Unrepeatered (UR) transmission systems are characterized by an extremely long fiber span with no in-line electrically powered elements between the terminals and are typically deployed in submarine environments. The relative performance of UR links is constrained by four main factors: fiber properties, transponder technology, Raman amplification, and remote optically-pumped amplifier (ROPA) design and location , .
Submarine systems have enjoyed the tremendous advances in coherent transmission brought about by high DSP technology over the last few years. These techniques have brought us remarkably close to the Shannon limit. Increasing the transmission capacity further is achievable by selecting improved system parameters such as a smaller fibre attenuation coefficient and a lower nonlinear coefficient.
The capacity of ultra-long haul (ULH) submarine systems is ultimately limited by the electrical power that can be delivered to the submerged amplifiers within the repeaters. An excessive power feed voltage will cause damage to the cable and joints and so the system should be designed to maximize the power available per repeater using the available voltage.
Larger network bandwidth to meet ever increasing traffic demand requires higher transport capacity. For unrepeatered applications, several reports have already been published on both high capacity and long reach unrepeatered transmission [1-8]. However, these results often require either a mix of fiber types in the span [2, 4-7], very strong (> 5W) high-order Raman pumping [1, 4, 7], or offline-channel processing [3, 6], which are less practical in real deployments.
557-km Unrepeatered 100G Transmission with Commercial Raman DWDM System, Enhanced ROPA, and Cabled Large Aeef Ultra-Low Loss Fiber in OSP Environment
Many applications would benefit from very-long, “skinny” (low-capacity) unrepeatered systems. One such application is terrestrial routes in remote and hostile areas (tropical forest, desert…) for which the use of unrepeatered transmission alleviates the need for intermediate amplification sites (and associated operational expenses).
Unrepeatered transmission systems provide a cost-effective solution in submarine networks to communicate between coastal population centers or in terrestrial networks to connect remote areas where service access is difficult. One of the main goals of unrepeatered systems is to achieve the longest reach without any inline active elements.
Unrepeatered transmission systems provide cost effective solution for:
- communication between islands or main-lands
- communication for oil platforms (for oil and gas company)
- skipping intermediate sites in hostile areas (tropical forest, desert etc.)
One of the factors limiting the potential capacity of a long submarine cable is the need to power the optical amplifiers which compensate for fibre loss. This paper shows how a number of routes to higher capacity all result in higher power requirements and uses the Shannon formula to assess upper limits on capacity. The results suggest that a promising, and practical, route to raising the limit would be to improve the powering of the optical amplifiers.
Field Deployment of Advanced Photonic Technologies for Ultra-High Bit Rate and Ultra-Long Reach Terrestrial WDM Transmission in Brazil
The insatiable quest for higher capacity in the optical backbone networks, fueled by video, cloud and other high capacity services, applies to both mature and emerging countries worldwide and in any geographical areas as well. Installation of new, high data rate backbone networks continues apace as a key driver for economic advancement in developing regions of the world.
Given the introduction of coherent 100G systems has provided enough fiber capacity to meet data traffic growth in the near term, enhancing network efficiency will be service providers’ high priority. Adding flexibility at the optical layer is a key step to increasing network efficiency, and both spectral and spatial functionality will be considered in next generation optical networks along with advanced network management to effectively harness the new capabilities.
It is well known that Raman amplification can improve system performance, but by what amount? I recently have written a review article on Raman amplification  based on an Invited Tutorial presented at OFC 2015 . This article provides a good overview on how Raman amplification works in real systems and describes a method to calculate the nonlinear transmission penalties in coherent links. It also describes the implementation and control of Raman modules.
Raman Amplification for Ultra-Large Bandwidth and Ultra-High Bit Rate Submarine and Terrestrial Long-Haul WDM Transmission
Insatiable capacity growth and lower cost per transported bit are part of the main challenges any long-haul optical transmission infrastructure needs to cope with. The advent at the beginning of this decade of 100G channel rate, with PM-QPSK modulation format associated with digital coherent detection , offered a 10-fold capacity increase compared to networks based on 10G waves.
Transmission of 400G PM-16QAM Channels over LongHaul Distance with Commercial All-Distributed Raman Amplification System and Aged Standard SMF in Field
PM-QPSK has become the standard long-haul transmission modulation format for practical carrier networks while PM-16QAM has been considered for practical metro and regional networks [1, 2]. Since much higher SNR is required for 16QAM signals compared with QPSK signals it has not been considered so far to run 16QAM signals for long-haul distance in a real field environment with lossy and aged fiber.
