CHAPTER 5

OPTICAL FIBER

5.1 Introduction

5.1.1 Introduction to Fibre Optics

Optical communication uses light waves to carry information. There were problems
associated with the source of light and the transmission medium during the early
days of optical communication. Transmission through the atmosphere proved to be
unreliable because signals are subjected to absorption and attenuation from fog, dust,
rain etc.
As a consequence of these atmospheric problems another kind of transmission
medium was needed. The future of optical communication seemed to lie in the suc-
cessful development of the optical fiber. An optical fiber is basically a light waveg-
uide which transmits energy at optical wavelength.
To make optical fibre competitive with existing systems it has to have low attenu-
ation. This allows sufficiently large spacing between the repeaters. Low attenuation
was achieved by the removal of absorbing impurities within glass fibers. Manufac-
turers have managed to reduce losses to such an extent that they compare favourably
with other transmission media, for instance copper wires.

Communication Systems I, First Edition. 95

By Osama A. Alkishriwo Copyright c 2015 John Wiley & Sons, Inc.
96 OPTICAL FIBER

The optical source will convert the electrical signal into light frequencies. The
light will then propagate through the optical fibre cable to the receiver. This con-
sists of optical detectors, for example silicon J photo diodes I which translate the
information back to an electrical signal.

5.1.2 Light Spectrum

Human eye can detect electromagnetic energy in the frequency range of 4.3×1014 Hz
to about 7.5 × 1014 Hz. Optical communication makes use of carrier frequencies
in the ranges of 100, 000 GHz (i.e 1014 Hz). For this it has enormous potential
of bandwidth for example if a carrier of 100, 000 GHz is used, and assuming 10%
bandwidth system then this system can carry 10, 000 GHz of bandwidth which is
equivalent to 2.5 billions of telephone channels or 10, 000 Gbps of data or 156 mil-
lions of PCM telephone channels, or 1.67 million TV programs.
The light wave lengths used in Fiber Optic system include 0.77 − 0.86 µm, and
1.1 − 1.6 µm with the most popular being used 0.82, 1.3, and 1.55 µm (1µm =
10−6 m = 104 nm (nanometer) = 104 Ao (Angstrom)).

5.1.3 Energy in a Light

Amount of energy in a light is given by
hv
u = hf = (5.1)
λ
where, h = 6.626 × 10−34 Joules-Sec is Plank’s constant, v is the velocity of
propagation, and f is the frequency where (1 Joule= 1 W Sec of energy 1 ev =
1.6 × 10−19 Joule).

EXAMPLE 5.1

(a) Determine the energy in a photon of red light (0.82 µm) in free space.
(b) Determine the power for a 0.1 µsec burst of this red light.

SOLUTION

(a)
hv 6.626 × 10−34 × 3 × 108
u= = = 2.42 × 10−19 J/photon
λ 0.82 × 10−6
(b)
u 2.42 × 10−19
P = = = 2.42 pW
t 10−7
 LIGHT PROPAGATION IN GLASS FIBERS 97

5.2 Light Propagation in Glass Fibers

When a single electromagnetic wave is incident on smooth interface separating the

fiber and the surrounding medium of different refractive index as shown in Fig. 5.1.

Figure 5.1 Light propagation in fiber glass.

Using Snell’s law,

n1 sin(θ1 ) = n2 sin(θ2 ) (5.2)
The critical angle of incidence θc , is the smallest angel from normal for which total
reflection occurs, i.e. θc occurs when θ2 = 90o .
 
−1 n2
θc = sin (5.3)
n1
where n is index of refraction given by n = c/vp , where c is velocity of light in
free space vp is phase velocity in the medium. Table 5.1 shows index of refraction
for various materials.

Table 5.1 Index of refraction

Material n
Vacuum 1.0
Water 1.33
Fused Quartz 1.46
Glass 1.5
Diamond 2.0
Silicon 3.4
Gallium Arsenide 3.6

Optical Fiber for communication are typically made of a glass core of n1 = 1.5,
and surrounded with a glass or plastic of slightly lower refractive index of n2 =
 98 OPTICAL FIBER

1.485 is common. This gives the fiber a large value for the critical angle of incidence
(81.9o ).

