide the most advanced integrated CPE, with the ability to do TR-069 remote provisioning and both broadcast and on-demand TV, while China Telecom is concerned with achieving sustained downstream bandwidth, with IPTV being a secondary concern, he says.
Heynen also highlights the devices’ software stack with its open application programming interfaces (APIs) that allow third-party developers to develop applications on top of features already provided in BroadLight’s software stack.
“Residential gateway software stacks used to be dominated by Jungo (NDS). But now chipset companies are developing their own, which helps to reduce licensing costs on a per CPE basis,” says Heynen. “If a silicon vendor can provide a hardware abstraction layer like this, it makes it very attractive to system vendors who need an easy way to customise feature sets for a wide range of customers.”
Is the Lilac GPON family fast enough to support XG-PON?
“We are deep in the design of XG-PON end-to-end: one team is working on the OLT and one on the ONT,” says Ivancovsky. “We expect an FPGA prototype very early in 2011.”
The first XG-PON product will be an OLT ASIC rated at 40Gbps: supporting four XG-PON or 16 GPON ports. One XG-PON challenge is developing a 10Gbps SERDES (serialiser/ deserialiser), says Ivancovsky: “The SERDES in Lilac is a 10Gbps one, a preparation for XG-PON devices.”
Meanwhile, the first XG-PON CPE design will be implemented using an FPGA but the control processor used will be the one used for Lilac. As for data plane processing, NPUs will be added to the OLT design while more NPUs cores will be needed within the CPE device. “The number of cores in the Lilac will not be enough; we are talking about 40Gbps,” says Ivancovsky.
Lilac device members
Ivancovsky highlights three particular devices in the Lilac family that will start appearing from the fourth quarter of this year:
- The BL23530 aimed at GPON single family units with VoIP support. To reduce its cost, a low-cost packaging will be used.
- The BL23570 which is aimed at the integrated GPON gateway market.
- The BL23510, a compact 10x10mm IC to be launched after the first two. The chip’s small size will enable it to fit within an SFP form-factor transceiver. The resulting SFP transceiver can be added to connect a DSLAM platform, or upgrade an enterprise platform, to GPON.
“This new system-on-chip is a technology improvement, especially with respect to the residential gateway software layer, which is a requirement among most operators,” concludes Heynen. “But it should be noted this is an addition to, not a replacement for, existing BroadLight chips that solve different infrastructure requirements.”
ROADMS: When "-less" is more
One only has to look at neighbouring IT and cloud computing in particular with its PaaS, IaaS and SaaS (Platform-, Infrastructure- and Software-as-a-Service).
But when it comes to agile optical networking and the reconfigurable optical add-drop multiplexer (ROADM), what is notable is the smarts that are being added and yet all are described using the “-less” suffix: colourless, directionless, contentionless and gridless.
These are all logical names once the enhancements they add are explained. But as Infonetics Research analyst Andrew Schmitt has pointed out, the industry could do better with its naming schemes. Even the most gifted sales person may be challenged selling the merits of a colourless, directionless product.
Colourless is a term long in use for such optical devices as arrayed-waveguide gratings. So to expect the industry to change now is perhaps unrealistic. But could better names be chosen? And does it matter?
Well, yes, if it undersells the benefits new products deliver.
The four smarts
Colourless refers to the decoupling of the wavelength dependency, so is “wavelength independent” better? What about colourful? Sales people are on a better footing already.
Then there is directionless. The idea here is that the latest ROADMs have full flexibility in routing a lightpath to any of the network interface ports. So instead of directionless, what about ROADMs that are omnidirectional or all-directional?
"Even the most gifted sales person may be challenged selling the merits of a colourless, directionless product."
Contentionless means non-blocking, a well-known term widely used to describe switch and router designs.
And gridless comes from the concept of relaxing the rigid ITU grid for wavelengths. Again, a perfectly logical name. But it sells short the adaptive channel widths that new ROADMs will support for data rates above 100 Gigabit-per-second.
So third-generation ROADMs are colourless, directionless, contentionless and gridless products. But does colourful, all-directional, non-blocking and adaptive-channel ROADMs sound better?
Suggestions welcome.