In the past few years, we have seen a tremendous surge in fiber capacity enabled by 100G optical channels built upon PM-QPSK (Polarization Multiplexing Quadrature Phase Shift Keying) modulation format. The quest for extreme bandwidth in optical backbone networks, however, seems never ending.
30 Gbaud opto-electronics and Raman technologies have quickly become the new standards for terrestrial backbone networks. Today, the different subsea market segments take benefit from these new technologies as well.
This insatiable demand for more bandwidth in optical backbone networks is fueled not only by increases in the number of users, the access methods and rates, but also the increased importance of data centers, cloud and the number of services (such as mobile, social media, and video in general).
Network operators who want to have frequent low bitrate client add/drop points on a long-haul 100G DWDM route have three legitimate options. They can 1) use a second pair of fibers for a parallel 10G DWDM system, 2) share the optical spectrum by adding 10G wavelength in with the 100G wavelengths on the same system, or 3) drop a 100G channl and mux/demux the lower rate clients from it at every add/drop point. Xtera can support three options.
Until recently the main purpose of line monitoring, or supervisory, was to monitor the repeater status, detect system degradation and locate faults. As systems become more complex the submerged equipment may need, however, to be controlled as well as monitored. The Branching Unit (BU) is a good example: while electrical switching can be controlled by power-feed currents, several suppliers have used supervisory signaling as a way of addressing some of the issues that this can create.
Below the water, laid on the sea floor, numerous high-capacity submarine cables connect countries and continents. Today, optical subsea cable systems carry more than 95% of the world’s international voice and data traffic. This figure clearly shows that very little international traffic is carried by satellite, which is still a popular misconception.
Traditional networks consist of two operating planes: the management plane and the transport or data plane. In this architecture, the data plane carries the user data. It consists of various network equipment, such as interface cards, switching equipment, and the fiber plant. Network operating information is managed by the management plane consisting of the Element Management System (EMS), the Network Management System (NMS) and the Operation Support System (OSS).
For the world’s complex infrastructure of information and communications technologies, industry standards are critical to the interoperability of all networks regardless of whether they are for voice, data or video. Industry standards allow these networks to “speak the same language,” and be able to exchange information among them.
Complimentary Fiber and Active Equipment Technologies that Deliver Extended Reach for High Data Rate Transmission in Long-Haul Networks
Installation of new, high data rate backbone networks continues apace as a key for economic advancement in developing regions of the world. Providing a reliable infrastructure for the movement of large volumes of telecoms data are as important for the economic development of a country as road network is for transporting physical goods.
From its inception, Xtera’s objective was to develop a Raman-centric solution that brings together operational excellence and simplicity, as well as outstanding reach and capacity in optical transmission performances, in field conditions. Unlike competitive offerings where integrating optical Raman amplifiers with EDFAs is an afterthought, Xtera designed its optical networking platform from the ground up to combine different optical amplification flavors, ranging from simple EDFA to all-distributed Raman amplification.
Continuous capacity growth – fueled by video and cloud services – and lower cost per transported bit are part of the main challenges any long-haul optical transmission infrastructure needs to cope with. The advent at the beginning of this decade of 100G channel rate, associated with digital coherent detection, offered a 10-fold capacity increase compared to networks based on 10G waves.
Arrayed Waveguide Grating (AWG)
An arrayed waveguide grating (AWG) is a passive optical device that is constructed of an array of waveguides, each of slightly different length. With a AWG, you can take a multiwavelength input and separate the component wavelengths on to different output ports. The reverse operation can also be performed, combining several input ports on to a single output port of multiple wavelengths. An advantage of AWGs is their ability to operate bidirectionally. AWGs are used to perform wavelength multiplexing and demultiplexing, as well as wavelength add/drop operations.
Bit Error Rate/Q-Factor (BER)
Bit error rate (BER) is the measure of the transmission quality of a digital signal. It is an expression of errored bits vs. total transmitted bits, presented in a ratio. Whereas a BER performance of 10-9 (one bit in one billion is an error) is acceptable in DS1 or DS3 transmission, the expected performance for high speed optical signals is on the order of 1015. Bit error rate is a measurement integrated over a period of time, with the time interval required being longer for lower BERs. One way of making a prediction of the BER of a signal is with a Q-factor measurement.
