Attenuation in an optical Fiber

Attenuation in an optical fiber refers to the reduction in optical power as a light signal propagates through the
fiber. This reduction in power is a key factor in determining the maximum transmission distance between a
transmitter and a receiver or an in-line amplifier in an optical communication system. Attenuation occurs due
to three primary mechanisms: absorption, scattering, and radiative losses.
The optical power P(z) decreases exponentially as the light travels a distance z along the fiber. This is
expressed by the formula:

Where:

• P(0) is the optical power at the origin (at z=0).

• P(z) is the power at a distance z along the fiber.
• αp is the fiber attenuation coefficient in units of km−1, representing the rate of power loss per kilometer.
To simplify the calculation of optical signal attenuation, the attenuation coefficient is often expressed in
decibels per kilometer (dB/km), using the formula:

This parameter is commonly referred to as fiber loss or fiber attenuation, and it is influenced by the factors
discussed above, including absorption, scattering, and radiative losses.
Absorption of Optical Power in Fiber
Absorption in optical fibers refers to the loss of optical power as light travels through the fiber material. This
absorption is caused by the conversion of optical energy into other forms, such as heat, and is due to three
primary mechanisms:
1. Absorption by Atomic Defects
Atomic defects are imperfections in the atomic structure of the fiber material. These defects can include
missing molecules, high-density clusters of atom groups, or oxygen defects in the glass structure. However,
the absorption losses caused by atomic defects are generally negligible compared to other types of absorption.
2. Extrinsic Absorption by Impurity Atoms
The most significant cause of absorption in silica fibers is the presence of impurities in the fiber material,
particularly OH (water) ions and transition metal ions such as iron, copper, chromium, and vanadium.
• Transition Metal Impurities:
In the 1970s, impurity levels of transition metals were around 1 part per million (ppm), resulting in
losses of 1 to 4 dB/km. These impurities cause absorption through electron transitions between energy
levels or charge transitions between ions. Modern fiber production techniques have greatly reduced
these impurities, allowing the fabrication of low-loss fibers.
 • OH Ion Impurities:
OH ions are introduced during the manufacturing process, mainly from the oxyhydrogen flame used
in the reaction of starting materials like SiCl₄, GeCl₄, and POCl₃. Early fibers had high OH
concentrations, leading to large absorption peaks at 725 nm, 950 nm, 1240 nm, and 1380 nm.
To reduce attenuation, water impurity levels below a few parts per billion (ppb) are required. Modern fibers
have very low OH content, resulting in nominal attenuations of less than 0.4 dB/km at 1310 nm (O-band) and
0.25 dB/km at 1550 nm (C-band). Further reduction of OH ions has opened up the E-band for data transmission
by diminishing the absorption peak around 1440 nm. Fibers optimized for use in the E-band are known as
low-water-peak fibers or full-spectrum fibers.
3. Intrinsic Absorption by Fiber Material
Intrinsic absorption is caused by the basic materials used in fiber construction, such as pure SiO₂ (silica). It
represents the fundamental limit of absorption when the material is in a perfect state, with no impurities or
structural variations.
• Electronic Absorption Bands:
In the ultraviolet (UV) region, intrinsic absorption occurs due to electronic transitions when a photon
excites an electron in the valence band to a higher energy level. The absorption in the UV region
follows Urbach’s Rule, expressed as:

Here:
C and E0 are empirical constants.
E is the photon energy.
Since photon energy E is inversely proportional to wavelength λ, UV absorption decreases exponentially with
increasing wavelength.
• Near-Infrared Absorption:
In the near-infrared (NIR) region, intrinsic absorption is caused by atomic vibrations in the glass
material.
The UV absorption loss at any wavelength λ can be further described by the empirical formula:

where x is the mole fraction of GeO₂ in the fiber material.

Scattering Losses in Optical Fibers
Scattering losses occur in optical fibers due to microscopic variations in the material's density, compositional
fluctuations, and structural inhomogeneities that arise during fiber manufacturing. These variations cause light
to scatter as it travels through the fiber, leading to a reduction in optical power.
Glass is made up of a randomly connected network of molecules, which naturally creates regions of varying
molecular density. In addition, since glass consists of multiple oxides such as SiO₂, GeO₂, and P₂O₅,
compositional fluctuations can also occur. These density and compositional variations cause changes in the
refractive index over small distances, which are much smaller than the wavelength of the light. These
variations result in Rayleigh scattering, a type of scattering where light is scattered in different directions. This
is similar to how sunlight scatters in the atmosphere, creating a blue sky.
The mathematical expression for scattering-induced attenuation caused by density fluctuations in a single-
component glass is:

