Design of Passive Optical Network

BRNO UNIVERSITY OF TECHNOLOGY
FACULTY OF ELECTRICAL ENGINEERING

AND COMMUNICATION
DEPARTMENT OF TELECOMMUNICATIONS
DESIGN OF PASSIVE OPTICAL NETWORK

MASTERS THESIS

AUTHOR Juan Salvador Asensi Pla

SUPERVISOR Ing. Pavel Reichert

BRNO, 2011

DESIGN OF PASSIVE OPTICAL NETWORK

Juan Salvador Asensi Pla

Tutor: Ing. Pavel Reichert

UNIVERSIDAD POLITCNICA DE VALENCIA
Mayo 2011

4

Brno University of Technology
Faculty of Electrical Engineering and Communication
Department of Telecommunications

Design of Passive Optical Network

Masters Thesis

Specialization of study: Communications

Author: Juan Salvador Asensi Pla

Supervisor: Ing. Pavel Reichert

ABSTRACT

The steady increase in the demand for broadband services and the consequent
increase in the volume of generated traffic in our communication networks have
motivated the need to implement next generation networks in our territories. This
master's thesis aims to explain the design and planning of a passive optical network
(PON). The main idea of this project is to build a fictitious environment that will allow
us to study in depth on FTTx networks and decide which the most optimal option for
this environment. To make this possible, first, we will look into the main element of
these networks, optical fiber, in order to get to know its operating principles and most
important features in detail. The theoretical part will be followed by the definition of
FTTx networks and point-to-point and point-to-multipoint configurations, moving to
elaborating at length on PON and conclude which of its technologies will be the most
optimal ones nowadays. Finally, in the practical part, simulation software that meets
the design requirements will be chosen, the design of passive optical network will be
made and in the simulation results which justify that the network is viable and can be
implemented in a real case will be obtained.

KEYWORD: Optical fiber, Passive Optical Network, FTTx, Point to Point, Point to
Multipoint, WDM, Simulation

5

Brno University of Technology
Faculty of Electrical Engineering and Communication
Department of Telecommunications

Design of Passive Optical Network

Proyecto Final de Carrera

Especializacin: Comunicaciones

Autor: Juan Salvador Asensi Pla

Supervisor: Ing. Pavel Reichert

RESUMEN

El constante incremento en la demanda de servicios de banda ancha y el consecuente
aumento del volumen de trfico que se genera en nuestras redes de comunicaciones,
han motivado la necesidad de implementar redes de nueva generacin en nuestros
territorios.
El presente proyecto final de carrera tiene como objetivo explicar el diseo y la
planificacin de una red pasiva de fibra ptica (PON). La idea principal de este proyecto
es construir un entorno ficticio que nos permitir profundizar en detalle en las redes
FTTx y decidir cul es la opcin ms optima para este entorno. Para hacer esto posible,
en primer lugar se va a hablar sobre el elemento principal de estas redes, la fibra
ptica, para conocer con detalle sus principios de funcionamiento y sus caractersticas
ms importantes. La parte terica seguir con la definicin de las redes FTTx y las
configuraciones punto a punto y punto a multipunto, para pasar a hablar
detenidamente sobre las redes PON y concluir cual de sus tecnologas ser la ms
optima hoy en da. Por ltimo, entraremos en la parte prctica en la que se elegir un
software de simulacin que se ajuste a los requisitos de diseo, se realizar el diseo
de la red ptica pasiva y se obtendrn los resultados de simulacin que justifiquen que
la red es viable y puede ser implementada en un caso real.

PALABRAS CLAVE: Fibra ptica, Red ptica Pasiva, FTTx, Punto a Punto, Punto a
Multipunto, WDM, Simulacin
6

PROHLEN

Prohlauji, e svoji semestrln prci na tma "Nvrh st FTTx" jsem vypracoval
samostatn pod vedenm vedouc semestrln prce a s pouitm odborn literatury a
dalch informanch zdroj, kter jsou vechny citovny v prci a uvedeny v seznamu
literatury na konci prce.
Jako autor uveden semestrln prce dle prohlauji, e v souvislosti s vytvoenm
tto semestrln prce jsem neporuil autorsk prva tetch osob, zejmna jsem
nezashl nedovolenm zpsobem do cizch autorskch prv osobnostnch a jsem si
pln vdom nsledk poruen ustanoven 11 a nsledujcch autorskho zkona
. 121/2000 Sb., vetn monch trestnprvnch dsledk vyplvajcch z ustanoven
152 trestnho zkona . 140/1961 Sb.

V Brn dne 17/05/2011 ....................................
podpis autora

7

ACKNOWLEDGEMENTS

I would like to thank my project tutor "Pavel Reichert" to offer me the opportunity of
working in this master's thesis and giving me help in the worst moments.

I would also like to thank and dedicate this project to my family and all the friends I
have met over the years.

8

CONTENTS

Chapter 1 - Introduction ......................................................................................................... 13
Chapter 2 - Fundamentals of Optical Fibers and Communication ................................. 15
2.1. Introduction ..................................................................................................................... 15
2.2. Basics of optical fibers ...................................................................................................... 16
2.2.1. Electromagnetic waves ............................................................................................. 16
2.2.2. Refraction and reflection .......................................................................................... 18
2.2.3. Operating principles of optical fibers ........................................................................ 21
2.2.4. Composition and geometry ....................................................................................... 23
2.2.5. Classification and types of optical fibers ................................................................... 26
2.2.5.1. Multimode optical fiber ..................................................................................... 27
2.2.5.2. Single-mode optical fiber ................................................................................... 28
2.3. Optical fiber parameters .................................................................................................. 31
2.3.1. Structural parameters ............................................................................................... 31
2.3.2. Transmission parameters .......................................................................................... 32
2.3.2.1. Attenuation ........................................................................................................ 32
2.3.2.2. Dispersion ........................................................................................................... 35
2.3.2.3. Nonlinearities ..................................................................................................... 37
2.4. Advantages and disadvantages of optical fibers .............................................................. 38
2.4.1. Advantages of optical fibers ...................................................................................... 38
2.4.2. Disadvantages of optical fibers ................................................................................. 40
Chapter 3 - Introduction to FTTx System ............................................................................. 41
3.1. FTTx networks .................................................................................................................. 41
3.2. General architecture of an FTTH network ........................................................................ 44
3.2.1. Point-to-point configuration ..................................................................................... 44
3.2.2. Point-to-multipoint configuration ............................................................................. 45
Chapter 4 - Introduction to Passive Optical Networks ...................................................... 46
4.1. Generic operation of a PON ............................................................................................. 46
4.1.1. Optical Line Terminal ................................................................................................ 47
4.1.2. Optical Network Terminal ......................................................................................... 48
4.1.3. Splitter ....................................................................................................................... 49
4.1.4. Description of operation of Passive Optical Network ............................................... 51
9

