to 1=r
3
co
, whereas fo r TM
01
modes, the loss is pro-
portional to 1/r
co
, where r
co
is the core radius of the fiber
[22]. Besides the refractive index contrast, modal radi-
ation loss of the Bragg fiber also depends on the number
of cladding layers, N, and scales as D
N
, where D is
a constant whose value depends on the mode being
considered [22].
3.4.2.3.2 Dispersion
Because of the multitude of index profile parameters
involved, Bragg fibers offer fiber designers wide latitude
in the choice of parameters to tailor their dispersion
characteristics for diverse functionality. Judicious choice
of the core radius, cladding layer thickness, and refractive
index contrast in the cladding layers can yie ld dispersion
characteristics that are virtually unique to Bragg fibers
and almost unachievable in conventional single-mode
fibers. For example, single-mode light propagation at
1.06 mm with zero dispersion at a wavelength z 1 mm
through a silica core Bragg fiber was experimentally
demonstrated in 2000 [9]. Subsequently, various designs
have been proposed to achieve multiple zero dispersion
[24] and high negative dispersion for dispersion com-
pensation purposes [24, 25] and single polarization
single-mode propagation [26].
3.4.3 Dispersion compensating
Bragg fiber
Dispersion compensators are an integral part of any long-
haul fiber optic telecommunications link. Conventional
high-silica single-mode fibers exhibit positive temporal
dispersion beyond a wavelength of 1310 nm. Conse-
quently, a signal carrying the LP
01
mode acquires positive
dispersion as it propagates through a conventional single-
mode transmission fiber, which unless canceled through
insertion of a component to combat it would limit signal
transmission capacity of the fiber. Thi s is achieved by
inserting a dispersion compensating module, which in
most cases is a dispersion compensating fiber (DCF; see
Chapter 1). A DCF is normally designed to exhibit a large
negative dispersion coefficient (D) across the signal
wavelength band such that a relatively short length (l
C
)of
the DCF could cancel the dispersion that the signal ac-
quires while propagating through much longer lengths of
the transmission fiber (l
Tx
). Equation (3.4.17) determines
the length of the DCF required for compensating the
dispersion accumulated in a transmission fiber.
D
C
l
C
þ D
Tx
l
Tx
¼ 0; (3.4.17)
where D
C
and D
Tx
are the dispersion coefficients of the
DCF and the transmission fiber, respectively. Besides
high negative dispersion, an efficient DCF must also
exhibit low loss in the operating wavelength range. The
overall efficiency of a DCF is measured in terms of an
integral parameter known as figure of merit (FOM),
which is defined as the ratio of dispersion to loss of the
DCF:
FOM ¼ D=
a
; (3.4.18)
where the loss is measured in dB/km, so that FOM is
expressed in units of ps/nm$dB. The larger the value of
FOM, the more efficient is the DCF. One can achieve
large negative dispersion in conventional silica fibers
through suitable tailoring of its refractive index profile.
The largest dispersion coefficient of 1800 ps/nm km
demonstrated to date was based on a dual-core DCF
design [27, 28]. However, the material loss of silica limits
the FOM that can be achieved, and typically FOM ranges
from w300 to 400 ps/nm$dB in commercially available
high FOM DCFs. Moreover, typical refractive index
profiles required to generate a large waveguide (negative)
dispersion to substantially offset the positive material
dispersion (within the gain spectrum of an erbium-doped
fiber amplifier) are necessarily characterized with a rela-
tively small mode effective area (A
eff
) z15–20 mm
2
,
which makes these DCFs sensitive to detrimental non-
linear effects unless the launched signal power into the
DCF is restricted.
Bragg fibers have evolved as an attractive alternative to
conventional DCFs, and various designs have been pro-
posed for realizing high negative dispersion fibers with
a relatively large A
eff
. One of the earliest designs reported
an estimated D up to approximately 25,000 ps/nm$km
through the hybrid HE
11
mode of a Bragg fiber [25].
However, the hybrid nature and higher loss of HE
11
mode (as compared with TE
01
mode) and small core
radius were major issues that limited the practical use of
this otherwise attractive design. Subsequently, it was
reported that high negative D could be achieved for the
TE
01
mode as well by intentionally incorporating a defect
layer within the periodic cladding layers [24]. The fiber is
so designed that the fundamental (TE
01
) mode of the
defect-free Bragg fiber and the defect mode supported
by the defect layer are in weak resonance, analogous to
phase matching of the sup ermodes in a fiber directional
coupler at a particular wavelength. Near the resonance
wavelength, the fiber exhibits very large negati ve dis-
persion. The thickness and localization of the defect layer
were crucial in determining precise dispersion charac-
teristics of such Bragg fibers. These parameters can be
altered suitably to match the dispersion and dispersion-
slope for achieving broadband dispersion compensation
characteristics. A fundamental design rule for such Bragg
fiber–based DCFs with an intentional defect layer(s) is
that its modal field must penetrate sufficiently into the
160
SECTION THREE Optical Fibers
Get The Optical Communications Reference now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.