Alcatel-Lucent reveals its 100Gbps-coherent hand
“It would be irresponsible of any system vendor to overlook a solution that can bring a cost advantage to their customer”
Sam Bucci, Alcatel-Lucent
What is being announced?
- Alcatel-Lucent has a commercially available 100Gbps optical transmission system.
- 40Gbps coherent transmission is also supported.
- Implemented as part of the 1830 Photonic Service Switch (PSS), the platform has a capacity of 500 Gigabit-per-second (Gbps) per rack, or a bay – made up of three racks – capacity of 1.5 Terabit-per-second.
- The system specification includes 88, 100Gbps dense wavelength division multiplexing (DWDM) lightpaths at 50GHz spacing that span the extended C-band; a reach of between 1,500 and 2,000km using coherent-optimised optical amplifiers, and the ability to operate alongside existing 10 and 40Gbps wavelengths without needing a guard-band between them (for more detail, click here).
- Some 20 operators are lined up to trial the 100Gbps technology. These include operators that have deployed the 1830 PSS and new ones.
- Telefònica and Softbank Telecom are two operators known to be trialling the 100Gbps system, Alcatel-Lucent will announce a third next week.
Why is the announcement important?
Alcatel-Lucent is the latest system vendor to announce a commercially available 100Gbps system. Until now Nortel’s Metro Ethernet Networks unit, now owned by Ciena, and Ciena itself had commercially available systems. Indeed Verizon Business deployed Nortel’s 100Gbps system for a route linking Paris and Frankfurt in late 2009.
"What will be a significant differentiator is the control/ management plane interworking across platforms - the integration of IP MPLS with optical networking products."
Ron Kline, Ovum
Alcatel-Lucent claims to be the first vendor to offer a 100Gbps system using a single carrier. Ciena/Nortel’s current offering is an extension of its 40Gbps coherent system and uses two 50Gbps sub-carriers that fit into a 50GHz channel.
But analysts downplay the significance of the advent of a 100Gbps single-carrier system. “Technology leads are short-lived,” says Ron Kline, principal analyst, network infrastructure at Ovum. "I’m not sure if there is a preference between single- versus dual-carrier from service providers either.”
What will be a significant differentiator, says Kline, is the control/ management plane interworking across platforms - the integration of IP MPLS with optical networking products. “Alcatel-Lucent is one of the few vendors which do both well and may have an edge pulling it off,” he says.
Ovum’s Dana Cooperson thinks it is significant that, like Ciena, and unlike some others, Alcatel-Lucent is also doing 40Gbps coherent. “I’ve heard some folks say they think 40 Gig coherent isn’t going anywhere, but the reasoning hasn’t made sense to me,” says Cooperson, Ovum’s vice president, network infrastructure. “If you have bad fibre, which loads of carriers do, and you want a mixed channel capability, which all carriers do, you’ll expect to get both in the same product.”
What’s been done?
Alcatel-Lucent’s 100Gbps system implements polarisation multiplexing quadrature phase-shift keying (PM-QPSK) modulation with coherent detection. The coherent receiver is based on an in-house application specific integrated circuit (ASIC) that includes high-speed analogue-to-digital (a/d) converters and a digital signal processor (DSP).
Alcatel-Lucent would not say if the ASIC uses a 60nm or 45nm CMOS process or what the sampling rate of its a/d converters are but it did say that it has built-in sufficient headroom to operate at a 64Gsamples-per-second rate. The system also uses hard-decision forward error correction (FEC) but, according to Sam Bucci, vice president, terrestrial portfolio management at Alcatel-Lucent, it is looking at a soft decision FEC scheme for a future version “that is not too far away”.
Additional system performance characteristics, according to Bucci, include the ability for the lightpath to travel through as many as 20 reconfigurable optical add/drop multiplexers (ROADMs) before needing optical-electrical-optical (OEO) conversion. The system also has a tolerance of at least 30ps for polarisation mode dispersion and 60,000 ps/nm for chromatic dispersion, says Bucci.
For 40Gbps coherent transmission, Alcatel-Lucent is using polarisation multiplexing binary phase-shift keying (PM-BPSK). Since less information is encoded on the symbol streams, this is a more demanding implementation because the implementation operates at 20Gbaud-per-second rather than the 10Gbaud-per-second of 40Gbps PM-QPSK coherent systems.“We were looking for a solution that was applicable not just for terrestrial but for submarine,” says Bucci. “Therefore the reach we were looking to achieve was greater than perhaps could be accomplished by other modulation formats.”