4.1.4.1. Downstream channel ......................................................................................... 51
4.1.4.2. Upstream channel .............................................................................................. 52
4.2. Advantages and disadvantages of PON ............................................................................ 53
4.2.1. Advantages of PON.................................................................................................... 54
4.2.2. Disadvantages of PON ............................................................................................... 55
4.3. PON technologies ............................................................................................................. 57
4.3.1. APON ......................................................................................................................... 57
4.3.2. BPON ......................................................................................................................... 57
4.3.3. EPON.......................................................................................................................... 59
4.3.4. GPON ......................................................................................................................... 61
4.3.5. Other PON technologies: NG-PON ............................................................................ 65
4.3.6. Comparative synthesis among xPON standards ....................................................... 70
Chapter 5 - Design of a WDM GPON ..................................................................................... 71
5.1. Considerations of designing and planning of a FTTx Network ......................................... 71
5.1.1. Layers of an FTTx network [32] ................................................................................. 71
5.1.2. Network elements ..................................................................................................... 73
5.1.3. Retail services ............................................................................................................ 76
5.2. Planning of the WDM-GPON ............................................................................................ 80
5.2.1. Project description .................................................................................................... 80
5.2.2. Access network design .............................................................................................. 81
5.2.3. Choice of network elements ..................................................................................... 89
5.3.1. Optisystem 9.0 simulation program ....................................................................... 99
5.4. Design of the WDM-GPON and Results ......................................................................... 103
5.4.1. WDM PON: Complete design .................................................................................. 104
5.4.2. WDM PON: Results .................................................................................................. 131
Chapter 6 - Summary and Conclusions .............................................................................. 139
GLOSSARY ................................................................................................................................ 143
REFERENCES AND BIBLIOGRAPHY ....................................................................................... 145

10

LIST OF FIGURES

Figure 1: Representation of a sine wave of 2 period ................................................................. 17
Figure 2: Electromagnetic spectrum of light ............................................................................... 18
Figure 3: Law of Reflection and Refraction ................................................................................. 19
Figure 4: Snells Law .................................................................................................................... 20
Figure 5: Principle of operation of optical fiber ........................................................................... 22
Figure 6: Optical fiber cable acceptance cone ............................................................................. 23
Figure 7: Structure of optical fiber cable ..................................................................................... 24
Figure 8: Geometry of optical fiber ............................................................................................. 26
Figure 9: Section of multimode optical fiber ............................................................................... 27
Figure 10: Section of single-mode optical fiber ........................................................................... 28
Figure 11: Multimode and single-mode optical fibers ................................................................ 30
Figure 12: Total attenuation in a fiber optic as a function of wavelength .................................. 33
Figure 13: Pulse broadening through polarization mode dispersion .......................................... 36
Figure 14: Some FTTx scenarios .................................................................................................. 43
Figure 15: OLT ............................................................................................................................. 48
Figure 16: Operation of a time-division multiplexing process ..................................................... 49
Figure 17: Generic operation of optical splitter .......................................................................... 50
Figure 18: Wavelength allocation plan in ITU-T G.983.3 ............................................................ 58
Figure 19: G-PON Access System ................................................................................................ 63
Figure 20: WDM-PON Network ................................................................................................... 66
Figure 21: 10G-GPON Network Diagram .................................................................................... 67
Figure 22: 10G-EPON Network Diagram ..................................................................................... 68
Figure 23: FTTx Network layers ................................................................................................... 72
Figure 24: Deployment scenario .................................................................................................. 82
Figure 25: Hospital area .............................................................................................................. 83
Figure 26: Business area .............................................................................................................. 84
Figure 27: Residential area .......................................................................................................... 86
Figure 28: School-Mall area ........................................................................................................ 87
Figure 29: Buildings area ............................................................................................................. 89
Figure 30: Installation in micro-trenches ..................................................................................... 91
Figure 31: ESP fiber optic cable ................................................................................................... 92
Figure 32: Stages of division ........................................................................................................ 94
Figure 33: SC connector ............................................................................................................... 95
Figure 34: WDM PON Complete design.................................................................................. 105
Figure 35: WDM PON Global parameters .............................................................................. 106
Figure 36: WDM PON Downstream spectrum into the first optical fiber ............................... 108
Figure 37: WDM PON SPLITTER Block .................................................................................... 109
Figure 38: WDM PON HOSPITAL Block ................................................................................... 110
Figure 39: WDM PON Downstream spectrum into HOSPITAL Block ...................................... 111
Figure 40: WDM PON ONU Block ........................................................................................... 112
Figure 41: WDM PON BUSINESS Block ................................................................................... 113
11

Figure 42: WDM PON Downstream spectrum into BUSINESS Block ...................................... 114
Figure 43: WDM PON RESIDENTIAL Block .............................................................................. 115
Figure 44: WDM PON Downstream spectrum into RESIDENTIAL Block ................................. 116
Figure 45: WDM PON SHOOL-MALL Block ............................................................................. 117
Figure 46: WDM PON Downstream spectrum into SCHOOL-MALL Block .............................. 118
Figure 47: WDM PON BUILDINGS Block ................................................................................. 119
Figure 48: WDM PON Downstream spectrum into BUILDINGS Block .................................... 120
Figure 49: WDM PON Business, Residential, School-Mall and Buildings ONU Block ............. 121
Figure 50: WDM PON Optical Time Domain Visualizer in BUSINESS Block ............................ 122
Figure 51: WDM PON Optical Time Domain Visualizer in RESIDENTIAL Block ....................... 124
Figure 52: WDM PON Optical Time Domain Visualizer in SCHOOL-MALL Block .................... 126
Figure 53: WDM PON Optical Time Domain Visualizer in BUILDINGS Block .......................... 128
Figure 54: WDM PON Upstream spectrum into the optical fiber ........................................... 129
Figure 55: WDM PON Upstream spectrum into the OLT receiver .......................................... 130
Figure 56: WDM PON HOSPITAL ONU Results ....................................................................... 131
Figure 57: WDM PON BUSINESS ONU Results ........................................................................ 132
Figure 58: WDM PON RESIDENTIAL ONU1 Results................................................................. 133
Figure 59: WDM PON RESIDENTIAL ONU2 Results................................................................. 134
Figure 60: WDM PON SCHOOL-MALL ONU1 Results .............................................................. 135
Figure 61: WDM PON SCHOOL-MALL ONU2 Results .............................................................. 136
Figure 62: WDM PON BUILDINGS ONU Results ...................................................................... 137
Figure 63: WDM PON OLT ONU Results ................................................................................. 138

12

LIST OF TABLES
Table 1: Comparative table of different types of networks ......................................................... 53
Table 2: BPON downstream/upstream bit-rate combinations.................................................... 58
Table 3: Comparative table xPON standards .............................................................................. 70

