1.

Optical Spectral Band with Operating Windows 🌈

(a) Electromagnetic spectrum & optical band
Light is a part of the electromagnetic (EM) spectrum, which ranges from high-frequency gamma rays to
low-frequency radio waves. The region relevant to optical communication is:
 Visible light: 400–700 nm (violet to red)
 Near-infrared (IR): ~700–1700 nm
Our eyes respond only to 400–700 nm, but optical fibers use IR wavelengths (mainly 850, 1310 and 1550
nm).
OFC-PPT [Link] (1)
(b) Why infrared is used instead of visible?
The attenuation (loss) in glass fibers is due to:
1. Intrinsic absorption – due to atomic vibrations of silica and electronic transitions.
2. Extrinsic absorption – due to impurities such as OH¯ ions (water content).
3. Rayleigh scattering – due to microscopic density fluctuations in glass; behaves as ∝ 1/λ⁴ (shorter
wavelengths scatter more).
If you look at the attenuation vs wavelength curves in the PDF (pages 28–30), attenuation is much higher in
the visible range and strongly decreases in the near-IR region, with a few peaks (water bands).
OFC-PPT [Link] (1)
So:
 At short wavelengths (visible): strong Rayleigh scattering → high loss.
 At too long wavelengths (>1.7 µm): strong vibrational absorption and thermal background → high
loss.
 In near-IR windows (0.8–1.6 µm): scattering and absorption are both relatively low → minimum
loss.
Hence, the optical spectral band used for fiber communication lies in near-IR, not the visible.

(c) Optical operating windows

From the attenuation curve (page 25, 29), three main “valleys” appear, known as optical windows.
OFC-PPT [Link] (1)
1. First window – near 850 nm
o Attenuation ~ 2–3 dB/km.
o Historically used with inexpensive LEDs.
o Mainly for short-distance LANs and low-cost links.
2. Second window – around 1310 nm
o Attenuation ~ 0.4–0.5 dB/km.
 o Chromatic dispersion of silica crosses zero near 1310 nm.
o Very attractive for systems where dispersion is more critical than loss.
o Used in metropolitan / access networks.
3. Third window – around 1550 nm
o Lowest attenuation ~ 0.2 dB/km in pure silica.
o Slightly higher dispersion than 1310 nm, but can be managed by dispersion-shifted fibers or
compensation modules.
o Compatible with Erbium Doped Fiber Amplifiers (EDFAs) → used in long-haul and
undersea cables.
Why not go beyond 1.6 µm?
At longer wavelengths:
 Contribution of molecular vibrational absorption increases.
 Thermal background radiation increases (blackbody emission), raising noise.
So the practical optical spectral band for silica fibers is roughly 0.8–1.6 µm, with the three operating
windows at 850 nm, 1310 nm, and 1550 nm.
In exams, for 10–15 marks you should:
 Sketch the attenuation vs wavelength curve showing three windows.
 Label first, second, third windows and mention typical attenuation.
 Explain why IR region is chosen (low scattering and absorption).

2. Ray Theory of Transmission – TIR, Acceptance Angle, NA, Skew Rays 🔦

Ray theory uses geometrical optics to describe light propagation, good for understanding multimode fibers.
OFC-PPT [Link] (1)
(a) Total Internal Reflection (TIR)
Consider a step-index fiber:
 Core refractive index = n₁
 Cladding refractive index = n₂ (n₁ > n₂)
When light in the core strikes the core-cladding boundary at an angle i (measured w.r.t. normal), Snell’s law
is:
𝑛 sin 𝑖 = 𝑛 sin 𝑟

For a ray trying to leave the core into the cladding:

 As i increases, r increases.
 When r = 90°, refracted ray grazes along the boundary. The corresponding i = φ_c (critical angle).
So:
 𝑛
sin 𝜙 =
𝑛

For i > φ_c, sin r > 1 is impossible → there is no refracted ray, and all light is reflected back into the core:
Total Internal Reflection.
This TIR at every bounce along the core keeps light confined and guides it along the fiber (see diagrams
page 13–14).
OFC-PPT [Link] (1)

(b) Acceptance angle (θₐ) & acceptance cone

At the input end of the fiber, light is usually launched from air (n₀ ≈ 1) into the core. Only a certain range of
input angles will eventually produce internal angles larger than φc and hence will be guided.
By simple geometry and Snell’s law at the air–core surface and core–cladding interface (shown conceptually
on pages 54–57), one can derive:
OFC-PPT [Link] (1)

sin 𝜃 = 𝑛 −𝑛

where θₐ is the maximum angle (with fiber axis) for which light entering the fiber will be guided.
All rays entering within a cone of half-angle θₐ around the fiber axis are accepted. This 3D cone is called the
acceptance cone.

