Xtera demonstrates 40 Terabit using Raman amplification

Feature: 100 Gig and Beyond. Part 2

• Xtera's Raman amplification boosts capacity and reach
• 40 Terabit optical transmission over 1,500km in Verizon trial
• 64 Terabit over 1,500km in 2015 using a Raman module operating over 100nm of spectrum  

Herve Fevrier

Optical transport equipment makers continue to research techniques to increase the data carried long distances over a fibre without sacrificing reach. The techniques include signal processing of the transmit signal to cram data-carrying channels closer in the C-band, advanced soft-decision forward error correction (SD-FEC) and receiver signal processing to counter transmission impairments.

 

System vendor Xtera is using all these techniques as part of its Nu-Wave Optima system but also uses Raman amplification to extend capacity and reach.

Raman amplification can more than treble the available fibre spectrum used for transmission while doubling the reach, claims Xtera. In a trial with US operator Verizon, Xtera demonstrated its Nu-Wave Optima platform using Raman amplification to send 15 Terabit over 4,500km, and 40 Terabit over 1,500km.

 

"We offer capacity and reach using a technology - Raman amplification - that we have been pioneering and working on for 15 years," says Herve Fevrier, executive vice president and chief strategy officer at Xtera.

 

EDFA and Raman

 

An Erbium-doped fibre amplifier (EDFA), long established as the amplifier of choice for wavelength division multiplexing (WDM) deployments, has superior power efficiency compared to Raman. Power efficiency refers to the optical pump power needed to pump an amplifier to achieve a certain gain, explains Fevrier. One way to improve the Raman's power efficiency is to use a high-efficiency fibre. However, unless an operator is deploying new fibre, the existing less-efficient fibre plant will be used as the Raman gain medium. Yet while the Raman amplifier has a lower power efficiency compared to an EDFA, it does deliver significant benefits at the system level.

 

An EDFA spans the C-band, the 35nm wide spectrum covering 1530-1565nm. Raman can amplify a much wider 100nm window, covering more than the C-Band and L-band (1570-1605nm) combined. This enables the transmission of many more wavelengths across a fibre, delivering the threefold increase in capacity Xtera mentions. And while operators can also deploy EDFAs in the L-Band, a different design is needed compared to the C-band EDFA such that an operator must stock two EDFAs types.

 

An EDFA performs lumped amplification, restoring the signal at distinct points in the network 80km apart. In contrast, Raman boosts the signal as it travels down the fibre. The smoother amplification of Raman also means the power needed per channel is lower, says Fevrier, increasing the margin before non-linear effects are triggered due to the optical signal's power.

 

The distributed amplification profile of Raman (blue) compared to an EDFA's point amplification. Source: Xtera

 

One way vendors are improving the amplification for 100 Gigabit and greater deployments is to use a hybrid EDFA/ Raman design. This benefits the amplifier's power efficiency and the overall transmission reach but the spectrum width is still dictated by Erbium to around 35nm. "And Raman only helps you have spans which are a bit longer," says Fevrier.

"100-Gigabit-plus will need Raman and people are using hybrids," he says. Instead, operators should consider deploying all-Raman for their future high-speed networks. "Go for it in terms of line system and then you will triple capacity." says Fevrier.  

 

 

Verizon trial

 

Xtera conducted several trials with Verizon using the Nu-Wave Optima with the Raman amplification operating over a 61nm window.

 

The first test sent 15 Terabit over 4,500km using 150 channels, each 100 Gig polarisation-multiplexed, quadrature phase-shift keying (PM-QPSK) modulation at 50GHz channel spacings. The second trial demonstrated 400 Gigabit super-channels comprising four 100 Gig PM-QPSK signals spaced 33GHz apart. The resulting capacity was 20 Terabit that  achieved a 3,000km reach. The final trial used a 400 Gig super-carrier based on two 200 Gig polarisation multiplexed, 16 quadrature amplitude modulation (PM-16-QAM) signals. Using 50GHz spacing a total of 30 Terabit was sent over 2,000km. Moving to a 37.5GHz channel spacing, capacity rose to 40 Terabits and a distance of 1,500km was achieved.

 

"People [at Verizon] were amazed that our system was linear," says Fevrier. "The non-linear penalty is extremely small."

 

Xtera expects that with a full 100nm spectral window, 48 Terabits could be sent over 2,000km and 64 Terabits over 1,500km. "To go to 100nm we need to be cost effective such that the cost of an 100nm system would match the 62nm one is today," says Fevrier.

 

The vendor's message to operators planning 100-Gig-plus deployments is that spectrum should be a key part of their considerations. "When you think of the investment your are doing, turning up a new system, I think it is really time to think of spectrum."

 

Meanwhile, Xtera is working on programable cards that will support the various transmission options. Xtera will offer a 100nm amplifier module this year that extends its system capacity to 24 Terabit (240, 100 Gig channels). Also planned this year is super-channel PM-QPSK implementation that will extend transmissions to 32 Terabit using the 100nm amplifier module. In 2015 Xtera will offer PM-16-QAM that will deliver the 48 Terabit over 2,000km and the 64 Terabit over 1,500km.

 

For Part 1, click here