"Introduction". In: Fiber-Optic Communication Systems

7.9. HIGH-CAPACITY SYSTEMS 313

have been developed for which the signs of bothDandSare reversed compared with
the conventional dispersion-shifted fibers. Dispersion map in this case is made using
roughly equal lengths of the two types of fibers.
Many experiments during the 1990s demonstrated the usefulness of DCFs for WDM
systems. In a 1995 experiment [148], 8 channels with 1.6-nm spacing, each operating
at 20 Gb/s, were transmitted over 232 km of standard fiber by using multiple DCFs.
The residual dispersion for each channel was relatively small (∼100 ps/nm for the
entire span) since all channels were compensated simultaneously by the DCFs. In a
2001 experiment, broadband DCFs were used to transmit a 1-Tb/s WDM signal (101
channels, each operating at 10 Gb/s) over 9000 km [158]. The highest capacity of
11 Tb/s was also realized using the reverse-dispersion fibers in an experiment [159]
that transmitted 273 channels, each operating at 40 Gb/s, over the C, L, and S bands
simultaneously (resulting in the total bandwidth of more than 100 nm).

7.9.2 Tunable Dispersion Compensation

It is difficult to attain full GVD compensation for all channels in a WDM system.
A small amount of residual dispersion remains and often becomes of concern for
long-haul systems. In many laboratory experiments, a postcompensation technique
is adopted in which the residual dispersion for individual channels is compensated by
adding adjustable lengths of a DCF (or a fiber grating) at the receiver end (dispersion
trimming). This technique is not suitable for commercial WDM systems for several
reasons. First, the exact amount of channel-dependent residual dispersion is not al-
ways known because of uncontrollable variations in fiber GVD in the fiber segments
forming the transmission path. Second, even the path length may change in reconfig-
urable optical networks. Third, as the single-channel bit rate increases toward 40 Gb/s,
the tolerable value of the residual dispersion becomes so small that even temperature-
induced changes in GVD become of concern. For these reasons, the best approach may
be to adopt a tunable dispersion-compensation scheme that allows the GVD control for
each channel in a dynamic fashion.
Several techniques for tunable dispersion compensation have been developed and
used for system experiments [160]–[167]. Most of them make use of a fiber Bragg grat-
ing whose dispersion is tuned by changing the grating period ̄nΛ. In one scheme, the
grating is made with a nonlinear chirp (Bragg wavelength increases nonlinearly along
the grating length) that can be changed by stretching the grating with a piezoelectric
transducer [160]. In another approach, the grating is made with either no chirp or with
a linear chirp and a temperature gradient is used to produce a controllable chirp [164].
In both cases, the stress- or temperature-induced changes in the mode index ̄nchange
the local Bragg wavelength asλB(z)=2 ̄n(z)Λ(z). For such a grating, Eq. (7.6.6) is
replaced with

Dg(λ)=

dτg

dλ

=

2

c

d

dλ

(∫L

g

0

n ̄(z)dz

)

, (7.9.3)

whereτgis the group delay andLgis the grating length. The value ofDgat any wave-
length can be changed by changing the mode index ̄n(through heating or stretching),
resulting in tunable dispersion characteristics for the Bragg grating.