A survey of some of the key developments at the OFC 2015 show held recently in Los Angeles.
Part 1: Line-side component and module developments
- Several vendors announced CFP2 analogue coherent optics
- 5x7-inch coherent MSAs: from 40 Gig submarine and ultra-long haul to 400 Gig metro
- Dual micro-ITLAs, dual modulators and dual ICRs as vendors prepare for 400 Gig
- WDM-PON demonstration from ADVA Optical Networking and Oclaro
- More compact and modular ROADM building blocks
Coherent optics within a CFP2
Integrating line-side coherent optics into ever smaller pluggable modules promises higher-capacity line cards and transport platforms. Until now, the main pluggable module for coherent optical transmission has been the CFP but at OFC several optical module companies announced coherent optics that fit within the CFP2 module, dubbed CFP2 analogue coherent optics (CFP2-ACO).
Oclaro, Finisar, Fujitsu Optical Components and JDSU all announced CFP2-ACO designs, capable of 100 Gigabit-per-second (Gbps) line rates using polarisation-multiplexing, quadrature phase-shift keying (PM-QPSK) and 200 Gbps transmission using polarisation-multiplexing, 16-quadrature amplitude modulation (PM-16-QAM).
Unlike the CFP, the CFP2-ACO module houses the photonics for coherent transmission; the accompanying coherent DSP-ASIC resides on the line card. The CFP2’s 12W power consumption is insufficient to house the combined power consumption of the optics and current DSP-ASIC designs.
With the advent of the CFP2-ACO, five or even six modules can be fitted on a line card. “With five CFP2s, if you do 100 Gigabit, you have a 500 Gigabit line card, but if you can do 200 Gigabit using 16-QAM, you have a one terabit line card,” says Robert Blum, director of strategic marketing at Oclaro.
Such line cards can be used not just for metro and regional networks but for the emerging data centre interconnect market, says Blum. Using line-side pluggables also allows operators to add capacity as required.
Oclaro says its CFP2-ACO module has been shown to work with seven different DSP-ASICs; five developed by the system vendors and two merchant chips, from ClariPhy and NEL.
Oclaro uses a single high-output power narrow line-width laser for its CFP2-ACO. The bulk of the laser’s light is used for the transmitter path but some of the light is split off and used for the local oscillator in the receive path. This saves the cost of using a separate, second laser but requires that the transmit and receive paths operate on a common wavelength.
In contrast, Finisar uses two lasers for its CFP2-ACO: one for the transmit path and one for the local oscillator source. This allows independent transmit and receive wavelengths, and uses all the laser’s output power for transmission. Rafik Ward, Finisar’s vice president of marketing says the company has invested significantly to develop its CFP2-ACO, and using it own in-house components. Finisar acquired indium phosphide specialist u2t Photonics in 2014 specifically to address the CFP2-ACO design.
At OFC, fabless chip maker ClariPhy announced a CFP2-ACO reference design card. The design uses the company’s flagship CL20010 DSP-ASIC with a CFP2 cage into which various vendors’ CFP2-ACO modules can be inserted. The CL20010 DSP supports 100 Gbps and 200 Gbps data rates.
“Every major CFP2 module maker is sampling [a CFP2-ACO],” says Paul Voois, co-founder and chief strategy officer at ClariPhy. Having coherent optics integrated into a CFP2 is a real game-changer, he says. Not only will the CFP2-ACO enable one terabit line cards, but the associated miniaturisation of the optics will lower the cost of coherent transmission.
“The DSP’s cost will decline [with volumes] and so will the optics which account for two thirds of the transponder cost,” says Voois. Having a CFP2-ACO multi-source agreement (MSA) also promotes interoperability, further spurring the CFP2-ACO’s adoption, he says.
NeoPhotonics announced a micro integrated coherent receiver (micro-ICR) for the CFP2-ACO. NeoPhotonics all but confirmed it will also supply a CFP2-ACO module. “That would be a logical assumption given that we have all the pieces,” says Ferris Lipscomb, vice president of marketing at NeoPhotonics.
5x7-inch MSAs: 40 to 400 Gig
Work continues to advance the line-side reach and line-speed capabilities of the fixed 5x7-inch MSA module.
Acacia Communications announced a 5x7-inch coherent transponder that supports two carriers, each capable of carrying 100, 150 or 200 Gigabit of data. The Acacia design uses two of the company’s silicon photonics chips, one for each carrier, coupled with Acacia’s DSP-ASIC.
Finisar announced two 5x7 inch MSAs: one capable of 100 Gigabit and 200 Gigabit and one tailored for submarine and ultra long-haul applications using 40 Gig or 50 Gig binary phase-shift keying (PM-BPSK).
Finisar claims it offers the industry’s broadest 200 Gigabit optical module portfolio with its 5x7 inch MSA and its CFP2-ACO. It demonstrated its 5x7-inch MSA also working with its CFP2-ACO at OFC. For the demonstration, Finisar used its CFP2-ACO module plugged into ClariPhy’s reference design.
Micro-ITLAs, modulators and micro-ICRs go parallel
Oclaro announced a dual micro-ITLA suited for two-carrier signals for a 400 Gig super-channel, with each carrier using PM-16-QAM.
