Additional testing requirements for DWDM-amplified links

Additional testing requirements for DWDM-amplified links

With the transmission capabilities of DWDM increasing the power of fiber-optic networks, testing procedures must evolve to measure wavelength parameters.

Booker H. Tyrone Jr. SBC Technology Resources Inc.

Xavier Lee Nortel Networks

Stephane Vigot EXFO Electro-Optical Engineering

The optical layer comprises various components or building blocks: transmitters and receivers, dense wave-length division multiplexers (DWDM) and demultiplexers, dispersion-compensation devices, optical amplifiers, optical add/drop multiplexers, and the optical fiber itself (see Fig. 1). Each one of these components has minimum performance expectations based on technical specifications, and each plays a critical role in overall performance at the system level. For example, slight crosstalk at the multiplexer level, which could be negligible during factory tests, can greatly influence transmission quality when combined with other components.

Building blocks

Not everyone needs to perform the same level of detailed testing. Equipment manufacturers and suppliers of these components are required to qualify their product specifications according to International Telecommunication Union (ITU) and Bell Communications Research (Bellcore) standards. But end- users of these systems, such as service providers, may only be required to test their fiber plant for chromatic dispersion or polarization-mode dispersion (PMD) impairments. Following are some of the testing requirements for the various building-block components.

Specific transmitter wavelengths for DWDM systems follow a wavelength grid based on ITU standards. Wavelength meters characterize the accuracy and long-term stability of the DWDM transmitter output. Laser linewidths can also be measured with a very high-resolution wavelength meter. Equipment testers can verify provisionable transmitter-output power levels with optical power meters and use optical spectrum analyzers (OSAs) to measure the sidemode suppression ratio. Optical return loss (ORL) from connectors is gauged using corresponding meters.

By using a programmable variable optical attenuator (pVOA) and a bit-error rate (BER) test set, parameters such as receiver sensitivity, input power range, signal degradation, and loss-of-signal (LOS) thresholds can be verified.

Multiplexers and demultiplexers

Multiplexers and demultiplexers in DWDM systems are actually sophisticated multipassband filters, because they combine and separate multiple wavelengths in and out of a single fiber in an optical link. As such, key parameters like the spectral shape and filtering characteristics need testing. In particular, center wavelengths, bandwidth, insertion loss, and the ripple or flatness of each passband should be characterized.

To ensure negligible crosstalk among all channels, equipment tests should reveal high isolation between worst-case adjacent and neighboring channels. It`s also important to verify insertion-loss uniformity among channels to test the differential loss or device ripple. If the multiplexers/demultiplexers have upgrade ports to accommodate future capacity growth, it`s wise to perform similar tests on the upgrade passband. All these measurements can be performed using an OSA.

High-density multiplexers and demultiplexers usually contain many connector ports, which increase the potential of obtaining a source of high reflection; therefore, ORL measurements on all connectors are essential. Polarization-dependent loss (PDL) as a function of wavelength can be verified using a combination of a tunable laser, polarization controller, and PDL meter or detector. Although PMD is expected to be quite small in these devices, it can adversely affect high-bit-rate systems and therefore must be measured with the appropriate equipment.

Dispersion-compensating devices

Using a dispersion-compensating device (DCD) to apply an equal but reversed dispersion on an optical pulse traveling through the fiber will counter the broadening of the pulse from chromatic dispersion. Currently, two types of DCDs are commercially available: dispersion-compensating fiber (DCF) and dispersion-compensating grating (DCG). In the case of the DCG, equipment testers should characterize the delay ripple and the insertion-loss ripple over the working bandwidth. They should also use dispersion analyzers to gauge the dispersion of these devices over the bandwidth of interest. Other tests and measurements include insertion loss and PDL over the bandwidth, ORL, and the use of PMD analyzers to verify the mean differential group delay (DGD).

Optical amplifiers

The natural gain spectrum of erbium-doped fiber amplifiers (EDFAs) is not uniform; the peak falls in the 1530- to 1535-nm region. Most commercially available optical amplifiers used in long-haul telecommunications systems have gain-flattening filters built into the EDFA construction. The gain of the amplifier is characterized as a function of wavelength; the amplifier`s spectral gain flatness requires evaluation, and its gain equalization needs to be ensured to guarantee uniform and balanced spectral amplification. This consideration is especially important in optical links having cascaded amplifiers. Characterizing the spectral content of the EDFA noise or amplified spontaneous emission (ASE) is also necessary to quantify the optical signal-to-noise ratio (OSNR--see Fig. 2).