Chapter 1 Introduction

13

Chapter 1
Introduction

The optical fiber is the most advanced transmission medium and the only one
capable of supporting next generation services. The main advantages of having a last
mile of optical fiber are many: higher bandwidth, longer distances from the central to
the subscriber, the more resistance to electromagnetic interference, increased
security, reduced signal degradation... Moreover, the fact of using PON technology
assumes the elimination of repeaters and optical amplifiers and therefore reducing the
initial investment, lower power consumption, less space, fewer points of failure...
The proliferation of broadband applications such as high definition television
(HDTV) broadcast, broadband Internet access or teleworking, has created a growing
interest among service providers by improving networks to supply broadband services
to residential and small business. To make this possible there must be a high efficiency
between the total cost of infrastructure and services that can be offered to end users.
The Wavelength Division Multiplexing (WDM) is the technology that can deliver the
best solution by considering the Dense Wavelength Division Multiplexing (DWDM) for
long distance applications and Coarse Wavelength Division Multiplexing (CWDM) as
the ideal solution for access networks and metropolitan areas, taking into account the
tradeoff between the cost of components and the ability of services offered. For
example, CWDM using a channel spacing of 20 nm can relax the tolerances of the
components used, such as filters, multiplexers and demultiplexers, routers or switches
and the operating characteristics of the active components such as lasers.
The objective of this thesis aims to demonstrate that the last mile can also be
constructed from fiber, giving to the same the benefits of the information transmission
through this medium. These networks, which have already taken their first steps in
some areas of eastern EU and Japan, are called FTTx (Fiber To The x). Therefore, it will
detail how an optical fiber deployment could be made in a fictitious environment
formed by five different areas that could easily fit to the expansion of a big city.
Throughout this document is going to provide a description of the theoretical
basis, major characteristics and parameters of the optical fiber, to help the reader to
understand how this transmission technology works and all the advantages that offers
over other types of networks such as copper or coaxial cable. Once seated the
knowledge of the optical fiber is passed to briefly define the types of FTTx networks
that exist and their configurations point-to-point and point-to-multipoint. With this, it
will have the information needed to describe in more detail the principles of Passive
Chapter 1 Introduction

14

Optical Networks (PON). In another chapter it will discuss the generic operation of a
PON, taking into account all elements of this type of network and which are its
advantages compared to Active Optical Networks (AON). Subsequently it will analyze
in detail the various types of PON networks, the new next-generation PON (NG-PON)
to conclude which will be the optimum technology for this design. Finally, taking into
account all design considerations, it will make a detailed planning of what will be the
network with all the necessary elements for deployment. The last step will be to
choose a simulation program that meets all design requirements, execute the design
and get the results that demonstrate that the design of passive optical network works
and therefore is viable for practical implementation.
To be able to establish high-speed services to end users, it is necessary that the
engineering design of the network passes through two clearly differentiated phases:
1. The design of the passive optical network. It is the first phase of design
and comprises all physical and structural elements that make up the
fabric of the network, since its beginning in a central office to each
household or businesses which the service is provided.
2. The design of the active network. It is the second phase of design and is
comprised of all communication equipments and other active elements
that allow the data transfer from the central office to the homes, and
which provide services to end users.
Since the second phase of the design consists on the installation of very specific
characteristic equipments of each manufacturer and will be supplied by the operator
which will provide the end services, it is not important to achieving the objectives of
this project.
Communications equipments that constitute the active part of an FTTx network
are not relevant for the transfer of information because they only filter user services
by one or more geographical areas. It will only be subject to development of this
project the passive network design. The fact that the study is very general allows the
physical network design can be extrapolated to any situation and any network of the
same type.

Chapter 2 Fundamentals of Optical Fibers and Communication

15

Chapter 2
Fundamentals of Optical Fibers and Communication

Given the importance of the transmission channel for a good communication,
throughout this chapter it will detail all general aspects with which must have take into
account in a network of fiber optic transmission. The main element of study in this
chapter is the optical fiber, around which it shall specify the other key elements in an
optical communication network.

2.1. Introduction

Optical fiber is a transmission medium commonly used in data networks, like
passive optical networks and others. It can be defined by a thin, transparent and
flexible glass or plastic, by which light pulses are sent in order to represent data to be
transmitted. The beam is completely contained and it spreads inside the fiber at an
angle of reflection above the critical angle of total reflection, according to Snells law.
Optical fibers are widely used in telecommunications, since they allow sending
large amounts of data at a great distance and having higher bandwidths (data rates)
than other forms of communication. They are the transmission medium par excellence
to be immune to electromagnetic interference and because signals travel along them
with less losses.
This transmission mode allows the transport of a multitude of information,
used for applications such as broadband Internet, telephone and cable television,
through more effective signals than copper wires.
Throughout the chapter, it will detail the generic characteristics of the optical
fiber particularizing the most influential aspects of it on the passive optical fiber
networks.


16

2.2. Basics of optical fibers

As mentioned above, optical fiber is a dielectric waveguide that operates at
optical frequencies and transmits information in the light form along its axis.
In this section it will explain the principles and physical laws that obey any type
of optical fiber and make possible their operation as optical waveguides.

2.2.1. Electromagnetic waves

An electromagnetic wave is the propagation of electromagnetic radiation
through space and theoretical aspects are related to the waveform solution that
supports Maxwells equations. Unlike mechanical waves, electromagnetic waves do
not require a material medium to propagate; i.e., they can move through the vacuum.
Light waves are electromagnetic waves whose frequency is within the range of
visible light. Each pulse of light is a single electromagnetic field in propagation or
mode. Particularized solutions for Maxwells equations for the various pulses of light
are those that determine the propagation of electromagnetic fields in waves through
optical fiber.
In light waves, like all electromagnetic waves, there are electric and magnetic
fields at each point of space, which fluctuate rapidly. As these fields also have a
magnitude and a given direction, they are vector quantities.
Other important concept in relation to the waves is the number of oscillations
per second at a point of the light wave, which it is known as frequency. Also important
is the concept of wavelength, which is defined as the distance along the direction of
propagation between two points in phase, i.e., points occupying equivalent positions
in the wave.
Below is an example of wavelength in a sine wave:

17

Figure 1: Representation of a sine wave of 2 period

There is another important parameter that defines another property of waves,
which is the speed propagation of it on a particular environment. Thus, there is a direct
relation between speed of propagation, frequency and wavelength. The speed of an
electromagnetic wave is defined as the product of frequency and wavelength:

(1)

In cases in which the wave propagates through a vacuum, the above equation is
particularized assigning the value of the propagation speed of light speed in vacuum c,
being 3 10
/ , and
the value of wavelength in a vacuum:

(2)

There are thus, infinite values of and f that satisfy the equation (1), and in
fact, there is a wide range of electromagnetic waves whose characteristics obey this
equation. The set of these waves is called spectrum given the enormous range of
variation of the wavelength.
The following figure shows the electromagnetic spectrum of light, depending
on the wavelength and frequency:

18

Figure 2: Electromagnetic spectrum of light
[http://chemed.chem.purdue.edu/genchem/topicreview/bp/ch6/index.php]

In the visible spectrum, different wavelengths are differentiated by colors. This
way, the visible range is from 350 nm (violet) to 750 nm (red). White light is obtained
from the mixture of all visible wavelengths.
The speed of a wave in a vacuum is always the same regardless of wavelength,
and therefore, it is equal for all wavelengths. However, the speed of light in material
substances is lower than in a vacuum and it varies for different wavelengths.
This will allow defining the index of refraction in the next section.

2.2.2. Refraction and reflection

When a ray of light propagates through a homogeneous medium and strikes
the surface of a second homogeneous medium with different composition, part of the
light is reflected and the other part penetrates as a refracted ray in the second
medium, which may or may not be absorbed. Then, refraction produces a change in
the direction of a ray or beam of light rays entering in the medium with different
propagation speed.
The plane of incidence is defined as the physical plane formed by the incident
ray and the normal (i.e., the vector perpendicular to the surface of the medium) in the
point of incidence.