The C-band is the “center” DWDM transmission band, occupying the 1530 to 1562nm wavelength range. All DWDM systems deployed prior to 2000 operated in the C-band. The ITU has defined channel plans for 50GHz, 100GHz, and 200GHz channel spacing. Advertised channel counts for the C-band vary from 16 channels to 96 channels. The C-Band advantages are:
- Lowest loss characteristics on SSMF fiber.
- Low susceptibility to attenuation from fiber micro-bending. EDFA amplifiers operate in the C-band window.
Chromatic Dispersion (CD)
The distortion of a signal pulse during transport due to the spreading out of the wavelengths making up the spectrum of the pulse. The refractive index of the fiber material varies with the wavelength, causing wavelengths to travel at different velocities. Since signal pulses consist of a range of wavelengths, they will spread out during transport.
A passive multiport device, typically 3 or 4 ports, where the signal entering at one port travels around the circulator and exits at the next port. In asymmetrical configurations, there is no routing of traffic between the port 3 and port 1. Due to their low loss characteristics, circulators are useful in wavelength demux and add/drop applications.
A coupler is a passive device that combines and/or splits optical signals. The power loss in the output signals depends on the number of ports. In a two port device with equal outputs, each output signal has a 3 dB loss (50% power of the input signal). Most couplers used in single mode optics operate on the principle of resonant coupling. Common technologies used in passive couplers are fused-fiber and planar waveguides.
WAVELENGTH SELECTIVE COUPLERS
Couplers can be “tuned” to operate only on specific wavelengths (or wavelength ranges). These wavelength selective couplers are useful in coupling amplifier pump lasers with the DWDM signal.
A dispersion compensation unit removes the effects of dispersion accumulated during transmission, thus repairing a signal pulse distorted by chromatic dispersion. If a signal suffers from the effects of positive dispersion during transmission, then the DCU will repair the signal using negative dispersion.
- Positive dispersion (shorter “blue” ls travel faster than longer “red” ls) for SSMF
- Dispersion value at 1550nm on SSMF = 17 ps/km*nm
DISPERSION COMPENSATION UNIT (DCU)
- Commonly utilizes Dispersion Compensating Fiber
- Negative dispersion (shorter “blue” ls travel slower than longer “red” ls) counteracts the positive dispersion of the transmission fiber… allows “catch up” of the spectral components with one another
- Large negative dispersion value … length of the DCF is much less than the transmission fiber length
Dispersion Shifted Fiber (DSF)
In an attempt to optimize long haul transport on optical fiber, DSF was developed. DSF has its zero dispersion wavelength shifted from the 1310nm wavelength to a minimal attenuation region near the 1550nm wavelength. This fiber, designated ITU-T G.653, was recognized for its ability to transport a single optical signal a great distance before regeneration. However, in DWDM transmission, signal impairments from four-wave mixing are greatest around the fiber’s zero-dispersion point. Therefore, with DSF’s zero-dispersion point falling within the C-Band, DSF fiber is not suitable for C-band DWDM transmission.
DSF makes up a small percentage of the US deployed fiber plant, and is no longer being deployed. DSF has been deployed in significant amounts in Japan, Mexico, and Italy.
Erbium Doped Fiber Amplifier (EDFA)
The power source for amplifying the signal, typically a 980nm or 1480nm laser.
ERBIUM DOPED FIBER
Single mode fiber, doped with erbium ions, acts as the gain fiber, transferring the power from the pump laser to the target wavelengths.
WAVELENGTH SELECTIVE COUPLER
Couples the pump laser wavelength to the gain fiber while filtering out any extraneous wavelengths from the laser output.
Prevents any back-reflected light from entering the amplifier.
EDFA Advantages are:
- Efficient pumping
- Minimal polarization sensitivity
High output power
- Low noise
- Low distortion and minimal crosstalk
EDFA Disadvantages are:
- Limited to C and L bands
Fiber Bragg Grating (FBG)
A fiber Bragg grating (FBG) is a piece of optical fiber that has its internal refractive index varied in such a way that it acts as a grating. In its basic operation, a FBG is constructed to reflect a single wavelength, and pass the remaining wavelengths. The reflected wavelength is determined by the period of the fiber grating.