Alcatel-Lucent says PM-BPSK is also better able to withstand non-linear effects such as cross-phase modulation.
Is Alcatel-Lucent open to adopting an ASIC from a third-party developer for its future 100Gbps systems? “It would be irresponsible of any system vendor to overlook a solution that can bring a cost advantage to their customer,” says Bucci. “If there is a solution that can fit into the scheme we have developed, then yes, we would have to consider it if it produces an economic advantage.” Such an 'economic advantage' would have to be significantly more than just a 10 percent cost-saving, he says.
Volume production of the 100Gbps system will begin at the end of June 2010. Two client-side interface boards are available: a 10x10Gbps and a 100Gbps native port using a pluggable CFP transceiver.
The InfiniBand roadmap gets redrawn
“We can already demonstrate in silicon a 30Gbps transmitter."
Marek Tlalka, Luxtera
“Our June 2008 roadmap originally projected 4x EDR at less than 80Gbps data rate for 2011,” says Skip Jones, director of technology at QLogic and co-chair of the IBTA’s marketing working group. “The IBTA has increased the data speeds for 2011 due to demand for higher throughput.” A 26Gbps channel rate - or 104Gbps for 4x EDR - is to accommodate the overhead associated with 64/66bit encoding.
The IBTA has also added an interim speed, dubbed Fourteen Data Rate (FDR), operating at 14Gbps per channel or 56Gbps for 4x FDR. This, says the IBTA, is to address midrange enterprise applications in the data centre. “Many server OEMs’ backplanes can support speeds up to 56Gbps,” says Jones. “For those OEMs doing a server refresh using existing backplanes, 56Gbps will be the solution they’ll be looking to implement.”
The IBTA dismisses claims by some industry voices that the re-jigged roadmap is to stop InfiniBand falling behind 100 Gigabit Ethernet (GbE) while FDR is to advance InfiniBand while laser vendors grapple with the challenge of developing 26Gbps vertical-cavity surface-emitting lasers (VCSELs) for EDR.
Jones points out that 4x Quad Data Rate (QDR) InfiniBand (4x10Gbps) now accounts for between 60 and 70 percent of newly deployed InfiniBand systems, and that 100Gbps EDR will appear in 2011/ 2012. “The IBTA has a good track record of releasing products on time; as such, 100Gbps InfiniBand will come out much faster than 100 Gigabit Ethernet.” FDR, meanwhile, will benefit from 14Gbps VCSELs for Fibre Channel that will be available next year. Jones admits that developing a 26Gbps VCSEL poses a challenge but that “InfiniBand markets are mostly electrical interconnects”.
“The 4x25G short reach is not going to rise and dominate for quite awhile."
Scott Schube, LightCounting
“VCSELs are going to have a tough time at 26Gbps per lane, though they'll get there,” says Scott Schube, senior analyst and strategist at optical transceiver market research firm, LightCounting. “There's definitely a push to go to 26Gbps per lane to reduce pin counts, and the chip guys look like they will be ready before the VCSELs.”
One company looking to benefit from the emerging market for EDR is Luxtera. The silicon photonics specialist says its modulator has already been demonstrated at 30Gbps. This is fast enough to accommodate EDR, 100 Gigabit Ethernet (a 4-channel design) and the emerging 28Gbps Fibre Channel standard.
“We can already demonstrate in silicon a 30Gbps transmitter using the same laser as in our existing products and modulated in our silicon waveguides,” says Marek Tlalka, vice president of marketing at Luxtera. “That allows us to cover 14Gbps, 26Gbps EDR, parallel Ethernet as well as 28Gbps for serial Fibre Channel.”
Luxtera will need to redesign the transistor circuitry to drive the modulator beyond the current 15Gbps before the design can be brought to market. It will also use an existing silicon modulator design though the company says some optimisation work will be required.
There are two main product offerings from Luxtera: QSFP-based active optical cables and OptoPHY, one and four-channel optical engines. Luxtera’s OptoPHY product is currently being qualified and is not yet in volume production.
