FIBER OPTICS

Prof. R.K. Shevgaonkar

Department of Electrical Engineering

Indian Institute of Technology, Bombay

Lecture: 3
Propagation of Light in an Optical Fiber

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 1
 Light energy can be modelled in three different forms which relate the
particular model of light to the context in which it is talked about. Light can be
characterized in any one of the following models

Ray Model
Wave Model
Quantum Model

In the simplest possible context, light is treated as a ray and the different
phenomena exhibited by light are explained in terms of the ray-model of light. Some
phenomena exhibited by light are not adequately explained by the Ray-Model of
light. In that case, we resort to the more advanced nature of light such as the wave
and the quantum models. In this section we shall mainly deal around the ray model
of light and attempt to explain the propagation of light in an optical fiber treating light
as a ray.

Constructionally, an optical fiber is a solid cylindrical glass rod called the core,
through which light in the form of optical signals propagates. This rod is surrounded
by another coaxial cylindrical shell made of glass of lower refractive index called the
cladding. This basic arrangement that guides light over long distances is shown in
figure 2.5.

Fig. 3.1: Constructional Details of an Optical Fiber

The diameter of the cladding is of the order of 125 m and the diameter of the
core is even smaller than that. Thus it is a very fine and brittle glass rod that we are
dealing with. In order to provide mechanical strength to this core-cladding
arrangement, other coaxial surrounding called the buffer coating and jacketing layers
are provided. They do not play any role in the propagation of light through the optical
fiber, but are present solely for providing mechanical strength and support to the
fiber.

The light energy in the form of optical signals propagates inside the core-
cladding arrangement and throughout the length of the fiber by a phenomenon called
the Total Internal Reflection (TIR) of light. This phenomenon occurs only when the
refractive index of core is greater than the refractive index of cladding and so the
cladding is made from glass of lower refractive index. By multiple total internal

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 2
reflections at the core-cladding interface the light propagates throughout the fiber
over very long distances with low attenuation. We shall now discuss the essential
requirements of the propagation of light through an optical fiber, over long distances
with minimum loss, in detail.

Figure 3.2 shows a section of the core of an optical fibre. If a ray of light is
incident on the core of an optical fibre from the side, the ray of light simply refracts
out from the fibre on the other side. The ray shown in figure 3.2(in green)
demonstrates the situation.

Figure 3.2: Launching of light into an optical fiber.

No matter what the angle of incidence of the light is, any light that enters the
fiber from the side does not propagate along the fiber. The only option thus available
with us is to launch the light through the tip of the fiber. That is, in order to guide light
along the fiber, the light must be incident from the tip of the optical fiber. The red ray
of light in figure 3.2 explains this situation. In other words, if the tip of the optical fiber
is not exposed to light, no light will enter the fiber. Although there may be ambient
light, as long as the tip is protected, no light from the sides propagates along the
fiber. Equivalently, if there was propagation of light through the fiber, no light would
emerge from the sides of the fiber. This characteristic of the optical fiber imparts the
advantage of information security to the Optical Fiber Communication Technology.

At this juncture, one basic question that may come to the readers mind is that
whether a partial reflection at the core-cladding interface suffices the propagation of
light along the fiber over long distances? The answer to this question is very clearly a
no. The reason is that, at each reflection a part of the optical energy launched into
the optical fiber would be lost and after a certain distance along the length of the
fiber the optical power would be negligibly low to be of any use. Thus total internal
reflection is an absolute necessity at each reflection for a sustained propagation of
optical energy over long distance along the optical fiber. This precisely is the sole
reason of launching light into the fiber at particular angles so that light energy
propagates along the fiber by multiple total internal reflections at the core-cladding
interface.

