KAREN LIU MID-REACH OPTICS
STUCK IN THE MIDDLE?
SPEEDING UP MID-REACH OPTICAL NETWORKS
There is a technology gap opening up in mid-reach optics, those that span reaches beyond 10km and up to approximately 100km. Until now, the lack of dedicated technology for this application has not been a problem but a coming collision between new applications and physical limitations looks to be about to change that. 5G and edge datacentres are expected to increase demand for links that span tens of kilometres just as data rates increase to speeds of 100+ Gbps where the optics become more physically challenging, writes Karen Liu , VP of sales and marketing at Lightwave Logic.
T he main original application for 40km optics is for metro telecom networks that connect adjacent central offices. A similar application in the cable network distributes data to head-ends. According to market research firm LightCounting, about a million units of merchant 1G and 10G Ethernet transceivers for 40km and 80km ship every year and the number continues to increase. Adding in other types of transceivers and captive shipments, LightCounting estimates about three million units of 40km and 80km optics ship annually. MID-REACH OPTICS STAND AT A TECHNOLOGY BORDER The history of 40km, and to a lesser extent 80km Ethernet, traditionally called ER (extended-reach) and ZR, illustrates the forces at play. This part of the optical interconnect market has long been a border zone between shorter and
at 1550nm for 10G, though about half the market has not been exactly 1550nm, but adjacent wavelengths that enable dense-wavelength division multiplexing (DWDM) sharing of fibre. Technically, solutions for 100G and above, implemented as multi-wavelength 25G, now have good reason to be in the ~1310nm CWDM camp. Increased chromatic dispersion effects at higher speeds force them to stay near the fibre minimum dispersion point. But note that the shift from 1550nm to 1310nm for 40km happened when both solutions were technically feasible, pointing to market forces as the root cause. The shorter-reach 10km products sell in much higher volumes than longer- reach products. It is cost-effective to ‘import’ components such as lasers to also service 40km. By ‘binning’ or selecting the best of the manufacturing distribution for devices designed for 10km, manufacturers can use slightly higher power transmitters and slightly
longer-reach approaches. The history of something as simple as wavelength choice between 1310nm and 1550nm anecdotally illustrates the technology border that exists at 40km. For all speeds of Ethernet optics from Gigabit up to the latest proposed IEEE standards for 100G, the 10km wavelength has been 1310nm. Until now, 80km has been non-standard but universally 1550nm, the fibre loss minimum point, was also used for long-haul links of hundreds and thousands of kilometres. At 40km, both 1310nm and 1550nm solutions are offered. The IEEE did standardise 10GBASE-ER with 1550nm but the more popular solution in the market proved to be 1310nm. The 10G technology was reused in the next generation by ganging together four data streams to implement 40G Ethernet. At 40G, both 10km and 40km were standardised on four coarse-wavelength division (CWDM) wavelengths clustered near 1310nm each carrying 10G. Meanwhile 80km remained
26
| ISSUE 16 | Q1 2019
www.opticalconnectionsnews.com
Made with FlippingBook - Online magazine maker