(c) Numerical Aperture (NA)

From the above expression, the Numerical Aperture is defined as:
𝑁𝐴 = 𝑛 sin 𝜃 (for medium of refractive index 𝑛 )

For launching from air (n₀ ≈ 1):

𝑁𝐴 = sin 𝜃 = 𝑛 −𝑛

This parameter measures the light-gathering power of the fiber:

 Larger NA → larger acceptance cone → easier to couple light from LEDs/lasers.
 Typical values: SMF: 0.12–0.20; MMF: 0.20–0.50.
OFC-PPT [Link] (1)
NA can also be expressed in terms of relative refractive index difference Δ:
𝑛 −𝑛 𝑛 −𝑛
Δ= ≈
𝑛 2𝑛
𝑁𝐴 ≈ 𝑛 √2Δ
This form is very useful in design.

(d) Skew Rays

In the simple ray picture we often draw meridional rays – those that pass through the fiber axis on each
reflection plane. But real fibers also support skew rays, which:
 Travel in a helical path around the axis.
 Do not intersect the core axis.
 Reflect from the core–cladding boundary at oblique azimuthal angles.
Because skew rays follow longer paths than corresponding meridional rays, they arrive later at the end of
the fiber, contributing to modal dispersion in multimode fibers.
In exam answers, briefly sketch meridional vs skew ray and state:
“Skew rays never cross the fiber axis and follow a spiral (helical) path, causing longer transit times.”

3. Electromagnetic Mode Theory for Optical Propagation 📡

Ray theory is an approximation. For accurate analysis (especially in single-mode fibers), we must treat light
as an electromagnetic wave that obeys Maxwell’s equations.
(a) Wave equation in cylindrical coordinates
In a homogeneous medium of refractive index n, the electric field E satisfies:
∇ 𝐸+𝑘 𝑛 𝐸 =0

where k = 2π/λ is the free-space propagation constant.

A cylindrical fiber has a core (radius a, index n₁) and cladding (index n₂). Using cylindrical co-ordinates (r,
φ, z) and solving the wave equation separately in core and cladding, we get field solutions involving Bessel
functions in the core and modified Bessel functions in the cladding.
OFC-PPT [Link] (1)
 Core: Jₗ(Ur/a)
 Cladding: Kₗ(Wr/a)
Here U and W are eigenvalue parameters related to β (propagation constant) and refractive indices (page
68).
OFC-PPT [Link] (1)
(b) Propagation constant and guidance condition
The longitudinal propagation constant β satisfies:
𝑛 𝑘<𝛽<𝑛 𝑘

 If β is in this range → guided mode (power mainly in core).

 If β < n₂k → oscillatory in cladding → leaky/radiation mode (page 65).
OFC-PPT [Link] (1)
(c) Mode classification (TE, TM, HE, EH)
Applying boundary conditions (continuity of tangential E and H at r = a) leads to allowed modes:
 TE₀m: E field has no longitudinal component; Ez = 0.
 TM₀m: H field has no longitudinal component; Hz = 0.
 Hybrid modes HEₗm and EHₗm: both Ez and Hz ≠ 0; typical in fibers (page 44–45).
OFC-PPT [Link] (1)
The fundamental mode in a practical circular step-index fiber is HE₁₁, which is nearly linearly polarized.
(d) Weakly guiding approximation & LP modes
In communication fibers Δ is very small (<1%) – weakly guiding. Under this condition:
 Hybrid modes come in nearly degenerate pairs with almost same β.
 These pairs can be combined into Linearly Polarized (LP) modes such as LP₀₁, LP₁₁, etc. (pages
46–48).
OFC-PPT [Link] (1)
These LP modes are easier to visualize and are widely used in design & specifications (e.g., “single-mode
fiber supports only LP₀₁”).

4. Modes in a Planar Guide 📏

Planar waveguide = thin film of core (n₁) sandwiched between two claddings (n₂). Thickness is 2d say. It’s
simpler to analyze than cylindrical fiber.
(a) Ray picture
 Light reflects at both interfaces via TIR, forming a zig-zag path.
 Only certain incidence angles lead to constructive interference of reflected waves in the core →
allowed guided modes.
Condition for constructive interference:
2𝑛 𝑑cos 𝜃 = 𝑚𝜆(𝑚 = 0,1,2, … )

or forms expressed in terms of β. Here θ is the angle inside the core. This quantization leads to discrete
modes TE₀, TE₁,… (see page 81–83).
OFC-PPT [Link] (1)
(b) TEₘ and TMₘ modes
 TEₘ modes: Electric field transverse (no Ez); Ez = 0.
 TMₘ modes: Magnetic field transverse (Hz = 0).
Mode number m is equal to the number of nodes (zero crossings) in the field pattern across thickness. The
lowest mode (m=0) has no node.
Planar analysis builds intuition:
  As thickness increases, more modes satisfy the condition.
 In a very thin film (below a certain cutoff thickness), only one mode (m=0) exists → analogous to
single-mode fiber.