Equipment testers should also conduct a series of tests to characterize the gain and noise figure (NF) across the working bandwidth. A series of gain and NF curves at several probe wavelengths can be obtained at varying gain levels for a reference wavelength. From these series of curves, equipment testers can determine the amplifier NF, gain ripple, and tilt across the working bandwidth at varying gain levels. OSAs with built-in EDFA programs can be used to analyze the spectral responses to determine these parameters.

Because some components in the EDFA may exhibit sensitivity to polarization or birefringence, polarization-dependent loss/gain (PDL/PDG) and PMD can be gauged. Other measurements include ORL as well as the optical amplifier`s performance sensitivity to backreflection, which is tested using backreflection reference meters.

Optical add/drop multiplexers

Optical add/drop multiplexers (OADMs) use different filtering technologies: dielectric, acousto-optic, fiber Bragg grating, arrayed waveguide gratings (AWGs), and others. Regardless of which filtering technology is used, the principle function of the OADM is to add and drop specific wavelengths without interfering with the remaining wavelengths that pass through the filter. Therefore, three routes need verification: the in-out (pass-through), the in-drop (drop), and the add-out (add) path. For each of these paths, equipment testers should gauge the passband insertion loss and ripple, channel isolation, ORL, and PDL.

Fiber plant

Before deploying a DWDM system in the field, the operator must first characterize certain parameters in the fiber plant to assess which system configuration and technology the fiber network can handle. As line rates increase, the effects of chromatic dispersion and PMD--due to distortion and dispersion-induced nonlinear effects--also multiply. Chromatic-dispersion analyzers measure the total dispersion in the fiber link and indicate how much dispersion compensation is required. PMD analyzers gauge the differential group delay (see Fig. 3). Optical time-domain reflectometers (OTDRs) can characterize span loss and fiber lengths.

Testing at the system level

With all the building-block components working together in the optical link, it is important to perform end-to-end optical testing at the system level. The net dispersion between the transmitter and receiver for all the wavelengths, including OADM wavelengths, requires characterization to ensure that it is within specified levels. Other key measurements include optical power and OSNR at the receiver site. In some DWDM systems, to achieve balanced OSNR levels among all the channels, an OSA is used to monitor changes in the OSNR with different transmitters.

From a vendor`s point of view, after a system is set up and optimized, testing the optically amplified system`s transmission performance is key to verifying the optical link budgets. To check the operating condition of the system, the optical link is stressed to the point at which the maximum permissible BER occurs, for example a BER of 10-12. One way to stress the optical link is by adding supplementary attenuation to it; another is to introduce noise to the system, hence reducing OSNR to determine the noise margin. Vendors should also characterize the penalty due to adding distortion to the system. The noise margin should then take into account allowances for aging, temperature variation, and transmitter/receiver compatibility; it should also correct for worst-case ripple and noise, power fluctuations, and link distortions due to nonlinear effects. The guaranteed end-of-life span loss is determined using the net margin and initial average span loss in the operating condition.

From the network operator`s perspective, a minimum number of system tests that demonstrate the overall system functionality and the integrity of the signal transmission is desired. The procedure to test and deploy Synchronous Optical Network/Synchronous Digital Hierarchy (sonet/sdh) single-channel optical transmission systems is well-documented. But emerging technologies like WDM lack such detailed procedures. Optical component properties and cable characteristics that could safely be neglected in systems using simpler transmission techniques must now be considered.

Continuing advances such as add/drop multiplexers are giving rise to wavelength routing over complex networks. Optical crossconnects, though still on the drawing board, offer the potential to change how future telecommunications networks are constructed. An effort is under way to define a set of system-level tests that characterize end-to-end network functionality such as optical signal transmission, multiplexing, supervision, performance monitoring, and survivability. These tests are categorized as follows:

Network compatibility

Transmission performance

Operations, administration, and maintenance

Multivendor operability

Architecture interconnections

Network compatibility tests ensure that the existing signal works with the fiber interfaces to the WDM network elements (NE). These tests also verify that the electrical power specifications of the WDM NE match those of the existing network. Some essential field tests measure end-to-end attenuation, optical reflections at the network interfaces, and dispersion when the systems are being turned up. End-to-end attenuation testing at 1310 and 1550 nm examines the fiber path for optical discontinuities and determines whether the fiber conforms to system requirements. Testing for reflected power from individual components, such as connectors and fiber joints, determines whether the impact of single and multiple reflections is within specified tolerance requirements.