19

Figure 3: Law of Reflection and Refraction

This way, we can say that the light beam incident on another medium with
different composition or transmission speed of light is divided into two beams: one
reflected in the first medium, and the other refracted crossing the edge.
From the above theory, in which a beam is subdivided into two beams of light
traveling on different mediums, it can extract two fundamental laws of optical physics
to quantify the two events: the Law of Reflection and the Law of Refraction, also
known as Snells Law.
The Law of Reflection says that when a light ray strikes on the boundary
between two homogeneous and isotropic transparent mediums, part of the incident
beam is reflected at the same angle and the other part is transmitted to other
medium.
Other important law to understand the operation of the light guided within the
fiber is the Law of Refraction (Snells law). To understand this law, first it must define
the index of refraction as the relationship between the speed of light in vacuum ()
and the speed of light in that medium (
).

(3)


20

Once this is understood, the law of refraction says that the relationship
between the sine of the angle of incidence and the sine of the angle of refraction is
equal to the ratio of the speed of the wave in the first medium and the speed of the
wave the second medium, or better known by the following formula:

sin
sin
(4)

Where:
= refractive index of the first medium,
= Angle of Incidence,
=
index of refraction of the second medium and
= angle of refraction.

Figure 4: Snells Law

Another important concept that it must take into account for the guiding of
light within the fiber is the Total Internal Reflection, which is based on the Law of
Reflection.
The Total Internal Reflection phenomenon occurs when, in the refraction, the
angle of incidence is greater than the critical angle, and therefore all light is reflected.
In the area of interest to this project, it is noteworthy that the optical fiber is a
new practical application of total internal reflection. Thus, when light enters through
an end of a solid tube of glass or plastic, it can be totally reflected on the outer surface
of the tube, and after a series of successive total reflections, out the other end.

21

All these concepts are needed to explain how information can be transported
over our optical fiber networks. In order to do this, in the next section it will explain
the operating principle of light within an optical fiber.

2.2.3. Operating principles of optical fibers

It is important to note that, to make possible transmission of information
through optical fiber, it is necessary to inject light beams from a light source (usually
LED or LASER) in the same fiber.
Incoming light pulses enter into the fiber core, and get passed through it
bouncing off the walls lining that surrounds the core. This is possible because the
refractive index difference between core and cladding, and because the refractive
index of the coating is less than the core. Finally, the light rays reach the opposite end
of the fiber, which arrive at the receiver and they are transformed into electrical
signals.
In short, the principle of operation of the optical fiber is based on the principle
total internal reflection, which has been explained in the previous section. The
transmitter emits rays of light at an angle of incidence greater than critical angle
allowing the transmitted beam is reflected first in the coating.
This same process is repeated again and again through the entire cable. Since
angles of incidence and reflection are equal, some beams continue to be reflected and
transmitted in a zigzag pattern through the entire length of the fiber. Thus, the light
energy from the transmitting source is contained in the nucleus.
However, not all the beams of light from the power source Light enter to the
core, or they enter but cannot be transmitted through internal reflections in the fiber.
This is because the light hitting the interfaces core-cladding, it makes this with a less
angle than the critical angle, losing that energy into the coating by refraction.
Below is the principle of propagation through optical fiber based on Snells law:


22

Figure 5: Principle of operation of optical fiber

In Figure 5, a reference rays are taken with different incident angles on the
fiber core, which holds that n1 > n2.
Looking from left to right, the first three rays have a lower incidence angle than
the critical angle, so that the beam is refracted on the cladding, losing energy in the
core of the fiber.
The fourth beam, has an incidence angle exactly equal to the critical angle,
therefore the phenomenon of total internal reflection occurs for the first time but
dissipates energy at the boundaries.
The fifth ray, which has an incidence angle greater than the critical angle, it
obeys the Law of Total Reflection for the remainder confined to the core.
If we consider that all light rays impinge at an angle above the critical angle are
trapped inside the nucleus, a virtual cone is generated and is called the cone of
acceptance. Therefore, all rays impinging on the fiber outside the cone of acceptance
do not propagate through it lengthwise. It can be represented graphically as follows:


23

Figure 6: Optical fiber cable acceptance cone

The larger the cone of acceptance of an optical fiber, the greater the amount of
rays can be transmitted by such mediums, and therefore does not need to be so
precisely at the time of emitting light from the transmitter. Moreover, a greater
number of beams of light inside the fiber, it involves a variation in the directions of
propagation of the rays, and their quality decreases.

2.2.4. Composition and geometry

The fiber optic cables are manufactured with various materials to achieve the
effect described above and adapt to the environment of the installation. Thus, external
cables must be strong, weatherproof and resistant to ultraviolet radiation (UV) and
maximum temperature variations that can occur during installation and throughout its
life. On the other hand, the internal cables must be strong and flexible, with the
required degree of fire resistance or smoke emission. The most common components
are: polyethylene, polyvinyl chloride (PVC), polyurethane, and polyfluorinated
hydrocarbons, ethylene vinyl acetate, aramid ropes and other materials for the steel
shell, torn wire, the central member and interstitial filling [6].
An optical cable consists of a set of generic elements present in any cable,
regardless of the structure and composition. It can be present the optical cable as a
structure consisting of six concentric layers, being these:
Optical fiber: This layer is composed of optical fibers responsible of
transmission of information. (1)
Secondary Protection: mechanical protection that usually is loose or
dense, with interstitial filler. (2)
Traction element: typically Aramid, Kevlar or fiberglass. (3)

24

Protective inner cover: made of PE, PVC or polyurethane. (4)
Protective shell, usually steel. (5)
Protective outer jacket: PE, PVC, polyurethane or fluoropolymer. (6)

In Figure 7 is shown the most important elements of an optical fiber that are
described above and encoded with numerology 1 to 6:

Figure 7: Structure of optical fiber cable

Based on the above generic settings and their respective concentric layers,
there are several basic structures of optical fiber cables. The most common are cables
with loose structure and cables with dense or adjusted structure, but there are other
less significant structures such as armored cables, self supporting aerial, submarines,
OPGW, hybrid and composite, etc
Turning now to the important part of the cable, which is the same optical fiber,
it can say that any optical fiber consists essentially of three concentric layers or
structures which are detailed below:

Core: The core, also known as transmitter tunnel, is the innermost part of the
fiber and is responsible for driving the optical signals from the light source to the
receiving device. It is one continuous glass fiber made from high temperature ultra-
pure quartz, plastic or silicon dioxide.

25

Often, the silica glass (SiO
2
) is usually doped with materials such as phosphorus
oxide (P
2
O
5
), germanium oxide (GeO
2
) or boron oxide (B
2
O
3
) to adjust its refractive
index.
It has a very small diameter, ranging between 10 to 300 mm. The higher is the
core diameter the greater the amount of light the cable can carry. Indeed, the fiber
optic cables across the board are classified according to their diameter.

Cladding: It is the middle part of the fiber, which surrounds and protects the
core. This medium has a refractive index lower than the core, so that acts as a
reflective layer which keep the waves trying to escape the nucleus.
It is manufactured with high temperature and made of silicon crystalline
nature, and is generally quartz or transparent plastic. This layer usually adds several
layers of plastic to absorb the potential impact or shock that the fiber can get and
provide extra protection against excessive cable bending.