“People are designing discrete line cards using micro-ITLAs, lithium niobate modulators and coherent receivers for 400 Gig, for example, and they need two lasers, one for each channel,” says Oclaro’s Blum. This is the main application Oclaro is seeing for the design, but another use of the dual micro-ITLA is for networks where the receive wavelength is different to the transmitter one. “For that, you need a local oscillator that you tune independently,” says Blum.
Avago announced a micro-ITLA device using its external cavity laser that has a line-width less than 100kHz. The micro-ITLA is suited for 100 Gig PM-QPSK and 200 Gig 16-QAM modulation formats and supports a flex-grid or gridless architecture.
Tunable SFP+
Oclaro announced a second-generation tunable SFP that has a power consumption below 1.5W, meeting the SFP MSA. The tunable SFP also operates over an extended temperature range of up to 85oC, but here the power consumption rises to 1.8W.
“We see a lot of applications that need these higher temperatures: racks running hot, WDM-PON and wireless front-hauling,” says Blum. Wireless fronthaul typically uses grey optics to carry the radio-head traffic sent to the wireless baseband unit. But operators are looking to WDM technology as a way to aggregate traffic and this is where the extended temperature tunable SFP+ can play a role, says Blum.
WDM-PON demonstration
ADVA Optical Networking and Oclaro demonstrated a WDM-PON prototype at OFC. WDM-PON has been spoken of for over a decade as the ultimate optical access technology, delivering dedicated wavelengths to premises. More recently, WDM-PON has been deployed to deliver business services and is being viewed for mobile backhaul and fronthaul applications.
The ADVA-Oclaro WDM-PON demonstration is a 40-wavelength system using the C- and L-bands. The system’s 10 Gigabit wavelengths are implemented using tunable SFP+ modules at the customer’s site.
The difference between Oclaro’s second-generation tunable SFP+ and the WDM-PON demonstration is that the latter module does not use a wavelength locker. Instead, a centralised wavelength controller is used to monitor all 40 channels and sends information back to the customer premise equipment via the L-band if a particular wavelength has drifted and needs adjustment. “We can get away with a very low-cost tunable laser in the customer premises [using this approach],” says Blum.
ROADM building blocks
JDSU showcased its latest ROADM line cards at OFC. These included its second-generation twin 1x20 wavelength-selective switch (WSS), part of its TrueFlex Super Transport blade, and its TrueFlex Multicast Switch blade that features a twin 4x16 multicast switch and a 4+4 array of amplifiers.
JDSU’s first-generation twin 1x20 WSS required more than two slots in a chassis; two slots for the twin WSS and another for amplification and optical channel monitoring. JDSU can now fit all the functions on one blade with its latest design.
The 4x16 multicast switch supports a four-degree (four directions) ROADM and 16 drop or add ports. The twin multicast switch design is used for multiplexing and demultiplexing of wavelengths. “This size multicast switch needs an amplifier on each of those four ports,” says Brandon Collings, CTO for communications and commercial optical products at JDSU. The 4+4 array of amplifiers is for the multicast switch multiplexing and the demultiplexing, “four amps on the mux side of the multicast switch and four amps for the demux side of the multicast switch”, says Collings.
NeoPhotonics announced a modular 4x16 multicast switch which it claims does not need drop amplifiers.
Being modular, operators can grow their systems based on demand, avoiding up-front costs and having to predict the ultimate size of the ROADM node. For example by adding multicast switches they can go from 4x16, 8x16, 12x16 to a full 16x16 switch configuration. “Carriers do not like to have to plan in advance, and they like to be future-proofed,” says Lipscomb.
The NeoPhotonics multicast switch uses planar lightwave circuit (PLC) technology and has a broadcast-and-select architecture. As such, the architecture uses optical splitters which inevitably introduce signal loss. By concentrating on reducing switch loss and by increasing the sensitivity of the integrated coherent receiver, NeoPhotonics claims it can do away with the drop amplifiers for metro networks and even for certain long-haul routes. This can save up to a $1,000 a switch, says Lipscomb.
NeoPhotonics’ multicast switch has already been designed on a line card and introduced into a customer’s platform. It is now undergoing qualification before being made generally available.
ROADM status
“This type of stuff [advanced WSSes and multicast switches for ROADMs] is what Verizon has been pushing for all these years,” says JDSU’s Collings. “These developments have been completed because operators like Verizon are getting serious.” Earlier this year, Verizon selected Ciena and Cisco Systems as the equipment suppliers for its large metro contract.
Some analysts argue that it is largely Verizon promoting advanced ROADM usage and that the rest of the industry is less keen. Collings points out that JDSU, being a blade supplier and not a system vendor, is one customer layer removed from the operators. But he argues that other operators besides Verizon also want to deploy advanced ROADM technology but that two milestones must be overcome first.
“People are waiting to see the technology mature and Verizon really do it,” he says. “[Their attitude is:] Let Verizon run headlong into that, and let’s see how they fare before we invest.” Collings says that until now, ROADM hardware has not been sufficiently mature: “Even Verizon has had to wait to start deploying this stuff.”
The second milestone is having a control plane to manage the systems’ flexibility and dynamic nature. This is where the system vendors have focused their efforts in the past year, convincing operators that the hardware and the control plane are up and running, he says.
“There is lots of interest [in advanced ROADMs] from a variety of carriers globally,” says Collings. “But they have been waiting for these two shoes to drop.”
For Part 2, click here