For multi-channel designs, Luxtera uses a continuous-wave 1490nm distributed feedback (DFB) laser fed to the modulated channels. Addressing 28Gbps Fibre Channel, an SFP+ form factor will be used. Luxtera may offer a transceiver product or partner with a module maker with Luxtera providing the optical engine. “It’s an open question,” says Tlalka.
“The IBTA has a good track record of releasing products on time; as such, 100Gbps InfiniBand will come out much faster than 100 Gigabit Ethernet.”
Skip Jones, IBTA
The company has said that the single-channel and four-channel 10Gbps OptoPHY engine consumes 450mW and 800mW respectively. Going to 26Gbps will increase the power consumption but only by several tens of percent, it says.
The first product from Luxtera will be a pluggable cable followed by a companion OptoPHY. The pluggable active optical cable from Luxtera will support 100GbE and EDR Infiniband. “I’d still place my bets on InfiniBand deploying first followed by 100GbE,” says Tlalka.
But Schube warns that Luxtera faces a fundamental challenge “Leading-edge designs based on proprietary technology to solve commodity problems - more bandwidth for out-of-the-box connections - are never going to get widely adopted, though Luxtera can fill a niche for awhile," he says.
There is also much work to be done before 100Gbps interfaces will be deployed. “The 4x25G short reach is not going to rise and dominate for quite awhile, no matter what the component availability is,” says Schube. That is because switch ASICs, backplanes, connectors and line cards will all first need to be redesigned.
Meanwhile the IBTA has also announced two future placeholder data rates on its InfiniBand roadmap: High Data Rate (HDR) due in 2014 and the Next Data Rate (NDR) sometime after. “We will refrain from identifying the exact lane speed until we are closer to that timeframe to avoid confusion and the possibility - and probability - of changing future lane speeds,” says Jones.
And Luxtera says its modulator can go faster still. “I think we can easily go 40 and 50Gbps,” says Tlalka. “After 50Gbps we’ll have to look at new magic.”
Infinera PICs 100Gbps coherent
Infinera is expediting its product plans, basing its optical transmission roadmap on coherent detection.
The company plans to launch a 100Gbps coherent transmission system in 2012. The design will be based on a pair of 5x100Gbps ultra-long-haul photonic integrated circuit (PIC) chips that will enable its systems to deliver 8 Terabits-per-second (Tbps) over a fibre.
“This change in roadmap is because of the successful development of our 100G coherent ASIC programme, and we have integrated five 100Gbps coherent channels onto one card.”
Drew Perkins, Infinera
Infinera also announced that it will be adding 40Gbps coherent detection to its DTN system in 2011. The 40Gbps will be based on optical modules and not its PIC technology. Using its planar technology and working with optical module suppliers to integrate its in-house coherent technology, Infinera’s DTN system will support 25GHz channel spacings to cram 160 lightpaths across the C-band, to deliver 6.4Tbps capacity.
Why is the announcement important?
Infinera had still to launch its 10x40Gbps PIC. This announcement marks a shift in Infinera’s strategy to focus on 100Gbps and gain a technology edge by offering the highest line speed at an unmatched density.
“It’s a good roadmap for Infinera,” says Jimmy Yu, a director at the Dell'Oro Group. “From an optical market perspective, I think 2012 is the right time for having a 100Gbps DWDM long-haul system. And it'll definitely be coherent.”
Dell’Oro expects to see early adopters of 100Gbps in 2010 and 2011, but it will be 2012/2013 when the market for 100Gbps will ramp.
What has motivated Infinera’s shift has been its success in developing coherent technology, says Drew Perkins, Infinera’s CTO. Coherent technology in combination with PICs is the best of all worlds, he says, marrying the two most significant optical developments of the last decade.
Perkins admits Infinera has been slow in offering 40Gbps technology.
“We are late to a very small market,” he says. “We think there is a 40G squeeze going on – it took the industry so long to get 40Gbps right with coherent technology such that 100Gbps is now just around the corner, as we are proving here.”
Yet Infinera will offer 40Gbps next year and will seek to differentiate itself with 25GHz channel spacing. “But it [the 40Gbps design] will be rapidly superseded by our 100Gbps, 8Tbps technology and then we believe we will be early to market with 100Gbps,” says Perkins.