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 3
 We have already stated that for explaining propagation of light in an optical
fiber, the Ray-Model of light shall be used. The Ray-Model of light obeys the Snells
laws. Following figure depicts a situation of a typical refraction phenomenon taking
place at the interface of two optically different media having refractive indices n 1 and
n2:

Figure 3.3: Refraction of light at a media interface

The angles measured in the expression for Snells law are measured with
respect to the normal to the media interface at the point of incidence. If n2 > n1 , then
the angle of refraction is greater than the angle of incidence and the refracted ray is
said to have moved away from the normal. If the angle of incidence (1) is increased
further, the angle of refraction (2) also increases in accordance with the Snells law
and at a particular angle of incidence the angle of refraction becomes 90 o and the
refracted ray grazes along the media interface. This angle of incidence is called the
critical angle of incidence (c) of medium 2 with respect to medium 1. One should
note here that critical angle is media-relative. That means, the same optically denser
medium may have different critical angles with respect to different optically rarer
media. If 1 is increased beyond the critical angle, there exists no refracted ray and
the incident light ray is then reflected back into the same medium. This phenomenon
is called the total internal reflection of light. The word total signifies that the entire
light energy that was incident on the media interface is reflected back into the same
medium. Total Internal Reflection (TIR) obeys the laws of reflection of light. This
phenomenon shows that light energy can be made to remain confined in the same
medium when the angle of incidence is greater than the angle of reflection. Thus we
can see that there are two basic requirements for a TIR to occur:

1. The medium from which light is incident, must be optically denser than the
medium to which it is incident. In figure 3.3 n2 > n1.
2. The angle of incidence in the denser medium must be greater than the
critical angle of the denser medium with respect to the rarer medium.

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 4
LAUNCHING OF LIGHT INTO AN OPTICAL FIBER
Light propagates inside an optical fiber by virtue of multiple TIRs at the core-
cladding interface. The refractive index of the core glass is greater than that of the
cladding. This meets the first condition for a TIR. All the light energy that is launched
into the optical fiber through its tip does not get guided along the fiber. Only those
light rays propagate through the fiber which are launched into the fiber at such an
angle that the refracted ray inside the core of the optical fiber is incident on the core-
cladding interface at an angle greater than the critical angle of the core with respect
to the cladding. But before delving into rigorous mathematical calculations, let us first
visualise how light energy can be launched into a fiber. Figure 3.4 shows one of the
possibilities of launching light into an optical fiber where the light ray lies in a plane
containing the axis of the optical fiber. Such planes which contain the fiber axis are
called meridional-planes and consequently the rays lying in a meridional-plane are
called meridional-rays. Meridional rays always remain in the respective meridional
plane.

Figure 3.4: Launching of Meridional Rays

There may be infinite number of planes that pass through the axis of the fiber
and consequently there are an infinite number of meridional planes. This indirectly
indicates that there are an infinite number of meridional rays too, which are incident
on the tip of the fiber making an angle with the fiber-axis as shown in the above
figure. These meridional rays which get totally internally reflected at the core-
cladding boundary meet again at the axis of the optical fiber as shown in the figure
3.5 below. In the figure the meridional plane is the plane of the paper which passes
through the axis of the fiber and the incident rays, refracted rays and the reflected
rays lie on the plane of the paper. Though only two rays are shown in the figure for
the sake of clarity, in practice there would be a bunch of rays that would be

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 5
convergent at the same point. Meridional rays are classified into bound and unbound
rays. The rays that undergo TIR inside the fiber core remain inside the core at all
times along the propagation and are called as bound rays. The rays that fail to
undergo TIR inside the core are lost into the cladding and are called unbound rays.
The dotted ray shown in figure 3.4 is an unbound meridional ray.

Figure 3.5: Meridional Rays meeting at the axis.

Since all the reflected rays meet at the same point a region of high optical
intensity is generated at that point (point A in figure 3.5). Since these rays undergo
multiple TIR at the core-cladding boundary, they meet repeatedly at the axis at
regular intervals along the fiber. This causes multiple regions of maximum intensity
along the axis of the fiber. Also, different sets incident rays would have different
angles of reflection at the core-cladding boundary and consequently have different
points of maximum intensities along the fiber axis. Thus it can be visualized easily
that at the output end of the optical fiber, maximum intensity will be in the axial
region of the fiber core and the intensity would gradually decrease as we move away
from the axis towards the periphery of the core.

Another way of launching a light ray into an optical fiber is to launch it in such
a way that it does not lie in any meridional plane. These rays are called skew rays. A
pictorial representation of launching a skew ray is shown in the figure 3.6 below.

