The coherent digital signal processor (C-DSP) is the cornerstone technology for optical network transmission, with every new generation of the C-DSP advancing its performance. The latest 800G C-DSP will enable transmission of 800G some hundreds of kilometers, and 200–400G over many thousands of kilometers. The 800G C-DSP will incorporate highly advanced performance enhancing algorithms including “State of Polarization” and “Faster than Nyquist (FTN).”
In addition, advances in transmission, optical line amplification, and reconfigurable optical add-drop multiplexer (ROADM) switching technologies are enabling more wavelengths to be transported per optical fiber pair. Historical solutions were capable of transmitting 80 wavelengths per fiber pair. Now, the industry is preparing for “Super C-band” and C+L solutions capable of supporting up to 216 wavelengths per fiber pair – a 2.7× improvement.
Continuing advances in C-DSP technology, and advances in system transport capabilities, are paving the way for 100Tbps transmission; an industry milestone.
Today’s core network currency is 100G
Today’s core network currency is 100G. Data centers are internally connected with 100G, with core network routers and switches interconnecting via 100G ports. Optical network equipment includes 100G client ports and 100G lineside ports. Network currency facilitates seamless equipment interconnection and network-to-network interconnection, avoiding multiplexing steps and enabling volume and scale.
200G: Cost effective, ultra-long-haul transport today, with a long lifecycle
200G has also emerged as a cost effective metro, regional, and ultra-long-haul transport option. 200G supports the 100G ecosystem as 100G clients are cost effectively transported over 200G line rates. Today’s 200G is capable of transmitting over thousands of kilometers. The next generation 800G C-DSPs will improve upon that performance, enabling a long lifecycle for 200G.
Figure 1: High speed dense wavelength-division multiplexing (DWDM): global deployment and the emergence of 200G
400G: The next major network currency enabled by 800G C-DSPs
800G C-DSPs, with programmable modulation formats, will support multiple capacity-system reach scenarios. At the 800G speed, 100km reaches will be achievable. At 400G, over 1,000km reaches will be achievable. The first generation 400G solutions were capable of 100km reaches, limiting the application space to metro distances. With the 800G C-DSP, the 400G speed application space will broaden to include ultra-long-haul and subsea – catalyzing high-volume market growth. 800G C-DSP will be a key technology in transitioning the entire telecom, datacom, and digital economy ecosystem to a 400G network currency.
Two sub-markets: high performance, and space-power optimized
The advanced optical transmission market is divided into two distinct sub-markets:
a space and power optimized market for shorter reach applications.
a high-performance market that maximizes system reach and spectral efficiency.
In metro connectivity applications, customer requirements focus on minimizing both network equipment form factor and power consumption. Due to the relaxed reach requirements, many functions and algorithms needed to mitigate long-distance fiber impairment are no longer needed in the metro environment. An optimized and reduced function set allows for the development of a more compact and less power-hungry DSP.
The high-performance DSP is designed around maximizing system reach, and includes all the industry’s leading algorithms to mitigate fiber impairments. Customers have a clear choice based on application needs.
800G C-DSP, technology foundations
The 800G C-DSPs are underpinned by the latest generation of silicon, 7nm CMOS. With each advance of silicon, more calculations can be performed per second, enabling highly sophisticated algorithms to be executed, improving performance.
Advanced “secret sauce” algorithms enable maximum performance
State of polarization mitigation reduces potential outside plant bit-error rates
The outside plant environment is harsh, with many hazards for optoelectronics. The fiber plant is subject to dramatic changes in temperature over time. Vibrations can occur, in some cases dramatically, and the fiber can be subject to tension stress. Changes in the external environment stress the fiber plant, and can lead to errors in transmission. Many of the environmental conditions are known and can be countered with corrective algorithms. The “State of Polarization” algorithm is designed to recognize, counter, and mitigate negative environmental conditions, reducing the potential bit error rate in optical transmission.
FTN spectral shaping improves spectral efficiency
Fiber transmission consumes a “bell curve” of space in the available spectrum. The long tails of each wavelength can potentially interfere with the transmission of adjacent wavelengths. FTN is a sampling technique that squares off the wavelength, reducing potential interference with adjacent wavelengths. The squared off signals can then be moved closer together with no interference, allowing more wavelengths to be transmitted in the available spectrum. FTN improves the overall system’s spectral efficiency, raising capacity transmitted per fiber pair.
Further capacity expansion enabled by Super C and C+L bands
The long-standing historical standard has been 80 wavelengths transmitted in the C-band only per fiber pair. The next steps in the optical transmission evolution will be utilizing the Super C-band and the L-band. There are 120 wavelengths in the ultra-wide C-band and, with C+L, there are an additional 96 wavelengths in the L-band. The evolution from 80 wavelengths up to 216 wavelengths transmitted per fiber pair is a major step up in system bandwidth – a 2.7× improvement.
Figure 2: Raising system capacity by utilizing all viable bands of spectrum
Source: The International Telecommunications Union (https://www.itu.int/rec/T-REC-G.694.1-201202-I/en)
Both Super-C and C+L require the development of technology ecosystems to support the enhanced spectrum range. Transmission lasers, amplifiers, and switching elements within ROADMs all require an upgrade to support the expanded spectrum. Super C will require the tunable lasers and modulators to operate in the greater Super C-band range. Laser output power will need to be flat over the extended wavelength range. The Super C-band amplifiers also need to work over the Super C range, with minimal gain variation. New doping elements have been considered to operationalize amplifier gain over the Super C range. Liquid crystal on silicon (LCOS) is the volume switching technology deployed in ROADMs today. To achieve Super C switching, the resolution must be improved to adjust and control the more numerous wavelengths precisely.
Super L-band is the ultra-wide version of regular L-band, and is able to provide more than 96 channels based on 50GHz standard channel spacing. To date, L-band has been niche, and has not achieved the status of universally adopted, practical technology. Super C and Super C+L are both in the classic technology/economic ecosystem market development catch-22: high volumes are required to enable cost effective systems, but CSPs will not invest until they see very attractive economics.
The catalyst for igniting this market will most likely be CSPs, enterprises, or others that are fiber constrained. Organizations that have only a fiber pair on a route will be highly motivated to maximize the transmission bandwidth on their scarce fiber pairs. Those organizations will endure low volume, early market equipment costs that will be less expensive in aggregate than procuring additional fiber pairs. Market momentum will build from there, enabling the Super C and the C+L ecosystems to potentially develop into volume markets.
The industry focus is higher speed per wavelength. By utilizing all of the economically viable spectrum, network operators can deploy more wavelengths per fiber pair, leading to lower overall cost per bit systems.