If the pattern of the grating is periodic, a FBG can be used in wavelength mux / demux applications, as well as wavelength add / drop applications. If the grating is chirped (non-periodic), then a FBG can be used as a chromatic dispersion compensator.
Four Wave Mixing (FWM)
The interaction of adjacent channels in WDM systems produces sidebands (like harmonics), thus creating coherent crosstalk in neighboring channels. Channels mix to produce sidebands at intervals dependent on the frequencies of the interacting channels. The effect becomes greater as channel spacing is decreased. Also, as signal power increases, the effects of FWM increase. The presence of chromatic dispersion in a signal reduces the effects of FWM. Thus the effects of FWM are greatest near the zero dispersion point of the fiber.
The gain from an amplifier is not distributed evenly among all of the amplified channels. A gain flattening filter is used to achieve constant gain levels on all channels in the amplified region. The idea is to have the loss curve of the filter be a “mirror” of the gain curve of the amplifier. Therefore, the product of the amplifier gain and the gain flattening filter loss equals an amplified region with flat gain.
The effects of uneven gain are compounded for each amplified span. For example, if one wavelength has a gain imbalance of +4 dB over another channel, this imbalance will become +20 dB after five amplified spans.
This compounding effect means that the weaker signals may become indistinguishable from the noise floor. Also, over-amplified channels are vulnerable to increase non-linear effects.
An isolator is a passive device that allows light to pass through unimpeded in one direction, while blocking light in the opposite direction. An isolator is constructed with two polarizers (45o difference in orientation), separated by a Faraday rotator (rotates light polarization by 45o).
One important use for isolators is to prevent back-reflected light from reaching lasers. Another important use for isolators is to prevent light from counter propagating pump lasers from exiting the amplifier system on to the transmission fiber.
The L-band is the “long” DWDM transmission band, occupying the 1570 to 1610nm wavelength range. The L-band has comparable bandwidth to the C-band, thus comparable total capacity. The L-Band advantages are:
- EDFA technology can operate in the L-band window.
A LASER (Light Amplification by the Stimulated Emission of Radiation) produces high power, single wavelength, coherent light via stimulated emission of light.
Semiconductor Laser (General View)
Semiconductor laser diodes are constructed of p and n semiconductor layers, with the junction of these layers being the active layer where the light is produced. Also, the lasing effect is induced by placing partially reflective surfaces on the active layer. The most common laser type used in DWDM transmission is the distributed feedback (DFB) laser. A DFB laser has a grating layer next to the active layer. This grating layer enables DFB lasers to emit precision wavelengths across a narrow band.
Mach-Zehnder Interferometer (MZI)
A Mach-Zehnder interferometer is a device that splits an optical signal into two components, directs each component through its own waveguide, then recombines the two components. Based on any phase delay between the two waveguides, the two re-combined signal components will interfere with each other, creating a signal with an intensity determined by the interference. The interference of the two signal components can be either constructive or destructive, based on the delay between the waveguides as related to the wavelength of the signal. The delay can be induced either by a difference in waveguide length, or by manipulating the refractive index of one or both waveguides (usually by applying a bias voltage). A common use for Mach-Zehnder interferometer in DWDM systems is in external modulation of optical signals.
- Combines multiple optical signals onto a single optical fiber
- Typically supports channel spacing of 100GHz and 50GHz
Separates individual channels from the aggregate DWDM signal
- Thin film filters
- Fiber Bragg gratings
- Diffraction gratings
- Arrayed waveguide gratings
- Fused biconic tapered devices
- Inter-leaver devices
Non-Zero Dispersion Shifted Fiber (NZ-DSF)
After DSF, it became evident that some chromatic dispersion was needed to minimize non-linear effects, such as four wave mixing. Through new designs, λ0 was now shifted to outside the C-Band region with a decreased dispersion slope. This served to provide for dispersion values within the C-Band that were non-zero in value yet still far below those of standard single mode fiber. The NZ-DSF designation includes a group of fibers that all meet the ITU-T G.655 standard, but can vary greatly with regard to their dispersion characteristics.