EXAMPLE 5.2

What is the critical angle beyond which an ideal underwater light source will
not shine into the air above.

Figure 5.2 Critical angle.

SOLUTION

   
n2 1.0
θc = sin−1 = sin−1 = 48.6o
n1 1.33

5.2.1 Numerical Aperture and Reflectance

Numerical aperture is the measure of the maximum angle at which the light rays
can be accepted at the input of the fiber and be totally reflected inside the fiber.
Light entering optical fiber inside the core of acceptance defined by θin (max) will
be totally reflected at the core-cladding interface, and propagate in the fiber core.
The reflectivity of a glass surface is found from r,

Figure 5.3 Cone of acceptance at the input of fiber.

 LIGHT PROPAGATION IN GLASS FIBERS 99

 2
n1 − na
Refracitivity = =r (5.4)
n1 + na

Reflected ray = r × 100% (5.5)

Transmitted ray = (1 − r) × 100% of the light (5.6)

This is for the light arriving at right angles to the face of the fiber.

Figure 5.4 Ray reflection.

EXAMPLE 5.3

n1 = 1.5, na = 1
 2
1.5 − 1 1
then, r = =
1.5 + 1 25
1
the amount of the reflected light = × 100% = 4%
25
i.e. Power Loss = 4%
The amount of the transmitted light = 100 − 4 = 96%

Numerical Aperture (NA) is given by,

NA = sin(θmax ) (5.7)
Using Snell’s Law,
na sin(θmax ) = n1 sin(α) (5.8)
For the reflection of the ray inside the fiber then,
n1 sin(θ1 ) = n2 sin(90) = n2
or n1 sin(90 − α) = n2 ⇒ n1 cos(α) = n2
s  2
n2 n2
cos(α) = , sin(α) = 1 −
n1 n1
100 OPTICAL FIBER

Figure 5.5 Ray reflection in the fiber input.

s  2
n2
q
na sin(θmax ) = n1 1− = n21 − n22 (5.9)
n1

na : refractive index of the input medium

na : 1 for air
n1 : refractive index of the fiber core
n2 : refractive index of the cladding

Figure 5.6 Power loss at the facing fibers.

• Effect of NA on Power Loss

NAs = sending end NA

NAr = receiving end NA

The power Loss due to numerical aperture NA is giving by,

 
NAr
NALoss = −20 log (5.10)
NAs
 OPTICAL FIBRE WAVEGUIDE 101

Figure 5.7

If NAr is smaller when coupling from a small diameter LED to a larger diameter
step index fiber with NAf the loss will be,

NAf = −20 log (NAf ) (5.11)

• Effect of Fibre Cross Section on the Power Loss

If As is the light source cross section and Af is fiber cross section, then the total
power is proportional to the ratio of areas,
Af
Pf = Ps , where Af < As
AS
 2
Df
or Pf = Ps
Ds

 
Df
Loss = −20 log (5.12)
Ds

where Df and Ds are fiber and source diameters, respectively.

Loss = 0 dB, if Af > As (5.13)

Thus large fibers with large NAs are efficient. However, large NA and diameter
can result in large mode dispersion in the fiber waveguide.

5.3 Optical Fibre Waveguide

5.3.1 Basics
The fiber waveguide with a core of constant index n1 surrounded a cladding material
of lower index n2 is called step index fiber waveguide. The index profile is shown.
The modes of propagation are the particular electric and magnetic field configura-
tions which satisfy the necessary conditions of propagation on the a waveguide. The
condition is that the phase of twice reflected wave is the same as incident wave. If
the diameter D of the core is small enough (on the order of one or two wavelengths),
then only the single mode will propagate in the waveguide. The number of modes
102 OPTICAL FIBER

Figure 5.8 Graded and step index fibers.

that can exist in a step index fiber waveguide of core D has been determined by nu-
merical methods, and can be found from the plot shown in Fig. 5.9. The horizontal
axis is written in forms of a normalized frequency or normalized wavelength from
fiber parameters.