Chromatic dispersion and PMD testing is not routinely done on newer fiber unless DCDs are present in the system because these parameters are generally specified by cable manufacturers. But it`s important to conduct these tests on older fiber plant considered for high-bit-rate transport, such as OC-192 (10 Gbits/sec), as well as with fiber distances great enough to encroach on system dispersion limitations.

Transmission performance tests measure the functioning of the transmitter/ transponder and receiver, optical amplifier, and BER. These tests also track the effects that accumulate through WDM NEs. Laser wavelengths may vary because of fabrication variations, thermal shifts, or frequency broadening. When a spectrally broadened signal propagates through a dispersive medium like conventional fiber, degraded error performance can result. The number of wavelengths available to a WDM system is limited by crosstalk, component drift, fiber nonlinearity, and cascaded filter alignment tolerances. Because of the use of optical amplifiers to compensate for component and transmission losses, the impact of noise accumulation and cross-gain saturation needs to be carefully controlled to minimize signal power variations in the amplifier.

Transmission performance tests verify end-to-end optical system performance. Channel wavelength, channel drift, OSNR, channel spacing, channel isolation, insertion loss, amplifier noise and gain flatness, BER, jitter, and transmission delay are all critical measurements.

Operations, administration, and maintenance tests confirm the capability of performing network operations and management functions. When maintaining a WDM system, the operator ensures the optical carriers remain within specifications, the wavelengths stay stable, and the power does not fluctuate more than the system can tolerate. These tests make use of data-communications (optical service) channels that enable communication between the WDM NEs (NE/NE) and the operation systems and network elements (OS/NE). Information that must follow a particular connection is transmitted via the embedded data-communications channel, local area networks, or optical supervisory channel.

Verifying security

Alarm surveillance and performance monitoring are tested. Key provisioning features and parameters are validated as provisioned locally at the craft interface or remotely from an operating system interface. System administration and security functions also need verification. These tests ensure that the operations communication functions meet the needs of individual architectures and allow the network operator to discover and control the state of the network.

As on any kind of system, backreflection is still a concern. The light reflected back could cause parasitic interference effects, which can lead to power fluctuations, phase noise, and pulse distortion. If the reflections are particularly strong, they may induce instability in the transmitting laser that triggers alarming and a switch away from the faulty path.

Multivendor operability tests substantiate that the WDM systems interoperate with other vendors` transport technology equipment. To date, optical translators are incorporated to provide an interface between various vendors` equipment. To achieve interoperability between vendors` equipment without using optical translators, the physical properties of the optical signal and the content (wavelength, line rate, coding format) at the network interface must be standardized. The ITU and other standards bodies are actively defining specifications for optical interfaces, architectures, and management. These interoperability tests verify transport of client layer signals across multivendor equipment.

Architecture interconnection tests corroborate the functionality and survivability of the networks` links. Because today`s WDM systems were designed primarily for point-to-point connectivity over long distances, optical network survivability is provided almost exclusively at the electronic layer by SONET. These systems are augmented by systems that allow static wavelength add/drop functionality. Service protection, an essential aspect of the architecture, is tested for network reliability, survivability, and service consistency.

Soon, however, WDM systems will work in conjunction with dynamically reconfigurable add/drop systems. High-performance switches and routers will route wavelengths over complex networks. Optical crossconnects and other advanced optical technologies will support flexible topologies. Restoration architectures based on self-healing rings, diverse routing, or mesh-based distributed algorithms are also receiving considerable attention.

The evolution of WDM technology will undoubtedly allow operators to get closer to the intrinsic capacity of their fiber-optic networks. Systems designers, as well as vendors, may choose different approaches to increase the power of their networks. Increasing the number of carriers--by expanding the operating WDM window and/or reducing the channel spacing--may work for some, while others opt to raise the data rate of each carrier. Whatever the solution, testing procedures will have to evolve along with the technology. u

Booker H. Tyrone Jr. is a member of the technical staff at SBC Technology Resources Inc. (Austin, TX). Xavier Lee is a systems engineer in the optical-layer applications division of Nortel Networks (St. Laurent, QC, Canada). Stephane Vigot is a product manager at the outside-plant division of EXFO Electro-Optical Engineering (Vanier, QC, Canada).

More in DWDM & ROADM