First available around 1996, NZ-DSF now makes up about 60% of the US long-haul fiber plant. It is growing in popularity, and now accounts for approximately 80% of new fiber deployments in the long-haul market. (Source: derived from KMI data)
Optical Add Drop Multiplexing (OADM)
An optical add/drop multiplexer (OADM) adds or drops individual wavelengths to/from the DWDM aggregate at an in-line site, performing the add/drop function at the optical level. Before OADMs, back to back DWDM terminals were required to access individual wavelengths at an in-line site. Initial OADMs added and dropped fixed wavelengths (via filters), whereas emerging OADMs will allow selective wavelength add/drop (via software).
Optical Amplifier (OA)
Placed immediately after a transmitter to increase the strength on the signal.
IN-LINE AMPLIFIER (ILA)
Placed in-line, approximately every 80 to 100km, to amplify an attenuated signal sufficiently to reach the next ILA or terminal site. An ILA functions solely in the optical domain, performing the 1R function.
Placed immediately before a receiver to increase the strength of a signal. The preamplifier boosts the signal to a power level within the receiver’s sensitivity range.
Optical bandwidth is the total data carrying capacity of an optical fiber. It is equal to the sum of the bit rates of each of the channels. Optical bandwidth can be increased by improving DWDM systems in three areas: channel spacing, channel bit rate, and fiber bandwidth. The current benchmark for channel spacing is 50GHz. A 2X bandwidth improvement can be achieved with 25GHz spacing.
Current benchmark is 50GHz spacing. A 2X bandwidth improvement can be achieved with 25GHz spacing.
- Laser stabilization
- Mux/Demux tolerances
- Non-linear effects
- Filter technology
CHANNEL BIT RATE
Current benchmark is 10Gb/s. A 4X bandwidth improvement can be achieved with 40Gb/s channels. However, 40Gb/s will initially require 100GHz spacing, thus reducing the benefit to 2X.
- PMD mitigation
- Dispersion compensation
- High Speed SONET mux/demux
Current benchmark is C-Band Transmission. A 3X bandwidth improvement can be achieved by utilizing the “S” & “L” bands.
- Optical amplifier
- Band splitters & combiners
- Gain tilt from stimulated Raman scattering
Optical fiber used in DWDM transmission is single mode fiber composed of a silica glass core, cladding, and a plastic coating or jacket. In single mode fiber, the core is small enough to limit the transmission of the light to a single propagation mode. The core has a slightly higher refractive index than the cladding, thus the core/cladding boundary acts as a mirror. The core of single mode fiber is typically 8 or 9 microns, and the cladding extends the diameter to 125 microns. The effective core of the fiber, or mode field diameter (MFD), is actually larger than the core itself since transmission extends into the cladding. The MFD can be 10 to 15% larger than the actual fiber core. The fiber is coated with a protective layer of plastic that extends the diameter of standard fiber to 250 microns.
Optical Signal to Noise Ratio (OSNR)
Optical signal to noise ratio (OSNR) is a measurement relating the peak power of an optical signal to the noise floor. In DWDM transmission, each amplifier in a link adds noise to the signal via amplified spontaneous emission (ASE), thus degrading the OSNR. A minimum OSNR is required to maintain good transmission performance. Therefore, a high OSNR at the beginning of an optical link is critical to achieving good transmission performance over multiple spans.
OSNR is measured with an optical signal analyzer (OSA). OSNR is a good indicator of overall transmission quality and system health. Therefore OSNR is an important measurement during installation, routine maintenance, and troubleshooting activities.
Optical Supervisory Channel
The optical supervisory channel (OSC) is a dedicated communications channel used for the remote management of optical network elements. Similar in principal to the DCC channel in SONET networks, the OSC inhabits its own dedicated wavelength. The industry typically uses the 1510nm or 1625nm wavelengths for the OSC.
Polarization Mode Dispersion (PMD)
Single mode fiber is actually bimodal, with the two modes having orthogonal polarization. The principal states of polarization (PSPs, referred to as the fast and slow axis) are determined by the symmetry of the fiber section. Dispersion caused by this property of fiber is referred to as polarization mode dispersion (PMD).
Raman fiber amplifiers use the Raman effect to transfer power from the pump lasers to the amplified wavelengths. Raman Advantages are:
- Wide bandwidth, enabling operation in C, L, and S bands.