Dell’Oro says 40Gbps is growing rapidly and it expects continued growing. “In 2009, 40Gbps wavelength shipments grew a little over 160 percent, and we’re forecasting it to grow nearly 90 percent in 2010,” says Yu. “If Infinera delivers 40Gbps on 25GHz channel spacing, it'll be a good interim step to 100Gbps.”
What’s being done?
Infinera has now scrapped its 10x40Gbps differential quadrature phase-shift keying (DQPSK) PIC, going to a 5x100Gbps polarisation multiplexing quadrature phase-shift keying (PM-QPSK) design instead. Interestingly, Perkins says that the 10x40Gbps transmitter PIC was designed from the start to also support 5x100Gbps PM-QPSK modulation.
The challenge is designing the coherent receiver PIC which is significantly different, and has required Infinera to gain coherent expertise in-house.
The receiver PIC also requires a local oscillator laser. “We have integrated the laser onto the receiver PIC per channel,” says Perkins. Infinera’s PICs already use lasers that are tuned over a significant number of channels though not the whole C-band so this is using technology it already has.
Another key aspect of the coherent receiver is the associated electronics that comprises very high-speed A/D converters, a digital signal processor and most likely advanced forward error correction. Developing such an ASIC is a significant challenge.
Is Infinera developing such a design? Infinera points to its Ottawa, Ontario-based research facility that was announced in September last year. “That team is working on ASIC level coherent technology,” says Perkins. “This change in roadmap is because of the successful development of our 100G coherent ASIC programme, and we have integrated five 100Gbps coherent channels onto one card.”
Did Infinera consider designing a 10x100Gbps PIC? “It comes down to the size of the line card,” says Perkins. Infinera believes the resulting terabit line card would have been too large a jump for the industry given the status of associated electronics such as switching technology.
What next?
Infinera says that in 2012 it will ship systems based on its 100Gbps coherent PICs to customers but it is unwilling to detail the key development milestones involved between now and then.
As for future product developments, Infinera claims it can extend overall capacity of its coherent technology in several directions.
It says it can integrate 10, 100Gbps channels onto a PIC. “Somewhere in the future we undoubtedly will”, says Perkins. The company also states that in the “fullness of time” it could deliver 100Gbps over 25GHz channel spacings.
Perkins also reconfirmed that Infinera will continue to advance the modulation scheme used, going from QPSK to include higher order quadrature amplitude modulation (QAM) schemes.
Is a datacom and telecom mini-boom taking place?
Daryl Inniss believes it is largely a reflection of cutbacks that have run their course. “The industry cut back swiftly and deeply when the market started to tank, cutting suppliers and capacity,” says Inniss, practice leader, components at market research firm Ovum.
“I think it's recovery dynamics - people ordering a tiny bit more and there are no parts available such that lead-times are stretching out simulating a boom.”
Brad Smith, LightCounting
Carriers supported the demand with inventory. “Now the industry needs to support both deployments and inventory and with the traffic continuing to grow suppliers cannot meet demand,” he says. Moreover this “bull-whip effect” impacts most severely suppliers furthest removed from the carriers i.e. component vendors.
Brad Smith, senior vice president at optical transceiver market research firm LightCounting, also explains the situation based on events last year.
“There is a shortage of certain parts in optical and semis as a result of cutbacks in manufacturing during 2009,” says Smith, “I think it's recovery dynamics - people ordering a tiny bit more and there are no parts available such that lead-times are stretching out simulating a boom.”
Late last year a research note highlighted industry reports that shortages were becoming more widespread, including components such as integrated circuits and fiber optic transceivers.
However one leading optical transceiver vendor commented that it is shipping everything it can make and that it can’t build stuff fast enough.
So is there a mini-boom after all?
DSL: Will phantom channels become real deployments?
Alcatel-Lucent is promoting its DSL Phantom Mode technology as a complement to fibre-to-the-x (FTTx) technology. Operators can use the technology to continue to extend services offerings to existing DSL subscribers as they roll out FTTx over the next decade or more.
But one analyst believes the technology could take years to commercialise and questions whether the announcement is not sending a wrong message to the industry by providing an alternative to fibre.
