Excerpted from our eBook, 100 Gbps Networks See the Light. Click here to download.
Wide-scale deployment of 100 Gbps technology in long haul will happen in earnest over the next two to three years, according to many industry sources. Generally, it could take longer to get into metro networks because the economics don't make sense yet. It is often more affordable for the carrier to relay on 40 Gbps or multiple 10 Gbps links than pay for the upgrade. Plus, the prices for 10 Gbps links continue to decline.
Still, vendors are working on the metro deployment case, and one recent innovation aimed at refining that case came from ADVA Optical Networking (XETRA: ADV.DE), which recently announced an alternative to the 100 Gbps optical platforms that use-polarization dual binary modulation and coherent signal detection.
Dual-binary modulation and coherent detection are considered the basis of many future long-haul 100G systems, but they also represent somewhat newer, and thus, more expensive technology options.
The transponder announced by ADVA instead uses dual binary modulation and non-coherent detection, and relies on four 28 Gbps channels (the extra gigabits are applied to error correction), each inserted into a 50 Ghz sub-channel, to achieve 100G at distances of up to 500 km, according to Jorg-Peter Elbers, vice president of advanced technology at ADVA.
"This can be used in the data center connectivity space or in the service provider metro, where you see distances of less than 500 km, or even less than 200 km," he said. "What we identified was a need over shorter distances. Core 100 Gbps solutions are not suited for this."
Michael Kennedy, principal analyst at ACG Research, said ADVA's announcement is an interesting innovation for what right now amounts to a niche market.
"At this point, the economics of 100 Gbps only make sense in long routes like in the service provider core networks," he said. "But, some metros are moving beyond the need for two or three 10 Gbps E links. Now, 10 Gbps is actually a small pipe in some cases, and this is where this transponder might come in. There is some efficiency to be gained in deploying that in a metro environment where you are already seeing a need for seven to eight 10 Gbps links."
ADVA's development doesn't dispute the need for DP-QPSK and coherent detection in the long run.
"We do coherent as well, but people are looking right now for a lower-cost, lower-power solution for the metro," Elbers said.
In fact, coherent detection will be very important to the future of 100 Gbps in all networks.
"Coherent technology actually makes 100 Gbps cheaper than non-coherent, though for now it is more costly because it's a newer technology," Kennedy said. "With coherent, think of it as a light beam with all of the light tightly aligned, but with non-coherent, the light would be more diffused. The coherent beam would look like a pen light if you shined it at the moon; the non-coherent would be like a flash light. Coherent reduces the need for amplification and increases the distance. It reduces chromatic dispersion, so you can have wavelengths set closer together, and you would need fewer subchannels."
Rick Dodd, senior vice president of global marketing at Ciena (Nasdaq: CIEN), said the result is more capacity. "With coherent detection you can encode more bits on a given amount of spectrum," he said. "You get more capacity, perhaps 10-times more bits on the same spectrum, and it gives you a much simpler deployment model."
Many of these benefits mean that coherent detection also helps network operators to better plan and manage their routes.
"Coherent gives you some new possibilities in network architecture," Dodd said. "It makes provisioning and restoration easier. We're doing proof of concept for someone using coherent to tune a receiver to a particular channel."
Coherent detection already has been a factor in some 40 Gbps deployments, but it is expected to become dominant as long-haul 100 Gbps deployments become the norm. Another management-related technology innovation that could play a role in future 100 Gbps networks-soft-decision forward error correction- is a much newer development. Using a soft-decision algorithm allows more dynamic allocation error correction for latency-sensitive traffic. It is not yet widely available in 100 Gbps equipment, but is expected to be more common as more vendors roll out 100 Gbps gear and feature upgrades later this year.
P.S. This article was featured in our latest eBook, 100 Gbps Networks See the Light.
Note: Yesterday we posted a photo misidentified as Jorg-Peter Elbers of ADVA Optical Networks. We've corrected the oversight.