Coating or buffer: The coating is the outer part of the fiber and acts as a shock
absorber, protecting the core and the cladding from damage and external agents. In
short, the cable provides some mechanical protection to manipulation. It is made of
plastic material, capable of protecting the fiber from moisture, crushing, rodents and
other environmental hazards.
In many cases the coating is clearly divided into two sublayers: the primary
coating (buffer) and the secondary coating (coating).
This cover can be easily removed by physical or mechanical means to splicing
and gives the fiber a predetermined outer diameter, that can be 125, 250 (in case of
having only one primary coating) or 500 or 900 m (if also have an additional
secondary cover).


26

Figure 8: Geometry of optical fiber

2.2.5. Classification and types of optical fibers

An optical fiber is a dielectric waveguide operating at optical frequencies. The
propagation of light along a waveguide can be described in terms of a set of guided
electromagnetic waves called the modes of the waveguide. These guided modes are
referred to as the bound or trapped modes of the waveguide, and each guided mode is
a pattern of electric and magnetic field lines that is repeated along the fiber at
intervals equal to the wavelength [7].
Once the ray has entered into the fiber core, there are a limited number of
optical paths that a light beam can take through a fiber. If the diameter of the fiber
core is large enough to allow multiple different pathways for light to travel along the
fiber, the fiber is called multimode. Single-mode fiber has a much smaller core that
allows light rays traveling exclusively by a single mode. In the following sections it will
detail these types of fiber:


27

2.2.5.1. Multimode optical fiber

As its name suggests, this type of fiber can propagate several transmission
modes simultaneously. This is possible because the core diameter of such fibers is wide
(50 microns or 62.5 microns), so that the coupling of light into its different modes is
simpler. The following figure shows a section of a multimode optical fiber:

Figure 9: Section of multimode optical fiber

In turn, multimode optical fibers are classified according to the index of
refraction of the core. They can be multimode step-index and multimode graded-
index, as detailed below:

Multimode step-index optical fiber: This group includes all those fibers
in which the value of the index of refraction in the core and the coating
is always constant, and also the value of the refractive index of the core
is significantly higher than the coating, so that, the refractive index
when moving from one area to another change abruptly.

28

Multimode graded-index optical fiber: In this type of optical fiber, the
core has not a constant refractive index, but is made from concentric
layers of optical material with refractive index value that decreases from
the center to the outside. Anyway, this value is always greater than the
refractive index of the coating.

2.2.5.2. Single-mode optical fiber

As its name suggests, this type of fiber allows propagation of only one single
mode of transmission. This is possible because the core diameter of such fibers is very
small (between 8 m and 10 m), so that only allow the propagation of a fundamental
beam.

Figure 10: Section of single-mode optical fiber


29

Although they delete the modal dispersion, these fibers are affected by other
type of dispersion called chromatic dispersion. Chromatic dispersion is produced by
the light source as a result of spectral coherence. There are no ideal light sources with
null bandwidth and infinite coherence, but always emit different residual wavelengths
around a fundamental wavelength.
The manufacture of the core is performed through a material whose index of
refraction is significantly higher than the coating, which leads to define it as a step-
index fiber.
High rates can be achieved with this type of fiber (on the order of hundreds of
Gbps), making it an attractive fiber for demanding high bit rate. This coupled with its
lack of modal dispersion makes it ideal for long distance links.
In addition, as a result of having a very small core diameter, the cone of
acceptance is also small and light-emitting have been to transmit a very precise and
narrow beam for a perfect fit. So, LEDs (which are very cheap) cannot be used but it is
necessary to use lasers that are more expensive.
In turn, single-mode optical fibers are classified according to the manufacture
of fiber materials in SMF, DSF and NZDSF fibers, whose main features are listed below:

Standard single-mode fiber (SMF): Standard single-mode optical fibers
have very basic features. Therefore, they have an attenuation of 0.2
dB/Km and chromatic dispersion on the order of 16 ps/kmnm in
wavelength of 1550 nm. For a wavelength of 1330 nm has a zero
chromatic dispersion, but the attenuation of the signal is doubled.

Dispersion shifted fiber (DSF): Single-mode optical fibers have
dispersion-shifted features that improve the chromatic dispersion over
the previous. They have attenuation of the order of 0.25 dB/km (greater
than SMF fiber) but have a zero chromatic dispersion in a wavelength of
1550 nm. Its main drawback is due to nonlinear effects, such as the
phenomenon of fourth wave mixing (FWM) that precludes the use of
these optical fiber systems by wavelength division multiplexing (WDM).
This fiber was introduced into the market in the mid 80s and remains in
use, yet it has never become as common as standard single-mode fiber.


30

Non-zero dispersion shifted fiber (NZDSF): Single-mode non-zero
dispersion shifted optical fibers arise to try to solve the drawbacks of
the previous fiber. They have a chromatic dispersion value close to zero
but not zero (hence the name) at a wavelength of 1550 nm. By contrast,
they achieve largely offset the effects of nonlinear phenomena by
chromatic dispersion.

In the figure 11 is graphically shown the different types of optical fiber
explained above:

Figure 11: Multimode and single-mode optical fibers


31

2.3. Optical fiber parameters

Once it is known the structure and composition of an optical fiber, it is necessary
to know the parameters that characterize the fibers. These parameters will help to
know which the most important characteristics of these cables and their strengths and
limitations are.
There are two groups of parameters which encompass each and every one of
them, and they are the structural parameters and transmission parameters. All of
them establish the conditions under which you can perform data transmission.

2.3.1. Structural parameters

The structural parameters are those related to the geometry and characteristic
structure of each fiber. These parameters determine and characterize the types of
optical fibers on the market, so that the classification of fibers directly depends on
them. There are many and varied types of parameters that characterize the fibers but
the most important are [9]:
Refractive index profile, which allows fibers with different dispersions.

Size of core and cladding, which determine the type of propagation:
single-mode or multimode.

Mode field diameter, which indicates how the geometrical distribution
of light occurs in the spread mode.

Numerical aperture, which indicates the number of rays that can enter
the core of an optical fiber transmission.

Cutoff wavelength, which determines that the optical fiber transmits on
an only single mode of propagation.


32

2.3.2. Transmission parameters

Signal attenuation is one of the most important properties of an optical fiber,
because it largely determines the maximum separation between a transmitter and a
receiver. Since repeaters are expensive to fabricate, install, and maintain, the degree
of attenuation in a fiber has a large influence on system cost. Of equal importance is
signal distortion. The distortion mechanisms in a fiber cause optical signal pulses to
broaden as they travel along the fiber. If these pulses travel sufficiently far, they will
eventually overlap with neighboring pulses, thereby creating errors in the receiver
output. The signal distortion mechanisms thus limit the information-carrying capacity
of a fiber [7].

2.3.2.1. Attenuation

The light that travels through an optical fiber loses power as it moves to along
it, and therefore, with distance. Attenuation losses limit transmission distance and
depend on the wavelength of light and material through which it spreads.
Signal attenuation is defined as the ratio of the optical output power P
out
from a
fiber of length L to the optical input power P
in
. The symbol is commonly used to
express attenuation in decibels per kilometer [7]:

10
log

(5)

As we can see in Figure 12, the lowest losses are at a wavelength of 1550 nm, a
value that is heavily used for long distance transmission, while the highest values are in
shorter wavelengths.