Figure 5.9 Optical fiber waveguide propagation modes for a step–index fiber.

πD NA
V = (5.14)
λ
where λ free space wavelength of the light the plot assumes typical fiber waveg-
uide that has a small percentage of difference in n, with respect to n1 . This index
difference is given by,
n1 − n2
∆= , and for typical optical fiber
n1
√
NA = n1 2∆ (5.15)

It is shown from the Fig. 5.9 that for single mode propagation V ≤ 2.4, other-
wise multimode operation exists. Multimode operation has advantage of propagat-
ing more power than does single mode. However pulse spreading due to the modal
dispersion limits the pulse and information rate.
 OPTICAL FIBRE WAVEGUIDE 103

EXAMPLE 5.4

A Fiber optic waveguide of a core of index 1.5 and cladding 1.485.

(a) How many modes will propagate if the core diameter is 4 µm and 820 nm light
source is used?
(b) What is the maximum core diameter for single mode propagation?

SOLUTION

p
NA = (1.5)2 − (1.485)2 = 0.21

(a)
4π × 0.21
V = = 3.2
0.82
From Fig. 5.9 above, 4 modes will propagate ( HE11 , T E01 , T M01 , HE21 )
(b)
2.4 × 0.82
Dmax = = 2.98 µm
0.21π

5.3.2 Types of Optical Fiber Cables

Optical fibers are classified by the way the light propagates. One type of classifica-
tion is a single mode or multimode type. Multimode fibers are divided further into
step index and graded index types. In terms of material used in the fiber they are
grouped into three types,
All glass fiber.
Mixed (glass and plastic) fibers.
All plastic fibers.
The following table gives the types of optical fiber cables.
The following are observed from the table,
Single mode fiber are superior in the propagation characteristics.
Within the multimode fibers graded index is better suited in the long distance
transmission (with a wider bandwidth) than step index.
All glass fibers have the highest performance in optical quality than either all
plastic or mixed type fibers.
104 OPTICAL FIBER

Table 5.2: Types of optical fiber cables.

Parameters
Optical Fiber Type Core Numerical Loss Bandwidth Characteristics
Diameter Aperture (dB/km) (MHz . km)
(𝝁𝒎)
Multimode Low loss
step index 50-100 0.18-0.25 2-6 20-60 Relatively low
cost
Multimode Low loss wide
All glass graded 50-100 0.2-0.25 2-6 200-2000 bandwidth
fiber index
Low loss
Single 5-10 0.1 0.5-1.5 More than extremely wide
mode 10,000 bandwidth
Multimode 50-150 0.2-0.3 5-20 20-60 Large core
step index diameter,
Mixed Multimode higher NA
fiber graded 50-150 0.2-0.3 5-20 200-1000 value, lower
index cost than all
glass fiber
Large core
All plastic 100-3000 diameter,
fiber Step index 100-300 0.3-0.6 (visible) Less than 10 higher NA
value, least
expensive

The disadvantage of all glass fibers, it has the highest cost.

The plastic fiber has the lowest performance with the most economical price.

The difference between multimode and single mode is summarized in the Table
5.3.2.

Table 5.3: Single mode and multimode parameters

Parameters Multimode fiber Single mode fiber
Transmission Loss About the same About the same
characteristics Bandwidth per 1 km ~ 1 𝐺𝐻𝑧 ~ 100 𝐺𝐻𝑧
𝜆 = 1 𝑛𝑚
Distance dependency 𝛼 𝐿−0.4 𝛼 𝐿−1
Connection Easy Difficult
(step index connector)
LD driver coupling Easy Difficult
Micro bending (Fiber wire, cabling) < 0.1 dB Less critical than
multimode
Analog transmission Problem Not much a problem