5. Phase and Group Velocity 🚀

(a) Phase velocity vₚ
For a wave travelling along z with field:
( )
𝐸(𝑧, 𝑡) = 𝐸 𝑒

A surface of constant phase corresponds to ωt − βz = constant. Differentiating w.r.t. time:

𝑑𝑧 𝜔
𝑣 = =
𝑑𝑡 𝛽

In a homogeneous nondispersive medium,

𝑐
𝑣 =
𝑛

(b) Group velocity v_g

Real optical pulses are not pure single-frequency waves – they are superposition of frequencies around ω₀.
The envelope of the wave packet moves at group velocity:
𝑑𝜔
𝑣 =
𝑑𝛽

This is the velocity at which energy and information propagate.

(c) Group index and dispersion
Define group index:
𝑐
𝑁 =
𝑣

In dispersive media, β(ω) is nonlinear. Expand β around ω₀:

1
𝛽(𝜔) = 𝛽 + (𝜔 − 𝜔 )𝛽 + (𝜔 − 𝜔 ) 𝛽 + ⋯
2

where:
 β₁ = dβ/dω = 1/v_g → gives group velocity
 β₂ = d²β/dω² → group velocity dispersion (GVD), main cause of pulse broadening in single-mode
fibers.
So, understanding phase and group velocity is essential to analyze dispersion and maximum bit rates of
fiber-optic links.
OFC-PPT [Link] (1)

6. Phase Shift with Total Internal Reflection 📐

When a plane wave undergoes TIR at a dielectric interface (core–cladding), reflection coefficient is not just
a real number −1 (like a mirror) but complex; therefore:
 Reflected wave undergoes a phase change (different for TE and TM polarization).
For an interface with n₁ > n₂ and incident angle i > φc, the Fresnel reflection coefficient for TE polarization
is:

cos 𝑖 − 𝑗 sin 𝑖 − (𝑛 /𝑛 )
𝑟 = =𝑒
cos 𝑖 + 𝑗 sin 𝑖 − (𝑛 /𝑛 )

Similarly for TM:

𝑛 cos 𝑖 − 𝑗𝑛 sin 𝑖 − (𝑛 /𝑛 )
𝑟 = =𝑒
𝑛 cos 𝑖 + 𝑗𝑛 sin 𝑖 − (𝑛 /𝑛 )

The magnitudes |r| = 1 (no power loss) but phase angles φ_TE, φ_TM are nonzero and depend on i and
refractive indices.
Consequences inside fiber:
 After many reflections, cumulative phase changes govern the allowed propagation modes (standing
wave condition).
 The difference between φ_TE and φ_TM leads to phenomena like Goos–Haenchen shift and
polarization effects in fibers (see phase change discussion on pages 84–85).
OFC-PPT [Link] (1)
Qualitatively, for exam:
 State that under TIR, reflected wave acquires a phase shift varying with angle and polarization.
 This phase shift is crucial in solving waveguide modes and explains lateral shift of beams.

7. Evanescent Field 🌊
As covered briefly, but now more structured.
During TIR, although no net power passes into the cladding, the electromagnetic field does not vanish
abruptly at the boundary. Instead, in the lower-index medium (cladding), you get:
( )
𝐸(𝑥, 𝑧, 𝑡) = 𝐸 𝑒 𝑒

where:
 x: distance normal to interface into cladding
 α: attenuation constant (related to how fast the field decays)
  β: propagation constant along z
This evanescent field:
 Decays exponentially with distance into cladding.
 Carries energy along z but no net energy flux in x-direction.
 Represents the “fringing” or “penetration” of the guided mode into cladding (see pages 63–64 & 86).
OFC-PPT [Link] (1)
Practical implications:
1. Cladding material choice
o Must be transparent at operating wavelengths: otherwise evanescent field suffers absorption
→ attenuation.
2. Cladding thickness
o Must be thick enough that evanescent field decays almost to zero before reaching outer
surface; otherwise external disturbances cause loss or mode coupling.
3. Evanescent-field devices
o Used in fiber sensors (refractive index sensing, chemical sensors).
o Used in couplers and tapers, where evanescent fields of closely spaced fibers overlap and
exchange energy.

