XXXX XXXX JOHN WILLIAMSON WDM
targets and still have a healthy supply chain to support growth, new efficiencies - especially in assembly and test - are called for that have not been emphasised in coherent transponders before.” Can 400G ZR reach beyond 120 km? Lipscomb observes that to extend the reach significantly, some improvements in performance are needed, especially in TX Optical Signal Noise Ratio (OSNR) and link OSNR; both would require a dedicated optical amplifier in the transmitter. He adds that a high TX OSNR also implies that a low loss coherent modulator should be used. He also reckons that another important factor is that a higher FEC coding gain would be required in a 7nm DSP to cope with longer distance metro optical networks. “However, when the FEC coding gain increases, the total power consumption could be increased proportionally,” he says. “This could result in a small form-factor pluggable module not being able to handle the required power consumption, and therefore a larger form factor pluggable module might be needed.” WAVES OF THE FUTURE What else might be coming down the WDM turnpike? Clarkson sees more of the technology migrating to the network edge to accommodate bandwidth- hungry services such as 5G and higher speed fixed access. “So smaller DWDM boxes and termination points for one or two wavelengths at the edge of the network become more important,” he reasons. Meantime, Gill believes that the technology will continue to extend its frequency spectrum repertoire. “Typically DWDM systems today operate in the C-band or extended C-band. But there are operators who are deploying systems that can operate with different wavelengths of light in the L-band,” he concludes. “When you do that you can potentially double the capacity of a fibre.”
of operator and service provider have different requirements and expectations of WDM. Jay Gill, director of marketing at Infinera, observes that a datacentre operator with three neutral colocation facilities in a city might require just a simple short-haul DWDM network to link those sites. “When you get to long haul networks, the longer you go the more the premium there is on fine-tuning technology for best performance,” he states. “And that goes to extremes in the subsea world.” DENSE MATTER In the DWDM arena there is presently much discussion about a move to 400G systems. Although the term 400G DWDM encompasses set-ups that use two 200G wavelengths, single carrier 400G systems are beginning to be investigated. In December 2018, for example, Acacia Communications announced the first 400G single carrier DWDM transmission on the 6,600 km Marea submarine cable between Virginia Beach, USA and Bilbao, Spain. The field demonstration was conducted by Acacia in collaboration with Microsoft and Facebook. As might be expected, 400G DWDM has its challenges. “The technology is very expensive. The optics are not the cheapest around,” suggests Clarkson. Reshef agrees. “Except for some tier-1 carriers of content service providers, I do not see mass deployment of 400G before the end of 2020, or beginning of 2021,” he says. “Until then the cost of both DSP and optics will be high. The technology will mature toward the end of 2020 and only then, will there be practical, low cost, low power modules for 400G DWDM.” DISAGGREGATION: BREAKING UP HARD TO DO? Disaggregation of the elements in WDM transport system in the form of Open Line Systems (OLS) is another hot topic right now. The OLS promise is that network builders and operators could put together interoperable, best-in-class functional transport network blocks sourced from different vendors. By doing so they would reap benefits in the form of: less vendor lock-in; potentially lower CapEx and OpEx; and improved access to leading edge technology. But there may be trade-offs attached to substituting OLS for single-source solutions. As remarked by Gill, there’s the argument that if transponders and line systems are supplied by the same vendor, then system performance could be optimised, as against the possibility of “lowest common denominator” performance achievable in multi-vendor networks. In the case of OLS, there’s also the issue of performance monitoring and fault isolation. “Network operators and their suppliers
So, you could end up here looking at some of the CWDM bands, maybe in the 1430 to 1450 range
ANTHONY CLARKSON EMEA SP SALES MANAGER, PROLABS
need to make sure that line systems and transponders operate nicely together,” he contends. “That’s not an insurmountable problem, and many operators have successfully implemented open line systems, but it does require a little more thinking through.” In this context, while some operators will countenance the use of proprietary management systems to control network elements supplied by different vendors, others are waiting for an overarching solution to manage and provision multi-vendor networks. “There’s a lot of interest in standardising open interfaces to all the components so that a higher layer SDN controller and provisioning system can look across all of those components,” explains Gill. “That’s achievable today with pre-standard interfaces, but full standardisation is still a work in progress. 400G ZR: COHERENT THOUGHTS According to Lipscomb, key technology considerations for the future development of 400G ZR include low power consumption and compact coherent DSP and optics. Whether the DSP can achieve the promised low power consumption and transmission performance is an uncertain factor. “We also see a major challenge to be ramping cost effectively,” he says. “While there is no fundamentally new technology involved, to meet the cost
While there is no fundamentally new technology involved, to meet the cost targets and still have a healthy supply chain to support growth, new efficiencies - especially in assembly and test - are called
for that have not been emphasised in coherent transponders before
FERRIS LIPSCOMB VP MARKETING, NEOPHOTONICS
www.opticalconnectionsnews.com
25
ISSUE 16 | Q1 2019
Made with FlippingBook - Online magazine maker