Where:
• n = refractive index of the glass
• kB = Boltzmann’s constant
• Tf = fictive temperature (the temperature at which density fluctuations are frozen into the glass)
• βT = isothermal compressibility of the material
Alternatively, scattering losses can be expressed as:

In this equation, pp represents the photoelastic coefficient. Both equations provide scattering losses in nepers
(base e units). To convert these values to decibels (dB) for practical attenuation calculations, multiply by:
10 log e= 4.343.
This conversion is necessary to express the scattering loss in decibels, which is commonly used in optical
power attenuation measurements.
Fiber Bending Losses
Radiative Losses occur when an optical fiber undergoes bending with a finite radius of curvature. There are
two main types of fiber bending:
Macrobending Bends
These bends occur when the fiber is bent with a large radius relative to its diameter, such as when a fiber cable
turns a corner. The loss due to macrobending is referred to as macrobending loss or simply bending loss.
o For slight bends, the loss is very small and not noticeable.
o As the radius of curvature decreases, the bending loss increases exponentially. Once the
curvature reaches a certain critical radius, the loss becomes more noticeable.
o If the bend radius becomes even smaller beyond this threshold, the bending loss increases
significantly, leading to large optical radiation losses.
The amount of radiated optical power from a bent fiber depends on:
• The field strength outside the fiber core.
• The radius of curvature (R) of the bend.
In multimode fibers, higher-order modes are less tightly bound to the core than lower-order modes. As the
fiber is bent, the higher-order modes are more likely to couple into the cladding region and radiate out of the
fiber. This reduces the total number of modes that can be supported by a curved fiber, compared to a straight
fiber.
The number of modes that can be guided by a bent fiber is given by the effective number of modes Meff, which
is calculated using the following expression:

Where:
• α = graded-index profile.
• Δ= core-cladding index difference.
• n2 = refractive index of the cladding.
• k=2π/λ = wave propagation constant.
• M∞ is the total number of modes in a straight fiber, given by:

Microbends
Microbends are small-scale, repetitive fluctuations in the radius of curvature along the fiber axis. These can
be caused by nonuniformities in the fiber’s manufacturing or by nonuniform lateral pressures during the fiber’s
incorporation into cables. These small bends lead to mode coupling, where energy is transferred between the
guided modes and the nonguided modes, resulting in increased attenuation. Figure 3.4 illustrates such random
microbends along the fiber.

To minimize microbending losses, fibers are often jacketed with a compressible material. This jacket deforms
under external forces, but the fiber itself remains relatively straight. For a graded-index multimode fiber
with core radius aaa, outer radius bbb (excluding the jacket), and index difference Δ, the microbending loss
αM in a jacketed fiber is reduced by the factor F(αM):
Where:
• Ej and Ef are the Young’s moduli of the jacket and fiber, respectively.
• The Young’s moduli of common jacket materials range from 20 to 500 MPa, while that of fused silica
glass (fiber material) is approximately 65 GPa.
By using this jacket, the microbending losses can be significantly reduced, thus improving the fiber’s
performance in terms of attenuation.
Signal Dispersion
Signal dispersion in optical fibers is caused by several factors, including intermodal delay (or modal delay),
intramodal dispersion (chromatic dispersion), polarization-mode dispersion, and higher-order dispersion
effects. These effects arise due to variations in the group velocity of the guided modes, which is the speed at
which energy in a particular mode propagates along the fiber.
Intermodal Delay (Modal Delay)
This effect only occurs in multimode fibers. Modal delay arises because each mode has a different group
velocity at a single frequency. Since different modes travel at different speeds, this causes the light signal to
spread out over time. The degree of modal delay impacts the information-carrying capacity of a multimode
fiber.
Intramodal Dispersion (Chromatic Dispersion)
Intramodal dispersion, also known as group velocity dispersion, occurs within a single mode and results from
pulse spreading. This spreading arises because the spectral emission width of the light source is finite. The
effect increases with the spectral width of the optical source. Sources like light-emitting diodes (LEDs) have
a spectral width of approximately 4–9% of the central wavelength (e.g., as Fig. 3.7 illustrates, if the peak
wavelength of an LED is 850 nm, a typical source spectralwidth would be 36 nm), while laser diodes have
much narrower spectral widths, typically 1–2 nm for multimode lasers and 10⁻⁴ nm for single-mode lasers.