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 6
 Figure 3.6: Launching of Skew Rays

Skew rays propagate without passing through the central axis of the fiber. In
fact the skew rays go on spiralling around the axis of the optical fiber. The light
energy carried by them is effectively confined to an annular region around the axis
as shown in figure 3.6. Consequently, at the output, skew rays will have minimum
energy at the axis of the optical fiber and it will gradually increase towards the
periphery of the core.

Thus when light energy is launched into an optical fiber, there arises two
possible energy distributions; one, which has maximum intensity at the axis due to
meridional rays and the other, which has minimum intensity at the axis due to the
skew rays. Thus, on the whole, there are two ways of launching light into an optical
fiber; light can be launched either as meridional or as skew rays.

Assuming that light is launched as meridional rays into the optical fiber, let us
now carry out a simple analysis. For that let us concentrate on figure 3.7 below. The
figure shows a cross-section of an optical fiber with a core of refractive index n 1 and
a cladding of refractive index n2. The incident ray AO (shown by dotted line) is
incident at an angle with the axis of the fibre. The refracted ray for AO in the core
(dotted line ON1) fails to be incident on the core-cladding interface at angle greater
or equal to the critical angle of the core w.r.t. cladding and hence refracts out of the
core and is lost to the cladding. In other words, the angle of incidence of a refracted
ray at the core-cladding interface in turn depends on the initial angle at which the
incoming ray was launched into the fiber. If this launching angle (with the fiber axis)
is decreased, the angle of incidence which the refracted ray makes at the core-
cladding interface increases. If this increase is such, as to exceed the critical angle
of the core-cladding interface, then total internal reflection of the refracted ray takes
place and the light remains in the core and is guided along the fiber. The ray CO is
launched into the fiber at such an angle that its refracted ray is incident at the
core-cladding boundary at its critical angle c. If any light ray is launched at an

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 7
angle more than then the refracted ray just refracts out to the cladding because the
angle of incidence of its refracted ray at the core-cladding interface is less than the
critical angle. Thus the angle is indicative of the maximum possible angle of
launching of a light ray that is accepted by the fiber. Consequently, the angle is
called the angle of acceptance of the fiber core. Since the optical fiber is symmetrical
about its axis, it is very clear that all the launched rays, which make an angle with

Figure 3.7: Launching of Light into an Optical Fiber

the axis, considered together, form a sort of a cone. This cone is called the
acceptance cone of the fiber as shown in the above figure. Any launched ray that lies
within this cone is accepted by the fiber and the light of this ray is guided along the
fiber by virtue of multiple TIRs as shown by the red ray BO in the figure 3.7.

NUMERICAL APERTURE OF OPTICAL FIBER

With the same initial assumption of meridional launching of light into an optical
fiber, let us consider the figure 3.8 below. The figure 3.8 shows a cross-section of a
core of refractive index n1 and a cladding of refractive index n2 that surrounds the
core glass. An incident ray AO is incident from medium1 at the tip of the fiber making
an angle with the axis of the fiber, which is the acceptance angle of the fibre. The
refracted ray for this incident ray in the core then is incident at the core-cladding
interface at the critical angle c of the core with respect to the cladding. The angle of
refraction for critical angle of incidence is 900 and the refracted ray thus grazes along
the core-cladding boundary along BC as shown in the figure 3.8. According Snells
laws, the incident and the refracted rays lie in the same meridional plane, which is

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 8
the plane of the paper in this case. Applying Snells law at the medium1-core
interface we get:

(3.1)

Figure 3.8

From the figure it is clear that, and so substituting this in

equation (3.1), we get:

(3.2)

From the basic trigonometric ratios,

(3.3)

Applying Snells law at the core-cladding interface we get:

( ) (3.4)

Substituting equation (3.4) in equation (3.2) we get:

Since the initial medium 1 from which the light is launched is air most of the
times, n = 1. The angle is indicative of light accepting capability of the optical fiber.

