Tutorial

tutorial
As passive components become active,
network designers face integration challenges.
active
lifestyles
This is an expanded version of the article
that appeared in oemagazine.
By Brian Lavallée and Frank Santillo,
Nortel Networks
As optical networks continually strive to increase line
rates, capacity, flexibility, and reach, significant changes to
both active and passive components are fast becomingenecessary and will soon be commercialized for wide-scale
deployment. Traditional passive components such as dis-
persion compensation modules (DCM), gain flattening
filters (GFFs), multiplexing/demultiplexing
(MUX/deMUX) filters, optical add drop module
(OADM) filters, and polarization-mode-dispersion
ILLUSTRATION BY HANK OSUNA(PMD) maintaining fiber will soon be offered as active
ing industry over the past few years, however, many ven-modules and sub-systems. Although these changes will
dors emerged that now offer active versions of these oncesurely add to the overall robustness, flexibility, reliability,
passive modules, claiming improved performance, reach,and performance of tomorrow’s optical network, they will
and capacity. However, most of these devices are offeredalso add to the overall complexity of the network as well.
in isolation, due to the specific expertise required for theirEach active module will operate dynamically and will
development. Thus, the onus is on the system integratortherefore be controlled by software and be integrally
to incorporate these exotic technologies into a unifiedincorporated within the system control loop to ensure
control system in wihich performance and flexibility leapproper operation of the codependent active components.
forward, but with an increase in system complexity. ThisFor example, tunable lasers will have to be synchronized
increased complexity must be designed into the system towith their paired tunable OADM filter settings to allow
enable simple network deployments and reduce operationfor propagation throughout the optical link. However,
expenditures. The level of complexity of today’s opticaltuning these wavelengths also requires subsequent changes
networks requires significant expertise that incurs consid-to the gain profiles of the optical amplifiers (erbium-
erable expenses which existing carriers, in the presentdoped fiber amplifiers (EDFAs) or Raman amplifiers),
environment of fiscal constraint, can ill afford to maintainthereby resulting in adjustments to the dynamic GFF
going forward.modules as well. Since all of these active components are
Although some passive devices such as dispersion com-inherently related at the optical propagation layer, any
pensation modules will go active, existing active moduleschanges in the properties of one module will have a sys-
such as lasers will increase performance and flexibilitytem-wide implication. These optical codependences mean
through advanced design. Each promises improvementsthat the control system will surely increase in complexity.
in network performance and flexibility only if they can beMany components in today’s deployed 10 Gb/s based
successfully integrated into a control system managed byoptical networks are passive (not electrically powered).
software both at the local control layer and the networkDue to the huge influx of capital into the optical network-
control layer. There must be provisions in such a system
30 spie’s oemagazine july 2003| to effectively cope with failures in a given module without the changing residual dispersion over time.
affecting overall system performance. For instance, if the com- In an ideal world, an optical amplifier system would pro-
munications link between the primary controller and its vide the same gain profile across the entire wavelength
subtending active optical modules is lost due to a fiber break, spectrum to ensure relatively identical signal-to-noise
these active modules must continue to ratios (SNRs) at the receiving demultiplexer site.
operate in a holdover mode until the com- Deployed wavelengths with the maximum spacing
munications path has returned. This is between one another would experience the same gain in
analogous to synchronous network ele- this theoretical yet unattainable model. Of course, the practical
ments operating in a local holdover world deviates significantly from the ideal world, which is
timing mode until the lost system timing essentially the primary need for the field of engineering in gen-
source is restored. Telecom companies eral. EDFAs actually exhibit a tilted gain that is
will not and should not tolerate network wavelength-dependent across the supported spectrum of wave-
availability below the defacto standard of lengths, with local fluctuations referred to as ripple.
99.999%, which is often cited as the “five Other non-linear factors such as polarization dependent loss
nines” of availability. and stimulated Raman scattering also contribute to increase the
The remainder of this article discusses resulting gain disparity between wavelengths. Consequently,
modules that are most likely to become active the optical SNR will be quite different from one wavelength to
in the coming generation of commercial opti- another at the individual receivers. Since the overall system per-
cal networks based on current industry trends formance is limited by the performance of the weakest channel,
and requirements. These emerging technolo- equalization processes (methods to balance the received optical
gies have been demonstrated in numerous lab SNR values) are required to optimize link performance and
experiments and described in generally avail- ensure overall network robustness. The idea is simple: steal
able technical conference papers. from the rich (strongest wavelength) and feed the poor (weakest
wavelength).
Slope compensation and Passive GFFs used today compensate for the non-linear gain
gain flattening profile of optical amplifiers, resulting in a window known as
Next-generation 40 Gb/s networks will the design flat gain where input channel powers achieve a mar-
implement quadrupled line rates that ginally flat gain profile. Other more sophisticated mechanisms
result in bit unit intervals that are four reside in the amplifier control loop that measure incoming sig-
times narrower than 10 Gb/s-based net- nals and automatically compensate them such that the
works. Consequently, the inverse square amplifier operates in the flat gain region even in the event of
relationship between dispersion suscepti- power transients (sudden loss or gain of wavelengths).
bility and line rate means that 40 Gb/s Although these devices minimize gain tilt, which provides some
networks are actually 16 times more sus- improvements in system performance, they do not offer indi-
ceptible to pulse spreading and distortion vidual wavelength control required to maximize system
than 10 Gb/s networks. Existing dispersion management strate- performance. For smaller systems, a centralized equalization
gies implement passive dynamic dispersion slope compensation strategy in which passive filters would serve to optimize link
modules (DSCMs) that may prove inadequate for 40 Gb/s net- performance is sufficient. As the demand for optical reach con-
works thus mandating the introduction of dynamic DSCMs tinues to increase, passive filters are no longer sufficientægreater
instead. This will enable dynamic dispersion compensation for individual wavelength control is required. In larger systems, a
any index of refraction changes along a given fiber route over distributed equalization scheme, such as one based on dynamic
time. Fixed dispersion compensation can still be deployed with GFFs, is instead required.
advanced active dispersion compensation modules managing A dynamic GFF is actually a wavelength-dependent attenua-
tor that can be used to control the amplifier tilt and ripple (see
figure 1). The power spectrum of the incoming optical signal,
P (f), is shown where P is the power of the input signalin
expressed in dBm and f is the frequency. P (f), the desiredout
output power spectrum, is used to calculate the attenuation
profile of the dynamic GFF as a function of frequency and can
be derived using the relationship:
A(f) = P (f) – P (f)in out
Phenomena that could create amplifier tilt and ripple, and
consequently affect system performance, include component
Figure 1. Operation of the dynamic GFF aging, non-linear effects, and changes in ambient operating
conditions. A variety of technologies allow dynamic GFF
july 2003 spie’s oemagazine 31|designs to actively compensate for unwanted energy transfer dency to achieve the desired tuned wavelength quickly.
from one wavelength to another. The second reason is dynamic wavelength routing, which
allows you to change the path of a wavelength from its ingress
point (transmitter) to its egress (receiver) point via a series ofMUX/deMUX
passive/active OADMs and/or switches. The latter applicationMUX/deMUX filters serve two fundamental roles when
requires a faster response time than the former given that thedeployed in optical networks. First, they multiplex (combine)
rerouting of wavelengths may be the result of a protectionindividual wavelengths into one fiber for dense wavelength
switch. In practice, the total recovery time, including the wave-division multiplexing transmission and demultiplex (separate)
length tuning, must not exceed 50 ms in order to comply withthese combined signals at the term