- Raman amplification can occur in ordinary silica fibers
Raman Disadvantages are:
- Lower efficiency than EDFAs
An optical amplifier performs a 1R function (re-amplification), where the signal noise is amplified along with the signal. For each amplified span, signal noise accumulates, thus impacting the signal’s optical signal to noise ratio (OSNR) and overall signal quality. After traversing a number of amplified spans (this number is dependent on the engineering of the specific link), a regenerator is required to rebaseline the signal. A regenerator performs the 3R function on a signal. The three R’s are: re-shaping, re-timing, and re-amplification. The 3R function, with current technology, is an optical to electrical to optical operation (O-E-O). In the future, this may be done all optically.
The S-band is the “short” DWDM transmission band, occupying the 1485 to 1520nm wavelength range. With the “S+” region, the window is extended below 1485nm. The S-band has comparable bandwidth to the C-band, thus comparable total capacity. The S-Band advantages are:
- Low susceptibility to attenuation from fiber micro-bending.
- Lowest dispersion characteristics on SSMF fiber.
Self Phase Modulation (SPM)
The refractive index of the fiber varies with respect to the optical signal intensity. This is known as the “Kerr Effect”. Due to this effect, the instantaneous intensity of the signal itself can modulate its own phase. This effect can cause optical frequency shifts at the rising edge and trailing edge of the signal pulse.
SemiConductor Optical Amplifier (SOA)
What is it?
Similar to a laser, a SOA uses current injection through the junction layer in a semiconductor to stimulate photon emission. In a SOA (as opposed to a laser), anti-reflective coating is used to prevent lasing. SOA Advantages are:
Solid state design lends itself to integration with other devices, as well as mass production.
- Amplification over a wide bandwidth
- SOA Disadvantages are:
High noise compared to EDFAs and Raman amplifiers
- Low power
- Crosstalk between channels
- Sensitivity to the polarization of the input light
- High insertion loss
- Coupling difficulties between the SOA and the transmission fiber
Engineering a DWDM link to achieve the performance and distance requirements of the application. The factors of Span Engineering are:
Amplifier Power – Higher power allows greater in-line amplifier (ILA) spacing, but at the risk of increased non-linear effects, thus fewer spans before generation.
Amplifier Spacing – Closer spacing of ILAs reduces the required amplifier power, thus lowering the susceptibility to non-linear effects.
Fiber Type – Newer generation fiber has less attenuation than older generation fiber, thus longer spans can be achieved on the newer fiber without additional amplifier power.
Channel Count – Since power per channel must be balanced, a higher channel count increases the total required amplifier power.
Channel Bit Rate – DWDM impairments such as PMD have greater impacts at higher channel bit rates.
Standard single-mode fiber, or ITU-T G.652, has its zero dispersion point at approximately the 1310nm wavelength, thus creating a significant dispersion value in the DWDM window. To effectively transport today’s wavelength counts (40 – 80 channels and beyond) and bit rates (2.5Gbps and beyond) within the DWDM window, management of the chromatic dispersion effects has to be undertaken through extensive use of dispersion compensating units, or DCUs.
SSMF makes up about one-third of the deployed US terrestrial long-haul fiber plant. Approximately 20% of the new fiber deployment in the US long-haul market is SSMF. (Source: derived from KMI data)
Stimulated Raman Scattering (SRS)
The transfer of power from a signal at a lower wavelength to a signal at a higher wavelength.
SRS is the interaction of lightwaves with vibrating molecules within the silica fiber has the effect of scattering light, thus transferring power between the two wavelengths. The effects of SRS become greater as the signals are moved further apart, and as power increases. The maximum SRS effect is experienced at two signals separated by 13.2 THz.
Thin Film Filter
A thin film filter is a passive device that reflects some wavelengths while transmitting others. This device is composed of alternating layers of different substances, each with a different refractive index. These different layers create interference patterns that perform the filtering function. Which wavelengths are reflected and which wavelengths are transmitted is a function of the following parameters:
- Refractive index of each of the layers
- Thickness of the layers
- Angle of the light hitting the filter
Thin film filters are used for performing wavelength mux and demux. Thin film filters are best suited for low to moderate channel count muxing / demuxing (less than 40 channels).
Optical networking often requires that wavelengths from one network element (NE) be adapted in order to interface a second NE. This function is typically performed in one of three ways:
- Wavelength Adapter (or transponder)
- Wavelength Converter
- Precision Wavelength Transmitters (ITU l)