“The investment required to upgrade DSL is quite small”
Stefaan Vanhastel, Alcatel-Lucent
What has been achieved?
The 300Mbps data rate is achieved using two copper wire pairs between the access equipment and a DSL modem although three DSL ports are required at each end. The rate drops to 100Mbps when the reach is extended to 1km. In comparison very high speed Digital Subscriber Line 2’s (VDSL2) data rate over a single line ranges from 20 to 40Mbps over 1km.
None of the three techniques that Alcatel-Lucent uses – bonding, vectoring and the phantom mode that creates an extra virtual channel alongside the two bonded pairs - is new. What the company claims is that it is the first to combine all three for DSL.
In March Ericsson announced it had achieved 500Mbps over 500m but it used six bonded pairs and vectoring only.
Why is the Phantom Mode important?
The significance of the announcement, according to Alcatel-Lucent, is that operators can continue to offer existing DSL customers new bandwidth-intensive services as they roll out FTTx.
“Rolling out FTTx will take a significant amount of time,” says Stefaan Vanhastel, director of product marketing, wireline networks at Alcatel-Lucent. “Operators are looking to reuse their copper infrastructure in the short-to-medium term - the next 5 to 10 years.”
An operator must have a central office or cabinet equipment 1km or less from the user’s residence as well as having two wire pairs per building or residence. “In many countries two pairs are available,” says Vanhastel.
However, one analyst questions the development and promotion of such copper-enhancing technology.
“I think Alcatel is being disingenuous when they say "fiber will take long to implement, this is an intermediary solution’,” says the analyst, who asked not to be named. “They know full well that customers would see this as a way to hold back on deploying fibre.
“Ultimately to me this is schizophrenia at work. Alcatel-Lucent wants to be all things to all service providers and may be sending the wrong message to the market that they need not invest to sustain the bandwidth demand growth, which is suicidal both for service providers and for Alcatel-Lucent in the long run.”
Alcatel-Lucent does believe operators will invest in DSL alongside FTTx.
“The investment required to upgrade DSL is quite small,” says Vanhastel. “Even with two ports it is a bargain; you get the investment back in one or two months.”
Even operators more advanced in their FTTx deployments will want to offer new higher bandwidth services such as high-definition TV to all their customers.
“What are you going to do? Offer your services to just 50% of your customers?” says Vanhastel “They [the remaining customers] will go elsewhere.”
Method used
The Bell Labs research arm of Alcatel-Lucent has used three techniques to enhance DSL’s speed and reach performance.
- Bonding: The combination of copper line pairs to boost the number of channels – in this case two are bonded - and hence the data rate between access equipment and the DSL modem.
- Vectoring: Noise cancellation techniques using digital signal processing to improve the overall signal-to-noise performance. “It involves measuring the noise on all the lines and generating anti-phase – the inverse signal – such that the two cancel out,” says Vanhastel.
- Phantom mode: The phantom mode technology uses two physical wires to create a third virtual one. The technology was first proposed in the 1880s as a way to add an extra virtual telephone line.
Two physical pairs and the third phantom one. Source: Alcatel-Lucent
Using the phantom mode, only two wire pairs are needed to connect the end equipment. The information on the third “virtual” line is shared over the two physical channels. Using analogue electronics, the data on the third channel is processed and recovered. “We add and subtract through the use of a bunch of transformers,” says Vanhastel. Where the circuitry is placed, whether in the DSLAM access equipment or elsewhere, is to be decided.
To create the virtual wire, a modem supporting three-pair bonding is required. In addition the chipset in the DSL modem must have sufficient processing performance to execute vectoring on three channels. That's because adding the phantom mode degrades the performance of all the channels due to crosstalk. The crosstalk is removed between the channels using vectoring.
What next?
The technology needs to be brought to market. “At the earliest it will be 2012,” says Vanhastel.
But the analyst points out that the technology is lab tested: “Between test labs and implementation, count a significant number of years.”
The concept could even be extended using more wire pairs. The relationship is (N-1) phantom channels for N wire pairs i.e. 1 virtual channel with two pairs, 2 with 3 pairs etc.
Alcatel-Lucent says it has already completed two VDSL2 bonding trials in Asia Pacific, while three operators are undertaking VDSL2 vectoring tests in their labs and will move to testing in the field using a single line this year.