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 9
Greater the value of , more is the light accepted by the optical fiber. In other words,
the optical fiber acts as some kind of aperture that accepts only some amount of the
total light energy incident on it. The light accepting efficiency of this aperture is thus
indicated by sin and hence this quantity is called as the numerical aperture (N.A.)
of the optical fiber. Thus for an optical fiber in air, with core refractive index n 1 and
cladding refractive index n2 and having an acceptance angle of is given by

(3.5)

Numerical Aperture is one of the most fundamental quantities of an optical

fiber. It indicates the light collecting efficiency of an optical fiber. More the value of
N.A. better is the fiber. For greater values of N.A. the difference on the right hand
side of equation 3.5 has to be maximized. For maximizing the difference, either the
refractive index of the core (n1) has to be increased or the refractive index of the
cladding (n2) has to be reduced. Since the core used is always glass, the value of its
refractive index n1 is thus fixed (approximately 1.5). The only option thus available
with us is to reduce the value of n2. But it too has a limitation of the lowest value of 1
for air because till date no material is known which has a refractive index lower than
that. If we make n2 =1, we would then get the maximum possible N.A. for an optical
fiber. But then we are basically talking about removing the cladding because, if there
is a cladding, the value of n2 will always be greater than 1. Thus one can clearly say
that from the point of view of light accepting efficiency, the presence of a cladding is
undesirable.

The above discussion suggests that although the optical fiber is made of core
and cladding, the presence of cladding is undesirable because it reduces the light
accepting efficiency of the optical fiber. However, with a deep thought, one can
realise that the prime concern behind prolonged research on optical fibers was not
just to put light inside an optical fiber with the best efficiency but also to propagate
the light over long distances with the least attenuation. That means if we have a
source of optical signal and an optical fiber with the highest light accepting efficiency
but is incapable of propagating the accepted light; the optical fiber is of no use in
spite of its high N.A. Thus judging the need of a cladding just on the basis of light
launching efficiency would be highly inappropriate. In other words, light launching
efficiency is just one of the key characteristic aspects of an optical fiber. There are
other attributes too which have to be given importance while determining the quality
of an optical fiber. One of such attributes of an optical fiber is its bandwidth. Large
bandwidths are desirable for high data rates of transmission.

When optical fiber is used for transmission of information, light signal

launched into it cannot be of continuous nature. For a carrier signal to carry
information, one or more of its characteristics has to be altered in accordance to the
data signal. In an optical fiber light is launched in the form of optical pulses to

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 10
transmit the required information. Light energy launched into the fiber may be
considered to travel in the form of numerous rays in accordance to the Ray-Model.
These rays travel different paths inside the core of an optical fiber because different
light rays are incident on the tip of the optical fiber at different angles within the
acceptance cone itself. This causes different light rays in the acceptance cone to
travel along different paths in the core of the optical fiber and accordingly take
different time intervals to travel a given distance too, which leads to a phenomenon
of pulse broadening inside the core of the optical fiber. Thus the pulse of light which
might originally be of width T seconds now might be of T+T seconds inside the fiber
core. The figure 3.9 below depicts a pictorial description of how light pulse broadens
inside the core of the fiber.

Figure 3.9: Pulse-Broadening inside optical fiber core

Any incident ray that lies within the acceptance cone gets guided inside the
optical fiber by virtue of multiple total internal reflections. Since the angle of refraction
different incident rays are different, they travel along different paths in the optical
fiber as shown in the above figure. This causes the initially launched narrow light
pulse to broaden as shown. The amount of broadening is measured in terms of the
increase in the pulse time width and is denoted by T. the value of T is given by:

(3.6)

Where, T= Pulse Broadening; c = velocity of light in free space; n1 =

refractive index of core and n2 = refractive index of the cladding.