8. Goos–Hänchen Shift ↔
Described in the PDF (pages 84–85).
OFC-PPT [Link] (1)
When a finite-width light beam (not an infinite plane wave) undergoes TIR at an interface, the reflected
beam behaves as if it were reflected from a point slightly inside the lower-index medium, leading to a small
lateral shift along the interface between incident and reflected beam footprints.
 This shift is called the Goos–Hänchen (GH) shift.
 It arises because the evanescent field penetrates into the cladding and re-radiates; the phase
variations result in effective displacement.
Mathematically, shift δ is related to derivative of phase with respect to incidence angle:
𝜆 𝑑𝜙
𝛿≈−
2𝜋 𝑑𝜃

Typical values are very small (~0.06–0.1 μm for glass–air interface at visible wavelengths). The PDF
stresses that though small, this concept provides important insight into guidance mechanism of dielectric
waveguides.
OFC-PPT [Link] (1)
Exam points:
 Define GH shift.
  Explain qualitatively why it occurs (penetration + phase change).
 State that it’s small but conceptually important for waveguide optics.

9. Cylindrical Fiber Modes 🌀

Now, specifically for cylindrical fibers (our real fibers).
(a) Geometry and field solutions
 Core radius = a, index = n₁
 Cladding index = n₂
Solving Maxwell’s equations in cylindrical coordinates gives fields:
 Core (r < a): Bessel functions Jₗ(Ur/a)
 Cladding (r > a): Modified Bessel functions Kₗ(Wr/a)
U and W are related to β (as in PDF page 68).
OFC-PPT [Link] (1)
Boundary conditions at r = a lead to characteristic equations whose solutions give allowed modes and
corresponding β.
(b) Hybrid & LP mode mapping
Based on symmetry and polarization, modes are labeled:
 TE₀m, TM₀m
 HEₗm, EHₗm
For weakly guiding fibers (Δ≪1), these combine into LP modes (page 46–47).
OFC-PPT [Link] (1)
Table on page 47 shows mapping, e.g.:
 LP₀₁ ↔ HE₁₁
 LP₁₁ ↔ HE₂₁, TE₀₁, TM₀₁
 LP₂₁ ↔ HE₃₁, EH₁₁
… etc.
The PDF also shows allowed LP modes vs normalized frequency V (page 48).
OFC-PPT [Link] (1)
(c) Role of normalized frequency V
2𝜋𝑎
𝑉= 𝑁𝐴
𝜆

 V < 2.405 → only LP₀₁ (fundamental) allowed → single-mode fiber.

 As V increases, more LP modes appear (LP₁₁ at V = 2.405, LP₂₁ around 3.83, etc.).
Thus, cylindrical fiber modes are controlled by core radius, NA, and wavelength via V. This is crucial for
designing single-mode vs multimode fibers.

10. Mode Coupling 🔁

Mode coupling is treated in PDF page 79.
OFC-PPT [Link] (1)
(a) What is mode coupling?
An ideal perfectly uniform fiber would let each mode propagate independently, maintaining its power. Real
fibers, however, have imperfections and perturbations:
 Core radius variations
 Random microbends and macro-bends
 Refractive index fluctuations
 Roughness at the core–cladding boundary
These perturbations cause energy to be transferred from one guided mode to another as the signal travels
→ mode coupling.
(b) Types of perturbations (from figures on page 79)
OFC-PPT [Link] (1)
1. Irregularities at core–cladding boundary
2. Local bends / twists in fiber
3. External stress or pressure
These effectively act as small scattering centers that redistribute power among modes.
(c) Effects of mode coupling
1. Modal noise reduction (sometimes beneficial):
o In multimode fibers with strong coupling, power tends to become evenly distributed among
modes → equilibrium mode distribution → reduces effect of input coupling conditions.
2. Increased attenuation:
o Coupling to leaky or radiation modes leads to power loss.
3. Change in dispersion:
o Mode coupling causes energy originally launched in faster modes to couple to slower modes
and vice versa, affecting pulse shape and broadening.
4. Speckle pattern variations in coherent detection.
In exam:
 Define mode coupling.
 List key causes with sketches (bends, interface irregularities).
 Explain its impact on losses and dispersion.
11. Step Index Fibers vs Graded Index Fibers 📊
This is a very common 10–15 marks question.
(a) Step Index (SI) fiber
Refractive index profile:
𝑛 𝑟<𝑎
𝑛(𝑟) = {𝑛 𝑟≥𝑎

 Abrupt “step” at core radius a (page 67).