“Bonding is here today, vectoring will be 2011 and the phantom mode will be after that,” says Vanhastel.
OFC/NFOEC 2010: Industry reflections
Here is a selection of their views.
“We heard again and again, that the internet service providers such as Google are still looking for solutions for their future bandwidth demand”
Andreas Umbach, u2t Photonics
OFC highlights
For many the story of the show was 40Gbps and 100Gbps long-haul transport.
“Of the many announcements I thought Opnext’s component technology and product announcements were notable as a big bet on bleeding-edge R&D by an established components player,” says Scott Schube, senior analyst and strategist at market research firm, LightCounting.
The developments Schube highlights include Opnext’s 100Gbps SerDes, 100Gbps DSP for coherent receivers, and various optical modules for 100Gbps, along with its field trial with AT&T announced before the show.
Andreas Umbach, CEO of u2t Photonics, notes the industry acceptance of dual-polarisation QPSK as the standard for 100Gbps. Oclaro also highlights the Optical Internetworking Forum’s (OIF) work towards a single solution, to avoid the proliferation of solutions that occurred at 40Gbps. “Most view this as a good sign, but it does remove some of the flexibility to introduce new innovations,” says Chris Clarke, vice president strategy and chief engineer at Oclaro.
OFC/NFOEC 2010 also marked the return of the system vendors, according to Karen Liu, vice president, components and video technologies at market research firm Ovum.
“There were two big trailers on the floor from Huawei and Ciena-Nortel. Each was a complete demo room showing multiple systems interoperating,” she says. “Ciena-Nortel with Corning had a demo involving 3000km of fibre.”
But others disagree. “The box makers have really pulled back at this show compared to those in the past,” says Neal Neslusan, a consultant at PhiBos Consulting. “With Supercomm now finally dead it makes one wonder what is the target show for the telecom box makers.”
“The biggest thing was the general sense that bandwidth on the fibre is no longer plentiful”
Karen Liu, Ovum
One datacom demonstration highlighted by LightCounting was Avago’s microPOD 120Gbps parallel optical modules shown as part of an IBM supercomputing blade. “Reflex Photonics and others have showed similar technology before, but Avago’s announcement shows that we might be getting close to true commercialization of “optics on board” products,” said Schube.
Liu was also taken by IBM’s water-cooled processor board with the Avago optical interconnect based on a novel connector made by injection-molded high-precision lensing: “There were 28 modules, 120Gbps each, on the board.”
What else besides 40 and 100Gbps?
Passive Optical Networks (PON) received huge attention, says David Menashe, vice president and chief scientist at RED-C Optical Networks. “There is a lot of activity around next-generation PON architectures including higher rates and longer reach, merging of access and metro networks and reducing the number of central offices,” he says
LightCounting highlights continuing innovation at 10Gbps which, it says, still has years of market dominance ahead of it. Schube notes the demos from several IC vendors of smaller, lower power, more integrated 10G PHYs and framers, as well as two companies introduced long-reach SFP+ modules, and Finisar’s demonstration of a new tunable XFP module.
“The tunable XFP seemed to gain general market acceptance with several suppliers announcing upcoming releases,” says Sinclair Vass, senior EMEA director at JDS Uniphase. He also noted a trend towards 1x23 type wavelength-selective switch (WSS) architectures with market agreement emerging as to what is required.
Learnt
For David Smith, CTO of CIP Technologies, the general acceptance of coherent detection as the solution for 100Gbps and beyond means that the two most important technologies driving the future of the industry will be photonic integration needed to realise the complex optics, and advances in electronic DSP technology to reap the flexibility and capacity.
“From the longer term research perspective orthogonal frequency-division multiplexing (OFDM) in its various guises increasingly seems to be the way people will see all this capability being harnessed,” he says
“40Gbps DPSK and DQPSK will live on and, in fact, grow very nicely,” says Neslusan. “I also learnt that at 40 and 100Gbps, the coherent detection offerings will not necessarily be adopted across the board; there is significant life for direct detection at both 40 and 100Gbps, specifically in the metro.”
