contact button

Meeting the 4 Challenges of C+L

headshot of Paul Momtahan

June 20, 2023
By Paul Momtahan
Director, Solutions Marketing

DWDM networks have traditionally used only C-band fiber spectrum. The C-band was preferred over the L-band for its lower dispersion and the availability of lower-noise EDFA amplification. The Japanese market was the exception to this as G.653 fiber was widely deployed, which made the L-band favorable for DWDM. Over time, the amount of C-band spectrum that vendors have supported in DWDM systems has evolved from 3.2 THz to 4.8 THz, the extended C-band. The L-band is also able to leverage EDFA amplification technology and provides the option to double the amount of spectrum, with an additional 4.8 THz in the extended L-band, as shown in Figure 1. As optical engine spectral efficiency approaches the Shannon limit and spectral efficiency gains become ever harder to achieve, increasing a fiber’s total capacity by utilizing C+L spectrum can provide a practical and cost-effective option for network operators.

C+L doubles spectrum to 9.6 THzFigure 1: C+L doubles spectrum to 9.6 THz

L-band Challenges 1, 2, and 3

SRS tilt shifts power from the C-band to the L-bandFigure 2: SRS tilt shifts power from the C-band to the L-band

However, extending into the L-band also creates several challenges. Many of these challenges stem from an optical effect called stimulated Raman scattering (SRS). SRS can be a highly desirable effect leveraged by Raman amplifiers to provide extended reach with lower noise. However, in C+L optical networks, this same effect can cause wavelengths in the C-band to amplify wavelengths in the L-band. Moreover, this effect causes both the C-band and the L-band to tilt, with power going from the C-band to the L-band, as shown in Figure 2. This creates challenges in the following areas:

  1. C+L recovery speed: Recovery times from a failure, such as a fiber cut or lost channels after transponder failure, can be long, around 10 minutes, without careful mitigation.
  2. C+L provisioning speed and complexity: The number of channels in the C- and L-bands may need to be carefully managed. Slow adaptation to changes can lead to provisioning delays. Complex loading rules and topology restrictions can also make C+L networks operationally challenging.
  3. C+L supported topologies: SRS can also impact which topologies can be supported. Point-to-point topologies are relatively straightforward. Ring, linear chain, and mesh topologies are more complex, as wavelengths in the C- and L-bands are added or removed from one span to the next.

L-band Challenge 4: Cost

The other key challenge of C+L networks relates to cost. C+L networks typically require doubling the number of components such as amplifiers and wavelength-selective switches (WSSs). Upgrading from a C-band network to a C+L network may also require a complete replacement of the C-band line system. Incurring the full cost of a complete C+L system when the L-band will only be required at some point in the future is also less than ideal for many network operators.

Infinera’s C+L Solution Addresses These Challenges

Infinera provides a comprehensive C+L solution to address these four challenges that includes the FlexILS optical line system and L-band versions of the ICE6-enabled CHM6 Xponder sleds for the GX G42.

Fast and Simplified C+L Provisioning

Infinera’s FlexILS supports C-band and L-band amplified spontaneous emission (ASE) idler cards. These ASE idlers, together with the local WSS, use ASE noise in 12.5 GHz slices to fill any unused spectrum, ensuring that the entire band is fully loaded at all times. ASE idlers therefore eliminate loading rules: any wavelength can be provisioned regardless of existing spectrum loading. Provisioning times are also greatly reduced. In addition, ASE idlers provide the ability to verify at system turn-up, when the network might be lightly loaded, that the network can in the future be fully loaded with wavelengths.

A second enabler for fast and simplified C+L provisioning is the ADAPT link control software that runs on each FlexILS. It provides resource and topology discovery, sets the gain of amplifiers, and optimizes the transmit power level of each wavelength. In terms of C+L-specific features, it provides fast automated C+L tilt control for the EDFA and Raman amplifiers in each span. ADAPT also includes a loading manager that controls the addition and removal of large traffic demands, preventing users from applying network or configuration changes that could place the network at risk. And finally, ADAPT provides automated ASE management when used with the ASE idlers.

Fast C+L Recovery

FlexILS’s C- and L-band ASE idlers, together with ADAPT’s automated ASE management and fast transit suppression, enable recovery times of less than 10 seconds, even with complex topologies such as mesh ROADM and major events such as fiber cuts. In addition, ADAPT’s Loading Manager function aids network availability by preventing users from applying network or configuration changes that could put the network at risk. Furthermore, the CHM6 and CHM6L sleds support a hybrid Y-cable client protection option with one wavelength in the L-band and the other wavelength in the C-band.

C+L Topology Flexibility

FlexILS supports C+L networks with a wide range of topologies, including point-to-point, linear chain, ring, and mesh. It even supports mesh networks with complex traffic patterns. And while simpler topologies can be deployed without ASE idlers, ASE idlers leveraging the automated ASE management within ADAPT provide many benefits, including simplified provisioning and fast failure recovery.

Cost-effective C+L

With L-band-expandable C-band ROADM and amplifier cards, FlexILS provides the option to initially deploy a C-band-only system. Then, as shown in Figure 3, when the extra capacity is required, an in-service upgrade to C+L can be performed. Furthermore, these upgrades need only be performed on those spans comprising routes that require the L-band. This avoids what for most network operators would be the unnecessary cost of having to prematurely upgrade the entire network. In addition, ICE6’s superior wavelength capacity-reach, enabled by baud rates of up to 100 Gbaud, and advanced features, including long-codeword probabilistic constellation shaping (LC-PCS)  and second-generation Nyquist subcarriers, minimize the coherent interface cost per bit in both the C- and L-bands.

FlexILS supports cost-effective in-service upgrades from C to C+LFigure 3: FLexILS supports cost-effective in-service upgrades from C to C+L

Telstra InfraCo: 61.3 Tb/s over 1,240 km

A recent trial with Telstra InfraCo provides a great example of what can be achieved with this C+L solution. In this trial, a simulated network demonstrated 61.3 Tb/s with 700 Gb/s ICE6-enabled wavelengths and a C+L FlexILS line system over the equivalent of 1,240 route kilometers between Melbourne and Sydney.

So, to conclude, as bandwidth demands continue unabated and transponder spectral efficiency gains become incremental, expanding the total amount of spectrum on the fiber by adding L-band support becomes an increasingly attractive option, especially in fiber-constrained environments, as is the case for many long-haul networks. With Infinera’s FlexILS and ICE6-enabled L-band Xponders, network operators can deploy a C+L solution that enables up to 80+ Tb/s per fiber pair with fast recovery over a wide variety of topologies, while also benefiting from automation and cost-effective in-service upgrades when and where the additional L-band-enabled capacity is required. For more information on this important topic, download the new Infinera application note on Infinera’s C+L Solution.