The quantity L is the horizontal distance travelled before suffering the first
total internal reflection by the refracted ray OB which corresponds to the incident ray
AO, incident at the acceptance angle as shown in the figure. The amount of pulse

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 11
broadening is effectively the difference in time of travel between the ray travelling
along the axis and the incident ray AO. This pulse broadening effect signifies that if a
second pulse is now launched into the fiber within the time interval T+T, the two
pulses will overlap and no identifiable data would be obtained on the output. Thus for
a given length L, there would be a corresponding value of T (from equation 3.6)
which would limit the rate at which light pulses can be launched into the optical fiber.
In other words, it limits the rate at which data can be transmitted along the fiber. This
indirectly limits the bandwidth available on the fiber. Thus we can say that more the
pulse broadening lower the bandwidth. That is:

(3.7)

Equation 3.7 suggests that for higher bandwidth of transmission the pulse
broadening, T should be as low as possible. In equation 3.6, we see that the value
of T is dependent on the value of L, the difference (n 1 n2) as well as the value of
n1/n2. But reducing the value of L would signify the reduction in the length of the
optical fiber, which is not desirable. As 1<n2<n1, the ratio, n1 / n2 is very close to 1.
Thus for low T values, the only option available with us is to decrease the value
(n1n2) or in other words, to increase the refractive index of the cladding n2. One can
now notice that a contradictory situation has been generated as to whether the
cladding should be removed for high NA or to use a cladding of large refractive index
value for higher bandwidth? The answer to this query is purely application specific.
That means if an optical fiber is used as a sensor (say), where lowest possible light
has to be accepted, we use fiber with low n2 values. When the optical fiber is used
for data communication, fibers with high values of n 2 are used. For practical
communication purposes the value of (n1 n2) is made of the order of about 10-3 to
10-4. If the cladding is removed, the value of n 2 becomes 1 and the value of the
above difference becomes about 0.5. The bandwidth corresponding to this value of
n1-n2 is of the order of few Kilohertz, which is far worse than that of a normal twisted
pair of wires. Thus cladding is an extremely important requirement for optical fiber
when the bandwidth is the prime concern of the application and its refractive index is
made as close to that of the core as the available technology permits, but not made
equal. This is brought about by varying the amount of doping in a single glass rod.
The differently doped regions have different refractive indices and serve as core and
cladding of the optical fiber.

PHASE-FRONT (WAVE-FRONT) BASED STUDY OF TIR

Let us now have a study of the phenomenon of total internal reflection at the
core-cladding interface on a backdrop of the wave-fronts of the incident and the
reflected light. Wave-fronts are nothing but the constant phase planes of the light
wave and are also called as phase-fronts. They are perpendicular to the direction of
propagation of the wave at every point. Any light ray launched meridional within the
acceptance cone will propagate along the fiber core by virtue of multiple total internal

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 12
reflections at the core-cladding interface. Thus in accordance to the ray-model of
light we may visualise a solid cone of light (having angle = double the acceptance
angle) that enters an optical fiber and propagates through the fiber by TIRs. Figure
3.10 below shows the phenomenon of total internal reflection of a ray of light at the
core-cladding boundary along with the wave-fronts of the incident and the reflected
rays. The red and green coloured dotted lines represent the wave fronts of the light
rays which are perpendicular to their direction of propagation. The light rays,
actually, are fictitious lines which, in reality, represent the direction of propagation of
these wave-fronts.

Figure 3.10: Total Internal Reflection of Light inside a fiber core.

The distance between a red and a green wave-front corresponds to a phase

difference of 1800 ( radians). The similar coloured wave-fronts have either 00 or
3600 phase difference between them. Thus, when two similar coloured wave-fronts
meet, they interfere constructively and dissimilar coloured wave-fronts interfere
destructively. This is evident from the interference pattern that sets up in the core as
shown in the above figure. In the core, the interference between the incident and the
reflected wave-fronts constitutes a standing wave pattern of varying light intensity
with discrete maxima and minima in a direction normal to the core-cladding interface.

Total internal reflection is also accompanied by an abrupt phase change

between the incident and the reflected rays at the core-cladding boundary. This
phase change depends on the angle of incidence of the incident ray at the core-
cladding boundary, the refractive index or the core and cladding and various other
parameters.