Micro bending refers to a displacement effect of light in a few micrometer range

occurring sharp curvatures. As optical fiber are placed inside the protective
jacket to take the form of cable, micro bending effect takes place inside the
cable. Micro bending is the major cause of energy loss in the cable.
 LOSES 105

5.4 Loses

5.4.1 Loss in optical fiber cable

There are several mechanisms causing loss of light energy during its transmission
through an optical fiber. We will look more closely at,

Scattering

– Rayleigh scattering
– Bend losses

Absorption

– Intrinsic
– Impurity

For a material to be considered as a possible optical fiber it has to be transparent to

the propagation of light. As with copper wires, an optical fiber cable should attenuate
the signal as little as possible. During the early days of optical fiber development,
experts suggested that an overall attenuation level of less than 20 dB/km would be
needed before optical fibers could be of use in (trunk) telecommunications. This
level was attained in 1970, making optical communication competitive with existing
systems.
As the light travels through a medium, its energy decreases because of losses
in the material. These losses are due to a number of effects, two being considered,
scattering and absorption. These effects are dependant on the wavelength. Scattering
losses are due to the redirection of light rays and absorption losses are caused by the
conversion of light energy into the energy at other frequencies, usually heat.

1. Scattering
Light energy that hits obstructions scatters in all directions and does not con-
tinue in the original direction of propagation. This light energy is lost. Obstruc-
tions are typically variations in the density of optical materials. By it’s nature,
glass is a disordered structure in which there are variations around the average
refractive index.

2. Rayleigh scattering
When obstructions are small, less than the wavelength, the scattering is called
Rayleigh scattering. Losses from Rayleigh scattering are related to λ−4 so that
the losses decrease to the fourth power of λ. When obstructions are larger than
the wavelength, the scattering effect is called Mie scattering. Present manufac-
turing processes make the Mie Scattering insignificant.

3. Fiber bend losses

Light rays could also be scattered into the cladding because of bending of the
fiber. There are two types of bends; macrobends and microbends. Macrobends
106 OPTICAL FIBER

are introduced during installation if the fiber is bent too sharply in cable ducts
in order to negotiate corners. They cause some of the light not to be internally
reflected but to propagate into the cladding and be lost.
4. Macrobends
Losses due to macrobends are based on the fact that the rays of light do not
propagate at an angle. They are spread over a range of angles, each of which
is steeper than the critical angle. All those rays travelling at a distinct angle are
called a mode. Due to the bending of the fiber, the angle of incidence on the
core/cladding interface is slightly changed so that some modes might no longer
be propagated.
5. Microbends
Microbends are very minute irregularities in core size or imperfections in the
core/cladding interface. They cause the light to leave the core. Microbending
can result from fibers being wound too tightly when forming a multifiber cable,
or when it is wound onto a drum for transportation. The effect is the same; some
modes might not be confined to the core and are lost.
6. Absorption
Some of the light energy transmitted along the fiber is lost not only due to
scattering, but by conversion into thermal energy and will be dissipated by con-
duction and radiation. The cause of absorption in glass is the interaction of light
waves with the molecular structure of the material or with impurities within it.
7. Intrinsic absorption
High frequencies, such as ultra violet light waves, are able to excite electrons
forming the lattice of the glass and set them in resonance. This means vibration
which gives rise to excessive absorption and therefore to a lower limit for ultra
violet light. It occurs at wavelengths below 1 µm. This absorption mechanism
is called electron absorption. The attenuation limit for higher wavelengths is
caused by lattice vibrations of the atoms themselves. These vibrations occur at
the wavelengths of infra red light and are the cause of material absorption. The
infra red limit becomes significant at wavelengths longer than 1.5 µm.
8. Impurity absorption
Another reason for absorption is impurities in the glass. Impurity absorption is
highly dependent on the wavelength of the impurity. This means that impurities
behave like resonance circuits with sharp peaks, attenuating light at specific
wavelengths e.g.
– Chromium: 0.625 µm
– Vanadium: 0.725 µm
– Copper: 0.850 µm
Still the major reason for impurity absorption is hydroxyl or OH ions. The
absorption peaks lie at the following values, 1.38, 0.92, and 0.72 µm. as shown
in Fig. 5.10.
 LOSES 107

The attenuation of specific wavelengths results in low loss windows. A window

is a range of wavelengths over which attenuation is a minimum, thus favourable
for transmission.