OFC-PPT [Link] (1)
 Rays in a multimode SI fiber follow long zig-zag paths; higher-order modes take longer paths →
strong modal dispersion.
Number of guided modes for multimode SI:
𝑉
𝑀 ≈
2

(page 70–71).
OFC-PPT [Link] (1)
(b) Graded Index (GI) fiber
Here the core index decreases gradually from center to edge (page 73).
OFC-PPT [Link] (1)
Typical profile:
𝑟 /
𝑛(𝑟) = 𝑛 [1 − 2Δ( ) ] ,𝑟 ≤ 𝑎
𝑎

and n = n₂ beyond core.

 α is profile parameter:
o α → ∞ → step index
o α = 2 → parabolic profile, optimum for minimum modal dispersion (page 74).
OFC-PPT [Link] (1)
In GI fiber:
 Rays near axis see higher index & travel slower.
 Rays near edge see lower index & travel faster.
→ These two effects compensate so that different modes have nearly equal transit times → greatly
reduced modal dispersion.
Approximate number of modes:
𝑉
𝑀 ≈
𝛼+2

For parabolic (α = 2),

𝑉
𝑀 =
4

i.e., half the number in corresponding SI fiber (page 75).

OFC-PPT [Link] (1)
(c) Comparison table (write this in exam)
Feature Step Index MMF Graded Index MMF
RI profile Abrupt step Smooth graded (often parabolic)
Modal dispersion High Much lower
Bandwidth Low High (up to Gb/s over km)
No. of modes ~V²/2 ~V²/4 (α=2)
Fabrication Easier Slightly more complex
Cost Lower Somewhat higher
Applications Short links, sensors LANs, MANs, medium-haul links
Also mention Single Mode Step Index as special SI fiber designed with small core such that V ≤ 2.405.

12. Single-Mode Fibers – Cutoff Wavelength, MFD & Spot Size 🎯

(a) Single-mode condition & cutoff wavelength
For a step-index fiber, the V-number is:
2𝜋𝑎
𝑉= 𝑁𝐴
𝜆

The fundamental LP₀₁ mode exists for all V > 0, but the next mode (LP₁₁) appears at V = 2.405 (first zero
of J₀).
To ensure single-mode operation, we require:
𝑉 < 2.405

Rearrange for λ:
2𝜋𝑎 𝑁𝐴
𝜆>𝜆 =
2.405
This λ_c is the cutoff wavelength; for λ > λc only fundamental mode propagates, making the fiber single-
mode (see discussion around V and cutoff on pages 48, 76–77).
OFC-PPT [Link] (1)
Designers choose core radius a and NA so that operating wavelength (e.g. 1310 or 1550 nm) is above λc but
not much larger – to balance coupling loss and bending loss.

(b) Mode Field Diameter (MFD)

In single-mode fibers, light is not confined strictly within the physical core radius a; instead, the
fundamental mode’s intensity extends into the cladding (due to evanescent field). Thus we define an
effective mode field radius w or mode field diameter (MFD = 2w).
PDF mentions this as a key parameter to match sources and detectors (page 87).
The fundamental mode intensity roughly follows a Gaussian:
𝑟
𝐼(𝑟) = 𝐼 exp (−2 )
𝑤

MFD depends on V. An approximate empirical formula (for telecom fibers) is:

𝑤 1.619 2.879
≈ 0.65 + + (1.5 < 𝑉 < 2.4)
𝑎 𝑉 / 𝑉

(You don’t need to memorize entire expression; just know that w increases as V decreases).
Importance:
 Coupling efficiency: Laser spot and fiber MFD should match.
 Splicing loss: Mismatch in MFD between two fibers causes loss.
 Bending loss: Larger MFD → more power in cladding → higher sensitivity to bends.

(c) Spot size

Spot size is closely related to MFD – it refers to radius at which optical power falls to 1/e² (or similar) of
central value. Some authors use “spot size” and “MFD” interchangeably; standards typically quote MFD at
1310 and 1550 nm (e.g. 9 µm–10.5 µm).
Design trade-offs:
 Smaller core / smaller MFD → better confinement, lower bending loss but more difficult coupling
from laser/detector (needs precise alignment).
 Larger MFD → easier coupling, but more bending and microbend loss.
In exam answers, emphasize:
 Cutoff wavelength formula and condition V < 2.405.
 Definition and physical meaning of MFD and spot size.
 Their importance in coupling and bending loss.