Solid-state field-deployable optical monitors are designed to measure and deliver key network data on absolute wavelengths, amplitudes, and OSNR-fast.
The success of dense wavelength-division multiplexing (DWDM) in increasing raw capacity on point-to-point fiber routes and the rapid growth of data communications bring new network-management challenges. The premise of all-optical networks requires the availability of tools to measure and control the smallest granular component of such networks--the wavelength channel. We'll outline how field-deployable optical-network monitors (ONMs) measure and report key channel parameters important in network operation. Some of these functions include the monitoring of amplifiers and switches at add/drop sites, the deployment and commissioning of DWDM routes, as well as the restoration and protection of networks.
The objective for ONMs is to remotely report the absolute wavelengths, amplitudes, and optical signal-to-noise ratio (OSNR) of many DWDM channels. This must be accomplished with speed and accuracy over an extended period of time. The key here is speed. Laboratory optical spectrum analyzers (OSAs) normally average the signal over long periods in order to improve OSNR. A remotely deployed network monitor, in contrast, needs to report optical impairments faster than the Synchronous Optical Network (SONET) 50-msec network restoration time.
ONMs support both the deployment and commissioning of DWDM routes. During the setup phase, ONMs establish an independent and easily accessible measurement point to check for provisioning problems such as bad splices, faulty connectors, incorrectly set laser wavelengths or under/over powered amplifiers. Improperly provisioned networks compromise the design margins typically reserved for component aging.In Figure 1, the optical-network monitor shows wavelength tracking data for two arbitrary distributed-feedback laser sources (A & B) exhibiting wavelength drift. The first (about 0.1 nm) is due to slow changes in room temperature, while the second (about 0.02 nm) is due to fast onboard temperature regulation. Start time is Friday at 5 pm.
It is common practice to document the baseline for all operating parameters such as signal power, bit-error rate, OSNR, etc., prior to network turn-on. During normal operation, the ONM reports any degradation events of the optical channel--due to environmental perturbations or component aging--to the network operations center (NOC). It also documents the degradation of the physical layer in time by saving optical performance data in an archival database (see Fig. 1). As DWDM channels are added to the system, network installers can monitor the system's status as these channels are gradually turned on, and check the power and OSNR of both the added and the existing channels.
With the deployment of DWDM systems, an emerging trend is to lease channels or wavelengths. Invariably, as channels are handed off between carriers, such as an interexchange carrier (IXC) and a regional Bell operating company (RBOC), problems can occur which require monitoring to resolve conflicts. Most of these issues occur at network boundaries.
Channel routing is a crucial function in all-optical networks. A variety of elements are deployed to support flexible routing such as add/drop multiplexers (ADMs), channel routers, wavelength converters, and switches. An emerging trend in network design and configuration is the move from static to dynamic wavelength provisioning. Here again, independent monitoring is needed to ensure quality and continuity of service.
Another monitoring approach is to modulate the carrier of each channel with a specific coded frequency. Channel power can be inferred by checking the amplitude of this filtered frequency. This method integrates the monitoring function with the multiplexer/transmitter for each channel and is effective for checking the amplitude of individual channels. However, channel wavelength itself remains unmonitored.
A form of error checking can be carried out at the SONET level by monitoring the B1 and J0 bits of the SONET stream. These bits indicate if errors have been received, but they do not supply channel-performance data. A more sophisticated approach is a Q-factor measurement of the signal, which can be done by sampling the received signal. This approach is very costly, however, since functionally a demultiplexer is needed at every node and a receiver is required for each channel. Another drawback to this method is that it does not isolate the optical from the electrical impairments if faults occur.
Desirable attributes for an ONM fall into three categories: power measurement accuracy, wavelength measurement accuracy, and management interface. The ONM's measurement accuracy should match that of benchtop instruments, particularly as capacity continues to increase per fiber. Furthermore, calibration needs for these instruments should be at a minimum since these units may be installed at unmanned sites. This attribute alone requires an instrument which is compact and is a solid state (no moving parts). Most laboratory benchtop instruments have scanning mechanisms that render them susceptible to shock and vibration; thus, requiring continuous calibration.
Since power-sampling taps are often used prior to the measurement terminal, it is advantageous to incorporate the tap ratios, or any inline attenuators, in the absolute power measurement reflected by the instrument. Absolute wavelength measurement accuracy is required, as referenced in the International Telecommunication Union (ITU) vacuum standards, with minimal recalibration. The emerging S (upper 1400-nm region) and L (1570- to 1625-nm) transmission bands demand that ONMs are adaptable to address these wavelengths. Current installations operate in the C-band, the erbium-doped fiber amplifier (EDFA) window of 1530 to1560 nm with channel spacing of 50 GHz.
Other ONM attributes are required to facilitate network management such as individualized minor and major alarm threshold settings per channel, fast alarm reporting, and a flexible interface to the NOC.
In a typical field installation of an ONM, a 1% tap off the transport fiber samples the optical signal. At the same time, the ONM correlates the measurement to the alarm thresholds, and reports the alarms and data to the NOC via the Ethernet channel (see Fig. 2).
Figure 3 demonstrates the ONM's accessibility over Ethernet using a commercially available browser. Multiple users can simultaneously access the same channel information and observe historical data as it is updated by the ONM on a per channel basis. This approach supports the reporting of all alarms to a centralized location such as the NOC. Other communication ports are required for local craft access.
The attributes described in this article can be found in an ONM, which is designed as a standalone, remotely deployed OSA. The solid-state design is based on a 256-element indium gallium arsenide (InGaAs) linear detector array and a conventional diffraction grating for spectral dispersion. The input wavelengths (1528 to 1562 nm) are dispersed across the array elements. The array is located at the focal plane of the optics, has a pixel pitch of 50 microns and total length of 12.8 mm. Other wavelength ranges are easily accessible with a simple internal adjustment.
The solid-state design of the monitor addresses stability issues over time. The absence of any moving parts allows for a long life without frequent recalibration. The instrument's sensitivity, down to -65 dBm, permits comfortable detection of transmitted optical signal even after significant attenuation. A 1% tap is sufficient to sample the signal in the fiber with minimal power penalty. For 100-GHz channel spacing, the unit's peak-to-valley contrast ratio is better than 25 dB, for 50-GHz channel spacing it's better than 15 dB.
A major advantage of this design is that the ONM can perform simultaneous measurement of the spectrum across all 256 elements. This parallel optical processing allows for high-speed channel detection, which minimizes alarm-reporting delays. In effect, it is 256 times faster than a single detector.
An embedded digital-signal processor is used for signal detection and instrument control. A peak-finding algorithm identifies channels and measures channel power, wavelength, and OSNR. Although each pixel corresponds to roughly 0.13 nm of spectral dispersion, an internal calibration improves the wavelength measurement accuracy by roughly a factor of 5.
Optical performance monitors are a necessary element in the realization of all-optical networks. Monitors support fault isolation and detection, as well as open the door for other system capabilities. For example, by reporting the amplitudes of all channels to the transmitter end, one can actively flatten the received spectrum of all channels simultaneously by pre-emphasis of the transmitted amplitudes.
A solid-state ONM design, based on a 256-element InGaAs linear detector array, provides fast acquisition and transfer of data to the NOC. This is highly advantageous in revealing transients due to changes of polarization dependent losses, thermal effects, or instabilities(information that previously may have gone undetected due to the slow response of OSAs based on scanning-and-averaging mechanisms.
Sami Hendow is director of engineering for fiber communications products and Salim Jabr is vice president of engineering at Ditech Communications Corp. (Mountain View, CA).