Figure 5.10 Total attenuation.

EXAMPLE 5.5

Using the attenuation graph below, determine the bandwidth available in the
fiber assuming maximum permutable attenuation of 1 dB/km. Also calculate the
number of simultaneous television channels that could be transmitted, assuming
a bandwidth of 5 M Hz for each TV channel.

SOLUTION
From the graph the transmission windows are between, 1.02 − 1.16 µm, 1.27 −
1.351 µm, and 1.45 − 1.68 µm. Using c = f × λ, this corresponds to frequency
windows, 2.94 − 2.59 × 1014 Hz, 2.36 − 2.2 × 1014 Hz, and 2.07 − 1.79 ×
1014 Hz. This is a bandwidth of (0.35 + 0.16 + 0.28) × 1014 Hz = 7.9 ×
 108 OPTICAL FIBER

1013 Hz. If each television channel bandwidth is 5 M Hz then,

7.9 × 1013
= 15.8 × 106 , 1.e. > 15 million TV channels
5 × 106
This is more than the most dedicated channel flickers could achieve in a lifetime.
Note This is a theoretical value and there many practical difficulties involved in
trying to reach this figure.

5.5 Sources of Optical Signal Loss

1. Air gap loss

– Reflectance loss obtained from

 2
n1 − na
Reflectivity r =
n1 + na
Reflectance loss = −10 log(1 − r)

na =for air, typical values is 0.18 dB loss.

– End separation loss

D/2
Les =
D/2 + ` × tan sin−1 NA


n

where,
D: Fiber diameter
L: Separation
n: Media index of refraction

2. Core diameter difference

 
NAr
NAloss = −20 log
NAs

If receiving NAr is smaller than sending NAs when coupling from a small
diameter LED to a larger diameter step index fiber with NAf the loss will be
NAloss = −20 log(NAf ).
When coupling from small LED to a larger diameter step index fiber with NAf ,
 
Df
Loss = −20 log
Ds
 BANDWIDTH–DISTANCE PRODUCT 109

Df is fiber diameter, Ds is light source ray diameter, and D1 < D3 , and

Loss= 0 dBs if Df ≥ Ds .
3. Axial misalignment or radial displacement
If the displacement is as much as 15% of the smaller core diameter. The loss
will be about 1 dB. Displacement of 5% will keep the loss below 0.3 dB.
4. Angular misalignment
The formula is very complex, it is a function of NA, θ, n medium for typical
fiber loss= 0.5 dB.
5. End surface roughness
The ends of connecting fibers must be highly polished (less than 0.5 µm ) if
scattering losses are to be avoided.

5.6 Bandwidth–Distance Product

Large numerical aperture and large diameter fibers will allow more light power to
propagate, these two factors allow more useful light to be coupled into the fiber, and
a large diameter allows more modes to be existed into the fiber. Step index fibers
with large diameters are very limited.
Various distortion effects in step index fibers primarily intra modal dispersion,
limit the bandwidth–distance product to about 20 M Hz km. This means that a
10 km fiber optic link with no repeaters will have only 2 M Hz of bandwidths.
Small diameter fibers with single mode propagation can have a bandwidth dis-
tance product three order of magnitude greater so that only power limits the repeater
less–link distance i.e. for step index,
(BW × distance)Single mode = 1000 (BW × distance)Multi mode (5.16)

A special fabrication process that produce a graded index core can yield bandwidth–
distance product in excess of 2 GHz km ( two order of magnitude over step index
intermode), and yet allow the higher power carrying capability of large diameter
multimode operation.
(BW × distance)Graded index = 100 (BW × distance)Step index (5.17)