33

Figure 12: Total attenuation in a fiber optic as a function of wavelength
[http://www.invocom.et.put.poznan.pl/~invocom/C/P1-9/swiatlowody_en/p1-1_2_2.htm]

Below 800 nm is not convenient to use the fiber due to high attenuation by
Rayleigh scattering. Above 1600 nm, attenuation problems are presented by the effect
of infrared radiation. In addition, the emitters and photodetectors technology for this
wavelength is very recent. Therefore, optical fiber communications typically operate in
a wavelength region corresponding to one of the following telecom windows [11]:
The first window at 800900 nm was originally used. Therefore, the first
telecom window is suitable only for short-distance transmission.

The second telecom window utilizes wavelengths around 1.3 m, where
the loss of silica fibers is much lower and the fibers chromatic
dispersion is very weak, so that dispersive broadening is minimized. This
window was originally used for long-haul transmission. However, fiber
amplifiers for 1.3 m (based on, e.g. on praseodymium-doped glass) are
not as good as their 1.5 m counterparts based on erbium. Also, low
dispersion is not necessarily ideal for long-haul transmission, as it can
increase the effect of optical nonlinearities.


34

The third telecom window, which is now very widely used, utilizes
wavelengths around 1.5 m. The losses of silica fibers are lowest in this
region, and erbium-doped fiber amplifiers are available which offer very
high performance. Fiber dispersion is usually anomalous but can be
tailored with great flexibility.

The signal attenuation into the fiber does not depend on bandwidth and
modulation because the carrier frequency is greater by several orders of magnitude to
frequency modulation which does not happen in other conventional waveguides. The
attenuation of light in an optical fiber is produced as a result of various effects and
losses can be classified into extrinsic and intrinsic losses:

Extrinsic losses: Such losses are due to extrinsic factors to the nature of the
fibers as are the parameters of curvature and the joints between optical fibers, or
external parameters due to the installation or temperature at which the fiber is
subjected. As a result of these, attenuations are produced in the light transmitted
along the fiber, which mainly lead to a loss of signal strength along the transmission by
reducing the value of the amplitude of the signal.

Intrinsic losses: Such losses are due to intrinsic factors to the nature of the
fibers, and therefore characteristic of the manufacturing and operational working of
the same. As also occurs with extrinsic losses, these losses originate decreases in the
transmitted signal power, reducing the amplitude of the signal. For example, the most
important intrinsic losses are:
- Inherent losses in the fiber during the manufacturing process
thereof.

- UV and IR absorption, which cause the called transmission
windows of the optical fiber or optimal wavelength bands for the
light transmission.

- Fresnel reflection, caused by a jump or change of refraction
index in the union of the fiber.

- Rayleigh scattering, occurs when light collides on its path with
foreign particles to the continuous medium by which it spreads.

35

- Absorption due to impurities is a major cause of signal loss in an
optical fiber. The most common impurity is the OH
-
molecule,
which remains into the fiber as a residue despite stringent
manufacturing techniques. It has a fundamental absorption at
2700 nm and its harmonics produce three absorption peaks at
1383 nm, 1250 nm, and 950 nm. These absorption peaks define
the three typical windows of operation.

2.3.2.2. Dispersion

Dispersion is an inherited property of the fiber that can be attributed to the
spreading of an optical pulse in time domain due to the difference in the velocities of
the various spectral components that are associated with that optical pulse. We have
to note that each optical pulse has different spectral components or multiple
frequencies. Each spectral component has its own velocity and can travel through a
different path. Because of this, each component reaches the exit end of a
communication channel (fiber) at different intervals of time. This difference in time
experienced by the various spectral components leads to a longitudinal spreading of
the pulse of a cylindrical waveguide [12].

Dispersion results in losses in the bandwidth of the signal, as well as loss per
distance traveled (as in the case of attenuation). The bandwidth of an optical fiber is a
measure of transmission capacity, limited by the total dispersion of the fiber or
widening of the transmitted pulse. This limits the ability to transmit information as
pulses are distorted and widened along the transmission, overlapping each other and
become indistinguishable to the receiver.

Likewise, dispersion restricts transmission distance so as the bandwidth of the
same and it is a function of the optical fiber length, since the higher the fiber length
the more pronounced the effect. This type of dispersion in the bandwidth can be
subdivided into three categories:


36

Modal dispersion, caused by the different ways or modes that follows a
beam of light in fiber, with the result that the different rays light travels
different distances and reach the other end of the fiber in different time
instants.

Polarization mode dispersion (PMD), produced because the fiber is not
exactly a cylindrical waveguide, which results in a phenomenon called
birefringence whereby causes a propagating pulse that loose the
balance between the polarization components.

Figure 13: Pulse broadening through polarization mode dispersion

Chromatic dispersion, which arise as a result of the dispersion material
and waveguide dispersion (own material and geometry of the fiber) and
it is due to the variation of refractive index of an optical medium with
the wavelength.


37

2.3.2.3. Nonlinearities

The fact of assuming that optical communication systems behave linear is an
appropriate approximation when operating at moderate power levels (magnitudes of
the order of mW) and transmission rates do not exceed 2.5 Gbps. But at higher speeds
(around 10 Gbps) or higher powers, effects of certain nonlinearities begin to be
important, and in the case of WDM systems these effects are very important even with
powers and moderate transmission rates, determining the number of channels in the
system and the separation between them.
The nonlinearities of the fibers can be classified into two categories:
Those that occur as a consequence of the dependence of the refraction
index with the applied field intensity, known as Kerr effect, which in
turn is proportional to the square of the amplitude: self-phase
modulation (SPM), cross-phase modulation (CPM) and the four wave
mixing (FWM).
Those produced by diffraction effects in the fiber, due to interaction of
light waves with photons (molecular vibrations) in silicon of the core:
stimulated Raman scattering (SRS) and stimulated Brillouin scattering
(SBS).


38

2.4. Advantages and disadvantages of optical fibers

Classical communications systems use electrical signals supported by coaxial cable,
copper, radio channel, etc. depending on the type of application to provide service.
These systems have some drawbacks and limitations that made it necessary to find
other ways to transfer data, based on other signals, such as fiber optics.
The main fact that a fiber use light energy as a transmission medium, it presents an
important set of advantages over other media used in the transmission of analog and
digital signals.

2.4.1. Advantages of optical fibers

The main advantages of optical fiber as transmission medium compared to
other systems based on electromagnetic signals over metals are [1]:

They allow higher throughput. The signals travel throughout optical
fiber cables approaching the speed of light ( 3 10
/ ) while
the electrical signals travel through the cables at a speed between 50 -
70% of it, depending on the type of cable.

Greater transmission capacity. They can be achieved speeds exceeding 2
Gbps, since the transmission rate increases with frequency. The fact to
have inherently greater bandwidth enables the simultaneous
transmission of large amounts of information. This property is directly
related to the previous.

They have total immunity to electromagnetic interference. Optical fiber
does not produce any electromagnetic interference and it is not
affected by radiation or nuclear electromagnetic pulses (NEMP)
accompanying nuclear explosions.