“The most vocal component companies appear to have a long way to go to reign-in to a 100Gbps product; the claims seem to be significantly beyond the capability to commercialise in the near-term,” says Oclaro’s Clarke. “There is also a growing consensus that vertical integration will be increasingly valued going forward.”
“The biggest thing was the general sense that bandwidth on the fibre is no longer plentiful,” says Ovum’s Liu, highlighting how Cedric Lam of Google referred to it and Glenn Wellbrock from Verizon also made reference to it in the context of wanting a move to “gridless” - no longer being bound by the ITU’s rigid wavelengths used for DWDM - for high speed transmission.
Surprises
JDS Uniphase and RED-C stress that technology tends to move incrementally and as such did not see any ‘quantum leaps’. One interesting observation, according to JDS Uniphase, was the reduction in OFDM papers in 2010 compared to 2009. But it warns not to read too much into it: “Two points don’t make a trend,” the company said.
Eve Griliches, managing partner at ACG Research was surprised by how OFC was “overrun” by content providers: “A new breeze blowing in town was how it felt,” she says.
“It is noticeable that the direction for the industry is increasingly being set by web service companies such as Google and Facebook, says CIP’s Smith.“These people are now the drivers of the capacity demand and it is interesting that their requirements and vision does not quite align with that of the long haul operators that traditionally set the tone at OFC.”
Reasons for optimism?
“There was plenty,” says RED-C’s Menashe. 100Gbps and coherent detection has injected a lot of energy into the industry, he says. Combined with the expected growth in access networks, this should lead to healthy growth for several years.
LightCounting is encouraged that the industry is moving to more complex modulation and detection schemes despite the near- and even medium-term economic case for 100Gbps components is suspect, says Schube. Such developments define a technology roadmap for at least another decade of network capacity doubling every 18 months.
“We heard again and again, that the internet service providers such as Google are still looking for solutions for their future bandwidth demand,” says u2t’s Umbach.
“The component vendor landscape for 40G and 100G is turning out to be just like that at the beginning of 10G – that is to say, very overcrowded”
Scott Schube, LightCounting
JDS Uniphase argues that from the attendance of mainly larger stable suppliers it is clear that the ecosystem has fully evolved from the “excessive speculation by outside forces” which caused and broke the bubble. The current situation doesn’t mean that the industry is “a full picture of health or profitability”, it said, but predictability and stability may be returning.
For CIP, the show was as busy as previous OFCs but its serious leads were up by 30%. Interest was much more focussed on custom designs and volumes for solid commercial applications rather than one-off blue sky projects, it said.
“This year marked the en masse return of the "optical gadget" suppliers at OFC, specifically from China,” says Neslusan. “I consider this a good sign.”
Did you attend OFC? Is there something noteworthy that you'd like to highlight?
Mobile broadband: congestion is inevitable
The table is taken from a recent report by Peter Rysavy of Rysavy Research, entitled Mobile Broadband Capacity Constraints And the Need for Optimization.
The report looks at the huge growth in mobile broadband services and the resulting congestion. The report includes a nice model showing how only a few intensive users can consume much of a cell's capacity. The report also discusses how operators must continue to add wireless capacity while being a lot smarter in the bandwidth consumed by applications.
To see a copy of the report, click here
|
Application |
Recommended Bandwidth |
|
Mobile voice call |
6 kbps to 12 kbps |
|
Text-based e-mail |
10 to 20 kbps |
|
Low-quality music stream |
28 kbps |
|
Medium-quality music stream |
128 kbps |
|
High-quality music stream |
300 kbps |
|
Video conferencing |
384 kbps to 3 Mbps |
|
Entry-level, high-speed Internet |
1 Mbps |
|
Minimum speed for responsive Web browsing |
1 Mbps |
|
Internet streaming video |
1 to 2 Mbps |
|
Telecommuting |
1 to 5 Mbps |
|
Gaming |
1 to 10 Mbps |
|
Enterprise applications |
1 to 10 Mbps |
|
Standard definition TV |
2 Mbps |
|
Distance learning |
3 Mbps |
|
Basic, high-speed Internet |
5 Mbps |
|
High-Definition TV |
7.5 to 9 Mbps |
|
Multimedia Web interaction |
10 Mbps |
|
Enhanced, high-speed Internet |
10 to 50 Mbps, 100 Mbps emerging |