If we refer to the electromagnetic wave theory of light, it shows that at total

internal reflection, the light intensity inside the cladding is not completely zero.
Instead, there exist some decaying fields in the cladding, which do not carry any
power but support the total internal reflection phenomenon by satisfying the

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 13
boundary conditions at the core-cladding interface. These fields are called as
evanescent fields. The Ray-model of light does not offer any explanation about the
evanescent fields, which indeed are as equally important as the fields in the core for
total internal reflection to occur. The importance of these evanescent fields in the TIR
can be clearly ascertained from the fact that even the slightest disturbance to these
fields in the cladding could lead to the failure of the TIR at the core-cladding
boundary accompanied by leakage of optical power to the cladding. This is one of
the instances when the ray-model of light becomes inadequate in explaining the
phenomena exhibited by light. Though the evanescent fields are decaying fields,
they never become zero, atleast theoretically. In other words, they remain present
upto infinite distance from the core-cladding boundary. But in practice, these fields
decay down to a negligibly small value as we move away from the core-cladding
boundary deeper into the cladding. Larger the value of the angle of incidence of the
incident ray at the core-cladding boundary, sharper is the decay of the evanescent
fields. Thus there must me a sufficient thickness of cladding provided for these
evanescent fields to be accommodated so that they decay to a negligibly small value
in the cladding and cannot be disturbed by external sources.

With this backdrop of knowledge about the propagation of light in an optical

fiber, let us now look into certain finer aspects of such propagation and investigate
the condition required for successful propagation of light in the fiber. Figure 3.11
below shows two parallel rays that are launched into an optical fiber and they
propagate as shown. The dotted lines represent the wave-fronts of the rays. The
refractive indices of core and cladding are n1 and n2 respectively. The diameter of the

Figure 3.11: Propagation of Light rays in an Optical fiber

core is d. The phase-front AE is common to both Ray 1 and Ray 2. The phase-front
DB is common to Ray 2 and BF. The Ray 2 is thus common to both the phase-fronts.
Hence for a sustained constructive interference, the distance between these two
phase-fronts must be multiples of 2In other words, it can be said that the phase
difference between the phase change undergone by Ray 1 in travelling distance s 1
and the Ray 2 in travelling s2 must be 0 or integral multiples of 2

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 14
 Mathematically,

(3.8)

( )
(3.9)

If is the phase change undergone in each TIR of Ray 1, then the total phase
change undergone by Ray 1 in travelling s1 is given by

(3.10)

Where n1 = refractive index of core; Wavelength of the light in the core.

The phase change undergone by Ray 2 in travelling s2 is given by

(3.11)

For a sustained constructive interference, both 1 and 2 must have a phase

difference of either 0 or integral multiples of 2. That is, for an integer m (=0,1,2,3,)
the following condition must be satisfied:

(3.12)

The significance of the equation 3.12 is that only those rays, which are
incident on the tip of the fiber at angles such that their angle of refraction in the core
satisfies equation (3.12), can successfully travel along the fiber. If we concentrate
on equation (3.12), we find that since m can take only discrete integral values, the
value of angle is also discrete. This suggests that there are only some discrete
launching angles within the acceptance cone (N.A. cone) for which the rays can
propagate inside the fiber core. A 3D visualisation reveals the significance of this
observation, i.e. the acceptance cone can no longer be assumed as a solid cone of
rays, launched at all possible angles (smaller than acceptance angle), but has to be
viewed as composed of discrete annular conical rings of rays which are launched at
the tip of the fiber core at angles which satisfy equation (3.12). Thus the condition
that the launching angle of the incident ray should be within the acceptance cone is
necessary but not sufficient. This angle has to be such that the equation (3.12) is

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 15
satisfied. Thus light can only be launched at certain discrete angles within the N.A.
cone leading to a further decrease in the light gathering efficiency of the optical fiber.
Any ray that is not launched at these discrete angles will not propagate inside the
optical fiber. This discretization in the values of launching angles lead to formation of
what are called as modes in an optical fiber, which are nothing but different patterns
of light intensity distribution around the axis of the core. In the subsequent sections,
we shall see the modal propagation of light in an optical fiber, in detail, and perform
some rigorous mathematical analysis of the light propagation in an optical fiber.

Fiber Optics, Prof. R.K. Shevgaonkar, Dept. of Electrical Engineering, IIT Bombay Page 16