5.6.1 Dispersion and Pulse Spreading

The performance of an optical fiber is reduced not only by losses, but also by dis-
persion. The term dispersion is used to describe the pulse broadening effect of a
transmission medium. This means that a pulse that appears at the output of a fiber is
wider than the input pulse. As a pulse of light travels along the fiber it will become
wider. There are two dispersion mechanisms, modal and chromatic dispersion.
1. Modal dispersion
Modal dispersion occurs due to the propagation of light in modes. Any mode
110 OPTICAL FIBER

has its typical angle and thus its own path of transmission through the fiber. A
mode propagating at angle zero mode would take minimum time to reach the
output. The maximum time would occur when a mode travels at the critical
mode.
The longer the fiber the bigger is the difference between the path of both modes
causing the dispersion to increase per unit length. During transmission through
the whole of the fiber, each single mode will be delayed by a particular amount
of time. The difference between the shortest and the longest transmission delay
is the time by which the pulse had been broadened. The broadening effect
increases not only with the length of the fiber, but also with the acceptance
angle as the span between the two extreme modes (zero and critical mode) gets
larger. This means, that a large acceptance angle is connected with high modal
dispersion.
This broadening could continue to such an extent that single signal pulses inter-
fere with each other. Eventually two successive pulses could merge and become
indistinguishable. This effect restricts the highest data rate applicable to a cer-
tain type of fiber.
The dispersion of a fiber is stated as the longest possible delay per kilometer
and is a measure of quality of the fiber. An example for the dispersion of a high
performance fiber is 10 ps/km.
The dispersion rises with the length of the fiber. Longer fibers result in more
pulse spreading, which leads to lower bandwidth. Therefore the data rate is
also length dependent. In conclusion, a certain data rate can only be applied
for a certain length of a fiber. After covering this distance repeaters have to
regenerate the signal. Thus, bandwidths are stated in respect to a unit length, as
the product of bandwidth and length, e.g. 500 M Hz.km. The increasing of the
fiber length is only possible at the expense of usable bandwidth.
2. Chromatic dispersion
Chromatic dispersion results from the fact that the light in the fiber consists of a
range of spectral components, i.e. of several frequencies. Each special compo-
nent travels at a slightly different speed, making delays between them possible.
Again, any delay during transmission produces pulse broadening. Therefore, it
is better to use a laser as the light source rather than a LED. The coherent light
transmitted by a user consists only of a very limited range of frequencies, whilst
the LED is broadband.
3. Total dispersion
The total dispersion is calculated from the modal and chromatic dispersion by
following equation,
q
∆ttotal = ∆t2modal + ∆t2chrom (5.18)

As a mathematical example we can state,

If ∆tmodal = 25 ns/km, and ∆tchrom = 20 ns/km, then
 BANDWIDTH–DISTANCE PRODUCT 111

p
∆ttotal = (25 ns/km)2 + (20 ns/km)2 = 32 ns/km

4. Pulse spread as function of refractive index

For a given digital data pulse, the pulse of light at the receiver, due to the low
order modes, extinguishes at the end of the pulse period Tb but the light from
the highest order modes is still arriving at the detector end finally extinguishes
τps seconds later. The amount of pulse spreading can be determined by tracing
the light rays for each mode and calculating the propagation time over that path.
The result of such analysis yields a pulse spread of,

τps = n1 L ∆/c (5.19)

where,
n1 : is the refractive index of the cable core
∆ = (n1 − n2 )/n1
L: cable length
c: velocity of light

The pulse spread per unit length is

τps (n1 − n2 )
= (5.20)
L c

5. Maximum allowable data rate

The maximum data rate for a digital optical fiber link may be determined from
the pulse spread per unit length τps /L. Thus the maximum bit rate is,

(
1/τps , for a narrow RZ format
rb max = (5.21)
1/(τ + τps ), for awide RZ format of bit duration τ )