It do not exist problems to return to ground, and cross-interference
reflection as in electrical transmission lines.
Attenuation in transmission increases more slowly with distance than in
the case of electrical cables, allowing greater distances between signal
repeaters.


39

It can get bit error rates typical of the order of 10
9
compared to rates of
the order of 10
6
to reach those coaxial cables. This increases the
effective speed of data transmission, reducing the number of
retransmissions or the amount of redundant information necessary for
detect and correct transmission errors.

They do not have a risk of short circuit or other damage of electrical
origin.

Fiber optic cables, weigh one-tenth of copper shielded cables, for
example. This is a very important consideration in military applications,
typical of ships and aircraft.

They usually have a much smaller diameter cable. In addition, they are
much more flexible and easier to install than electrical cables.

The fiber optic cables are suitable for use in a wide range of
temperatures, since they support extreme temperatures better than
cables with metallic origin.

Unlike the copper cables, the transmitted signal cannot be drawn from a
fiber without tampering it. This way, the optical fiber communication
provides 100% signal security.

It is possible to increase transmission capacity by adding new data
channels using different wavelengths to those already employed, by
WDM techniques.

Optical fiber has a higher resistance to environments and corrosive
liquids than electric cables.

Raw materials, which are used in the manufacture of optical fibers, are
very abundant, and it is hoped that eventually the manufacturing cost is
reduced to a level similar to metallic cables.

The operational half-life and average time between failures of a fiber
optic cable are much higher than an electrical origin cable.


40

2.4.2. Disadvantages of optical fibers

Optical fiber not only has advantages. It also has some drawbacks, which it is
not convenient to forget, because in some circumstances they limit studies and
deployment projects in some networks based on this transmission medium [1]:
The cost, either installation or transmitters and receivers, is usually high
compared with other types of communications. In addition, companies
do not charge for time used, but the amount of information transferred
to the computer that is measured in megabytes.

Electro-optical conversion. Before connecting an electrical signal to
optical fiber, the signal must be converted to light spectrum (850, 1310
or 1550 nm). This conversion is carried out by an electronic device on
the end of the transmitter, which provides an own format to the
communication signal and the signal becomes optics using a LED or a
solid state laser.

Particular installation. Because the fiber is predominantly silica glass,
special techniques are needed for engineering and installation of the
links. No longer apply conventional methods wire cable installation as,
for example, welding or crimping. It also requires the right equipment to
test and put into service cables. Technicians must be trained and
previously prepared for the installation and commissioning of optical
fibers.

Repairs. An optical fiber cable that has been damaged is not easy to
repair. Thus, repair procedures require a team technician with enough
skill and ability in handling equipment. In some situations it may be
necessary to repair the entire cable. This problem could be even more
complex, given the large number of people using this service. Therefore,
it is important to design an own system with physically different routes
that will address such contingencies.

With all this, you can perform an initial assessment, which highlights the
multitude of benefits that fiber has over electric transmission mediums. Despite the
drawbacks, most of them are decreasing over time. Being a fairly recent mode of
transmission, allows this research are very active. On the other hand, there is an
optical fiber support mature enough to serve very high performance.
Chapter 3 Introduction to FTTx System

41

Chapter 3
Introduction to FTTx System

The application of PON technology for providing broadband connectivity in the
access network to homes, multiple-occupancy units, and small businesses commonly is
called Fiber-to-the-x. This application is given the designation FTTx [14].

The next chapter consists of an extensive collection of FTTx optical transmission
systems, and specifically the FTTH system by which the design and deployment of the
proposed network in this project is based. Therefore, it describes general operation of
such networks, standards of architecture and operating structure, as well as the
current situation in the world and the services that they offer, and in the near future
will be able to offer. This will provide an overview of FTTH technology and in-depth the
knowledge of the particular circumstances of such networks.

3.1. FTTx networks

FTTH networks belong to the family of FTTx transmission systems within the
world of telecommunications. These networks, which are considered broadband, have
the ability to transport large amounts of data and information at very high bit rates up
to a point close to the end user.
The FTTx family comprises a set of technologies based on transport of digital
signals through optical fiber as transmission medium. Different levels of scope,
depending on the degree of optical fiber closer to the end user, which arise as a result
of a greater or lesser price reduction of these systems.
All FTTx networks support a logical network configuration of tree, star, bus and
ring, and all with the ever present possibility of using active components depending on
the location of users or end customers.


42

Depending on the degree of penetration of FTTx, these networks can be
classified into the following [14]:

FTTB, fiber-to-the-business, refers to the deployment of optical fiber
from a central office switch directly into an enterprise.

FTTC, fiber-to-the-curb, describes running optical fiber cables from
central office equipment to a communication switch located within
1000 ft (about 300 m) of a home or enterprise. Coaxial cable, twisted-
pair copper wires (e.g., for DSL), or some other transmission medium is
used to connect the curbside equipment to customers in a building.

FTTH, fiber-to-the-home, refers to the deployment of optical fiber from
a central office switch directly into a home. The difference between
FTTB and FTTH is that typically, businesses demand larger bandwidths
over a greater part of the day than do home users. As a result, a
network service provider can collect more revenues from FTTB networks
and thus recover the installation costs sooner than for FTTH networks.

FTTN, fiber-to-the-neighborhood, refers to a PON architecture in which
optical fiber cables run to within 3000 ft (about 1 km) of homes and
businesses being served by the network.

FTTO, fiber-to-the-office, is analogous to FTTB in that an optical path is
provided all the way to the premises of a business customer.

FTTP, fiber-to-the-premises, has become the prevailing term that
encompasses the various FTTx concepts. Thus FTTP architectures
include FTTB and FTTH implementations. An FTTP network can use
BPON, EPON, or GPON technology.

FTTU, fiber-to-the-user, is the term used by Alcatel to describe their
products for FTTB and FTTH applications.


43

Figure 14: Some FTTx scenarios
[http://policom.dei.polimi.it/education/comunicazioni/lezione-PON.pdf]

The use of fiber optics as medium of transmission to homes, and so, to end
users, ensures network completely adapted to the needs of both current and future.
The reuse of this physical infrastructure saves money over time despite strong outlay
in initial phase, amortizing it in a short time.


44

3.2. General architecture of an FTTH network

FTTH technology involves the introduction of fiber optic in global network, both
the backbone network operator as the last mile. In relation to the last mile, it includes
the fiber from the central office to each household that requires services.
The interconnection between the end subscriber and the distribution node
which is going to provide the services can be done through various physical
configurations, which are detailed below.

3.2.1. Point-to-point configuration

Point to point configuration, in terms of optical fiber is concerned, it is
specifically a link between the central node and the end user.
Transmission point to point links are operated by companies that have access
to fiber optic in outside plant and need to connect remote locations with some
communications capability, which can vary from a voice or telephone link, up to high-
speed data link.
As for the active part of the network is concerned, it is noteworthy that the
equipment used for transmission of information in point to point links are PDH or SDH,
besides WDM. These links have high capacity and are very useful in the business world.
However, not all are benefits. In case of use of this configuration for home
users, it would be a high cost of deployment, to which any operator, whether
telecommunications or neutral, is willing to address.
In addition, it would break the pattern of global network configuration, because
tree or star configuration, it would be dismantled, increasing the cost of network
expansion as well as operation and maintenance.


