Optical network maintenance made simple

Optical network maintenance made simple

The key to successful dwdm deployment is the provision of the same maintenance and monitoring performance levels delivered by traditional network approaches.

Larry Yu and Steven Hersey, Lucent Technologies

To any communications service provider that has operated a network in the last few years, the news of fiber exhaust is hardly new. The industry has witnessed a seemingly insatiable demand for bandwidth capacity--fueled by Internet access, voice, data, and video transmission--and this demand is accelerating at rates beyond even the most bullish growth calculations. One carrier has reported that the growth of its network in 1997 was nearly equivalent to the size of its entire network in 1991. Another has said that the size of its network doubles every six months.

In 1995, it was estimated that between 70% and 80% of the fiber in North America`s networks was currently "lit" or in use. So it comes as no surprise that many carriers today are nearing 100% capacity in certain portions of their networks.

If a potential capacity crisis is the old story, the new chapter should be the availability of a manageable solution that is resilient and reliable. Dense wavelength-division multiplexing (dwdm) allows service providers to exploit untapped fiber capacity by enabling more than one optical signal, or wavelength, to be carried on each fiber. The capacity potential of this technology is only now beginning to emerge, with systems capable of carrying up to 80 wavelengths or more and 400 Gbits/sec of traffic becoming available soon. By taking advantage of dwdm, the percentage of embedded fiber that may be lit is no longer an issue. Fiber capacity is now determined by counting the available wavelengths on each fiber, rather than counting unlit fiber strands.

While many network carriers are already migrating to dwdm, others are still evaluating the technology. Such evaluations likely will reveal that deploying these new optical transport systems doesn`t mean sacrifices in reliability, management, and maintenance. In fact, choosing the right optical elements will provide the same features and reliability as traditional networking solutions.

Built-in reliability

To take advantage of the capacity potential of a dwdm-based solution, optical systems must offer a range of maintenance capabilities that will allow the technology to be seamlessly integrated into the existing network. Seamless integration is most readily accomplished by selecting and deploying dwdm systems designed to incorporate a comprehensive array of management and maintenance capabilities familiar to those with sonet elements. As such, these optical transport systems must be similarly able to detect failures (and maintain service during such failures), monitor transmission performance, isolate faults to specific circuit packs, and report trouble conditions. To provide these vital maintenance capabilities, the dwdm systems must include a wide range of signal monitoring, fault isolation, protection switching, and fault-reporting features.

If there is any difference in the maintenance of a dwdm system as compared to sonet, it should only be the level where the maintenance features operate. The high degree of maintenance and management capabilities of the two systems should be equally comprehensive. In sonet elements, maintenance information can be embedded in the overhead portion of a signal. This information enables network elements and network-management systems to isolate failures within a maintenance region, often correctly isolating problems to a single circuit pack at a single site.

In the optical transport system, maintenance is performed at the optical level. Built-in optical monitoring and maintenance capabilities are essential to simplify management of multiwavelength optical networks by making it possible to locate isolated failures and manage each wavelength independently.

If you can find it, you can fix it

Having built-in maintenance functions is critical to guarantee operations within any network, and carriers should expect and rely on both reactive and proactive maintenance to support the same quality of service and reliability in dwdm optical networks that is being delivered through sonet networks today. Reactive maintenance allows carriers to identify a failure after it occurs and is performed by monitoring the power of the multiplexed optical signal. When a change in signal power is detected, automatic diagnostics enable the system to determine the exact cause of the change, isolating the failure to a circuit pack or signal level.

While detection and isolation of failures is essential, maximizing service reliability and ensuring that services unaffected by the failure are not degraded or compromised is equally important. Given that dwdm systems transport multiple wavelengths simultaneously, failures can occur that affect some, but not all, of the wavelengths (e.g., loss of signal to one input signal). As such, optical elements must ensure that wavelengths unaffected by a failure continue to be transported without impairment. To achieve this goal, the power output or gain of the amplifier must be readjusted to ensure the remaining channels are unaffected by the failure. Without this power adjustment, the remaining gain could be too high and easily corrupt the remaining "good" signals.

To avoid degrading service on channels unaffected by system failures, amplifier gain adjustment must be accomplished by the system automatically. Upon clearing the defect, normal operation of the optical amplifiers automatically resumes. Solutions that require manual gain adjustment will lead to service disruption that could have otherwise been prevented. Therefore, automatic amplifier gain adjustment should be a key requirement when selecting a dwdm system.

Fiber cut? No problem.

Another important consideration in ensuring high-quality, high-reliability service delivery is automatic optical protection switching. While effective detection and reporting are necessary to allow rapid response to repair failures, automatic optical-protection switching can significantly enhance the survivability of the network by enabling service to be maintained during the failure repair interval.

Optical-layer protection switching is particularly important when carrying asynchronous or other traditionally unprotected types of traffic such as Asynchronous Transfer Mode, Internet protocol, or video. The optical protection switch provides resilience against optical amplifier failures and/or fiber cuts in the optical network. dwdm systems can rapidly detect signal loss and switch traffic to restore service during the repair. As in sonet applications, optical protection switching can be accomplished in less than 50 msec, ensuring the robust performance in the optical layer of the network that service providers expect with today`s full range of traffic types.

Fix it before it fails

The proactive maintenance features of dwdm systems are equally important to the operation of a network. Proactive maintenance detects conditions not severe enough to be considered a failure, but may be indicative of an impending problem. A prime example of a proactive maintenance feature is per-wavelength signal-power monitoring, which detects degrading performance on one or more wavelengths by setting performance parameters in the optical system.

To illustrate, for each line and channel, signal power and laser-device health are continuously measured. Every time an optical channel is added or removed, a nominal threshold value is established. When a threshold value is crossed, a threshold-crossing alert is reported through the message-based interface to the operations system (OS). Based on the threshold-crossing alerts associated with a particular path, the OS is able to correlate alerts and identify the likely source of the degradation. The crossing alert reported for each performance-monitoring parameter can be enabled or disabled to allow for specific maintenance activities.

A reporting feature integrated into proactive maintenance information is an important tool for trend analysis. Performance-monitoring information can be stored in the network element in the form of registers, which should be available in hourly and daily intervals for retrieval by the OS to provide carriers with the data necessary for trend analysis.

The inevitable network

As service carriers move to the inevitable all-optical network, maintenance and management capabilities that ensure reliability in the network elements are critical. A comprehensive maintenance system requires that each dwdm network element be able to detect actual and impending failures, monitor performance degradation, isolate faults to specific circuit packs, automatically adjust output power, and provide protection switching.

While dwdm is clearly the right technology for establishing an infrastructure capable of keeping pace with long-term capacity growth, the ability to maintain and manage these systems is key to reaping the full benefits of this technology. Requiring a full range of maintenance capabilities from optical transport systems will assure carriers that their optical elements are just as reliable and robust as traditional networking solutions. Carriers who have these maintenance capabilities in check are paving the way for business growth through the security of a network with virtually limitless capacity. q

Optical-protection switching in conjunction with dwdm systems ensures network availablility during repair periods.