For NRZ data

Tb /2 + τps ≤ Tb ⇒ τps ≤ Tb /2

1 1 2
rb max = = = (5.22)
Tb + τps Tb + Tb /2 3Tb

EXAMPLE 5.6

A 10 km long, 50 µm diameter glass fiber with n1 = 1.5, n2 = 1.485 is to

transmit RZ data pulses. Determine the pulse spreading per unit length and the
maximum allowable data rate.
 112 OPTICAL FIBER

SOLUTION
This inside diameter will have many modes propagating. Thus pulse spread per
unit length is

τps n1 ∆ n1 − n2 0.015
= = = = 50 ns/km
L c c 3 × 105
Maximum data rate for 10 km link with narrow RZ pulses is
1 1
rb = = = 2 M bps
τps 50 × 10 × ×10−9

Figure 5.11 Maximum allowable data rate.

5.7 Power Budget

Energy per bit Eb = Pb τb , where Pb is the power per bit (pulse), and τb is the
effective pulse (bit) duration. Since the power is most conveniently measured with
an average reading photometer, we determine the power per bit from:
 
τb
Pb = P (5.23)
Tb

where P is the average power, and τb /Tb is the duty cycle of the pulse bit stream as
shown in Fig. 5.12.

Figure 5.12 Pulse Bit Stream.

The losses are,

Lt : is the power loss at the transmitter to fiber interface.
 POWER BUDGET 113

Figure 5.13 Pulse Bit Stream.

Lr : is the power loss at the fiber to receiver interface.

where,

Lt (dBs) = Reflected loss + NAloss + Area difference loss + Air gap loss
+Misalignment losses (5.24)

and the same applies for Lr .

Lc is the fiber to fiber coupling loss, and it is given by,

Lc (dBs) = Reflectance loss + Air gap loss + Misalignment losses (5.25)

These are usually given by the connector main fractures, but if not a 2 dB is a rea-
sonable assumption. Lf is the fiber loss and it is highly frequency dependent, and
produces the greatest loss. A dB/km is obtained from the fiber attenuation charac-
teristics at the operating frequency. The received power (detected light power) PD
is,

m
X m+1
X
PD = Pt − Lt − Lf i − Lci − Lr − Mg (5.26)
i=1 i=1

where m is the number of fiber sections and Mg is the system margin.

EXAMPLE 5.7

A repeater less fiber optic communication link is to be 10 km long. The LED

produces 8 mW at 0.82 µm has as emitting area with diameter of 80 µm and it
is closely coupled to a short fiber pigtail. Assume NA = 0.3 step index fibers
of 1 km Length with 2 dB/km, NA=0.3, core diameter 50 µm, and n = 1.5 are
used. Connectors each with 2 dB of loss from a mechanisms are to be used.
The receiver uses a ferrite connector with a very short fiber, and diode with an
antireflection surface and an input diameter of 100 µm and NA = 0.5. The
air gap is essentially zero, assume zero system margin. Determine the detected
light power.

SOLUTION
 114 OPTICAL FIBER

   
NAf Df
Lt = −20 log − 20 log
NAs Ds
 
50
= −20 log(NAf ) − 20 log
80
= 10.457 + 4.082 = 14.539 dB
   
NAr Dr
Lr = −20 log − 20 log + reflectance loss
NAf Df
 
NAr
NAr > NAf , then − 20 log =0
NAf
 
Dr
Dr > Df , then − 20 log =0
Df
"  2 #
1.5 − 1
reflectance loss = −10 log 1 − = 0.18 dB
1.5 + 1
⇒ Lr = 0.18 dB
10 km
Number of fiber sections = = 10
1 km
10
X
Lf i = 2 × 10 = 20 dB
i=1
Number of connectores is 10 + 1 = 11
11
X
Lci = 2 × 11 = 22 dB
i=1
PD = Pt − 14.5 − 20 − 22 − 0.18 = Pt − 56.68
Pt = 10 log(8) = 9 dBm
PD = 9 − 56.68 = −47.7 dBm