Two main causes of intramodal dispersion are:

Material Dispersion: This type of dispersion arises because the refractive index of the core material
varies with wavelength. Since the refractive index affects the group velocity of light, different
wavelengths in a pulse travel at different speeds, leading to pulse spreading. This phenomenon is
similar to how a prism spreads light into a spectrum, often referred to as chromatic dispersion.
Waveguide Dispersion: This type of dispersion happens because only part of the optical power is
confined to the core of the fiber. Shorter wavelengths are more tightly confined to the core, while
longer wavelengths propagate more in the cladding. Since the refractive index of the cladding is lower
than that of the core, the portion of light traveling in the cladding moves faster. The spatial distribution
 of light between the core and cladding also varies with wavelength, causing each spectral component
to propagate at a different speed, contributing to pulse spreading. Waveguide dispersion is particularly
significant in single-mode fibers, but it can often be neglected in multimode fibers (Fig. 3.8).

Polarization-Mode Dispersion (PMD)

This occurs in single-mode fibers due to the fact that light at a given wavelength actually occupies two
orthogonal polarization modes. Although these polarization states are initially aligned, variations in the fiber's
material cause each polarization mode to experience slightly different refractive indices. As a result, the two
polarization modes travel at different velocities, leading to pulse spreading. This difference in propagation
times between the two polarization modes is known as polarization-mode dispersion.
These various types of dispersion can all contribute to signal degradation in optical fibers, with each type of
dispersion having different impacts depending on the fiber structure and the light source characteristics.
Modal Delay Effects (Intermodal Dispersion)
Intermodal dispersion, or modal delay, only occurs in multimode fibers. It results from each mode having a
different group velocity at a single frequency. This phenomenon causes a distortion of the signal due to the
different speeds at which each mode travels through the fiber. Here's how it works:
In a multimode step-index fiber, rays propagate along the fiber at different angles. The steeper the angle of
propagation, the higher the mode number and consequently, the slower the axial group velocity for that mode.
This variation in the group velocities of the different modes leads to group delay spread, which is the cause of
intermodal dispersion.
The maximum pulse broadening caused by modal delay is the difference between the longest ray path (for the
highest-order mode) and the shortest ray path (for the fundamental mode). The difference in travel times is
referred to as the group delay spread.
To calculate the modal delay, the difference between the travel time Tmax of the longest ray path and the travel
time Tmin of the shortest ray path is given by:

Where:
• n1 is the refractive index of the core,
• c is the speed of light in a vacuum,
• L is the length of the fiber,
• θc is the critical angle.
• Δ is the index difference
Group Delay and Pulse Spreading in Optical Fibers

In digital transmission, signal dispersion is often examined in terms of pulse broadening. An optical source
modulated by an electrical signal excites all modes of the fiber equally. Each mode carries an equal amount
of energy and contains all the spectral components from the source. As the signal propagates, each spectral
component experiences a group delay, which causes pulse spreading.

Group Delay and Pulse Spreading

The group delay per unit length τg/L is the inverse of the group velocity Vg:

The group velocity Vg is the speed at which pulse energy travels along the fiber:

Since the group delay depends on wavelength, each spectral component experiences different delays,
resulting in pulse spreading.

Pulse Spreading and Delay Difference

For a narrow spectral width, the delay difference per wavelength is approximated by dτg/dλ. The total delay
difference δτ\delta\tau over fiber length LL is:

In terms of angular frequency ω:

The second derivative of β\beta with respect to ω\omega is called Group Velocity Dispersion (GVD),
which controls the amount of pulse broadening.

RMS Pulse Spreading

The rms pulse spreading σg relates to the rms spectral width σλ of the source:

This shows the relationship between pulse broadening and the spectral width of the optical source.
 Dispersion

The dispersion parameter D quantifies pulse spreading and is given by:

This parameter is measured in ps/(nm·km) and results from both material and waveguide dispersion.

In practical cases, material and waveguide dispersion are often calculated separately and then added
together. This approach helps estimate the total dispersion, impacting the fiber’s data rate and information-
carrying capacity.

Material-Induced Dispersion
Material dispersion occurs because the refractive index of the core material varies as a function of the optical
wavelength. This is exemplified in Figure 3.9, which shows the refractive index variation for silica. As a
consequence, since the group velocity Vg of a mode depends on the refractive index, the different spectral
components of a given mode will travel at different speeds depending on the wavelength. This leads to material
dispersion, which is an intramodal dispersion effect. It is particularly important for single-mode waveguides
and LED systems because LEDs have broader output spectra than laser diodes.

To calculate material-induced dispersion, consider a plane wave propagating in an infinitely extended

dielectric medium with a refractive index n(λ)n(\lambda) equal to that of the fiber core. The propagation
constant β is given by:

Substituting this into Equation (3.27) with k=2π/λ, the group delay τmat resulting from material dispersion is:

The pulse spread σmat for a source with spectral width σλ can be derived by differentiating the group delay with
respect to wavelength and multiplying by σλ:

where Dmat(λ) is the material dispersion.