5.8 Fiber Transmitter and Receiver Devices

5.8.1 Fiber Transmitter Devices

The light source for fiber optic communication that are most commonly used are
solid state devices. The most important devices are
Light emitting diode (LED).
The injection laser diode (ILD) or (LD)
Neodymium :yttrium Aluminum Garner (Nd: yAG) solid state laser
All of these three light sources can be intensity modulated by varying the input device
current. The LED devices produces non coherent emission, where is the laser output
 FIBER TRANSMITTER AND RECEIVER DEVICES 115

waves are coherent Consequently lasers have good power in a narrow special width
and can be used for single mode operation with small diameter step index fibers.
The optical power versus the bias current of the diode is shown below for LED and
injection laser diode (ILD).

Figure 5.14 The optical power versus the bias current.

The modulation of transmitter is performed utilizing the linearity of the LED and
ILD sections of the power curve. The bias current level can be pulsed for digital
amplitude shift keying ASK or on-off keying (OOK). For analog AM, the current
level is varied from a fixed bias level.

5.8.2 Fiber Receiver Devices

The most important devices used for detecting light In fiber optic system are,

Positive intrinsic negative doped (PIN) photo diode.

The avalanche photo diode (APD).

The photo transistor.

In general silicon photo diodes are used for small size, high sensitivity, fast re-
sponse time, and low noise. The receiver devices transform the light wave into elec-
trical signal which is then signal processed. Optical fiber receivers consist of the
photo diode or phototransistor and operational amplifiers to increase the current gain
and provide a compatible circuit output level.

5.8.3 Optical Fiber System Repeater

There are two types of repeaters,

1. Electrical Repeaters:
In this type the optical wave is transformed to electrical signal by (O/E) device
then this electric signal is amplified or regenerated then applied to the electrical
to optical (E/O) device and applied to the fiber as shown in Fig. 5.15.
116 OPTICAL FIBER

Figure 5.15 Electrical repeater.

2. Optical Repeaters:
In this type the optical wave coming in the fiber is amplified or regenerated
directly in optical form without transforming it to electrical.

Figure 5.16 Optical repeater.

5.8.4 Wave Division Multiplexing (WDM)

The concept of wave division multiplexing is the same as frequency division mul-
tiplexing (FDM), where is information signals m1 (t), m2 (t), · · ·, mn (t) each is
modulated by a light wave with specific wave length At λ1 , λ2 , λ3 , · · ·, λn . An re-
spectively then these light waves are combined and applied to the optical fiber cable
to be transmitted to the receiving side where they are demultiplexed as shown in Fig.
5.17.

Figure 5.17 Wave division multiplexing.

The shown WDM system is in one direction using one fiber.

5.8.5 Examples of Typical Systems

Typical repeater less system
 FIBER TRANSMITTER AND RECEIVER DEVICES 117

Figure 5.18 Modulated signals.

The cable used is graded index type consisting of six fibers.

The bit rates are 2.048, 8.448, and 34.368 M bps.
The length is 7.6 km.
It is repeater less.

The parameters of these systems are shown in Table 5.2 below.

Table 5.2 System characteristics.

Bit rate (Mbps) 2.048 8.448 34.368

Permissible line att. (dBs) 53.5 47.5 41.5
Approx. permissible repeater section length (km) 11 9.5 8
Transmit power mean value (dBm) −2 −2 −2
Min. receiver input signal (dBm) −63
Max Prob of error 10−10 10−10 10−10
DC supply voltage (V) 60/48 60/48 60/48
Power consumption of terminals (W) 16 14 25

Typical repeater spacing as compared with coaxial cable is shown in Table 5.3.

Table 5.3 Typical repeater spacing.

Transmission Repeater spacing (km)
rate
Coaxial cable 9.5 mm Mono mode fiber 1.3 µm
34 M bps 9.3 km 50 km
140 M bps 4.65 km 39 km
565 M bps 2 km 27 km
1200 M bps 1.4 km 21 km