45

3.2.2. Point-to-multipoint configuration

Point to multipoint configuration, in terms of fiber optics is concerned it is in
which FTTH networks are based. Typically, this configuration is called PON (Passive
Optical Network).
Architecture based on PON or passive optical networks is defined as a global
system devoid of active electronic elements in the last mile. As it is one of the most
important points in this project, it will devote an entire chapter to explain what they
are and how such networks work.

Chapter 4 Introduction to Passive Optical Networks

46

Chapter 4
Introduction to Passive Optical Networks

A passive optical network is a network which by its nature provides a variety of
broadband services to users through optical fiber access. PON allows removing all
active components between the server and client introducing in place optical passive
components to guide the traffic throughout the network. Its principal element is the
optical splitter.
The usage of passive architecture can reduce costs and are mainly used in FTTH
networks. By contrast, the bandwidth is not dedicated, but rather multiplexed in a
single fiber in the network access points. In short, this is a point-to-multipoint
configuration network.
Moving from the network to the user, it can say that PON architecture consists
of the following equipment: an Optical Line Terminal (OLT) at the service providers
central office and a number of Optical Network Units (ONUs) or Optical Network
Terminals (ONTs) close to end users.

4.1. Generic operation of a PON

As has been described in the previous section in general terms, a passive optical
network always works under transmission between the OLT and the different ONTs
through optical splitters, which multiplex or demultiplex signals based on their origin
and destination.
Therefore, they appear three distinct devices in the network: the OLT, the ONT
and the splitter, each of which has a necessary and priority function on the passive
optical network. Below are detailed the function and general characteristics each of
them.


47

4.1.1. Optical Line Terminal

The OLT is located in a central office and controls the bidirectional flow of
information across the ODN (Optical Distribution Network). An OLT must be able to
support transmission distances across the ODN of up to 20 km (currently could be
more with EDFA). In the downstream direction the function of an OLT is to take in
voice, data, and video traffic from a long-haul network and broadcast it to all the ONT
modules on the ODN. In the reverse direction (upstream), OLT accepts and distributes
all the traffic from the network users [14].
Simultaneous transmission of separate service types on the same fiber in the
ODN is enabled by using different wavelengths for each direction. For downstream
transmissions, a PON uses a 1490 nm wavelength for combined voice and data traffic
and a 1550 nm wavelength for video distribution. Upstream voice and data traffic use
a 1310 nm wavelength [14].
Each OLT is tasked to avoid interference between the contents of downlink and
uplink channel, using two different wavelengths superimposed. For this, techniques for
WDM (wavelength division multiplexing) are used, and are based on the use of optical
filters.
An optical power measurement at the OLT is also required to ensure that
sufficient power is delivered to the ONTs. This should be done during the initial
activation because it cannot be repeated without interrupting service for the entire
network once the network has been connected.

Finally, note that the OLT does not emit the same light output at all ONT fairly,
but depends on the distance they are from the plant. Therefore, a user close to the
central need less power, while a remote user will need a higher power.


48

Figure 15: OLT
[http://www.motorola.com/Video-Solutions/XP-EN/Products-and-Services/Broadband-Access/Optical-Line-
Terminals/AXS2200_XP-EN]

4.1.2. Optical Network Terminal

An ONT is located directly at the customers premises. There its purpose is to
provide an optical connection to the PON on the upstream side and to interface
electrically to the customer equipment on the other side. Depending on the
communication requirements of the customer or block of users, the ONT typically
supports a mix of telecommunication services, including various Ethernet rates, T1 or
E1 (1.544 or 2.048 Mbps) and DS3 or E3 (44.736 or 34.368 Mbps) telephone
connections, ATM interfaces (155 Mbps), and digital and analog video formats [14].
A wide variety of ONT functional designs and chassis configurations are
available to accommodate the needs of various levels of demand. The size of an ONT
can range from a simple box that may be attached to the outside of a house to a fairly
sophisticated unit mounted in a standard indoor electronics rack for use in large MDU
or MTU applications, such as apartment complexes or office buildings. At the high-
performance end, an ONT can aggregate, groom, and transport various types of
information traffic coming from the user site and send it upstream over a single-fiber
PON infrastructure. The term grooming means that the switching equipment looks
inside a time-division-multiplexed data stream, identifies the destinations of the
individual multiplexed channels, and then reorganizes the channels so that they can be
delivered efficiently to their destinations [14].


49

In summary, we can say that the ONT are elements capable of filtering the
information associated with a particular user from the OLT. They also have the
function of encapsulating a users information and send it toward the OLT header to
redirect it to the appropriate network.
Each ONT receives all the signals sent by its corresponding header ONT, like the
rest of ONTs of the same stage. Information of the OLT is transmitted by broadcast
TDM, and reaches all ONT by alike. However, the ONT has the task of filtering the
information that only goes directed himself (at a given time interval).
The figure below shows graphically Operation of a time-division multiplexing
process:

Figure 16: Operation of a time-division multiplexing process [14]

4.1.3. Splitter

Splitters are passive power dividers that allow communication between the OLT
and their respective ONT who serve. However, not only are dedicated to multiplex or
demultiplex signals, but also combine power: they are bidirectional optical distribution
devices with one input and multiple outputs:
The signal which enters from input port (downlink), it proceeds from the
OLT and it is divided among multiple output ports.

The signals which enter from the exits (uplink), they come from ONT
and they are combined at the entrance.

50

The fact of being completely passive elements, it allows them to operate
without extern power, lowering their cost of deployment, operation and maintenance.
They just introduce optical power loss on communication signals, which are inherent in
nature.

There is an inverse mathematical relationship between the losses introduced by
the divisor and the number of outputs of the same, being this:

10
1

(6)

Therefore, a splitter with two outputs, in the worst case, it loses 3 dB (half
power) at each exit.
Graphically, it can be expressed the operation of a splitter with the following
Figure:

Figure 17: Generic operation of optical splitter


51

There are various types of splitters, as not all are built from same technology.
However, the common divisors are two types:

For devices with large number of outputs (> 32 outputs) splitters based
on planar technology are used.

For devices with fewer outputs (<32 outputs) splitters based on fused
bi-conical couplers are used.

4.1.4. Description of operation of Passive Optical Network

Once detailed all the elements that build a PON, it is necessary to know how
the global system works and the behavior of the network with all the interconnected
elements, from the head OLT towards ONT users, and vice versa.
The most important thing to note in the generic operation of the network is the
existence of two channels, one ascending and one descending. However, both
generally work through the same physical cable, so different wavelengths are assigned
to each transmission channel and, depending on traffic, coexisting in the same fiber at
least 3 different wavelengths: one for video flow in the upstream channel, and two for
data flow of uplink and downlink respectively.
Below is going to be analyzed in more detail both transmission channels:

4.1.4.1. Downstream channel

The downstream channel is the direction of information from the OLT operator
to the ONT located on the end user. In this network, the PON behaves like a point-
multipoint network.
The OLT includes plenty of added voice and data frames that go towards PON,
through the P-OLT (voice and data) and the V-OLT (video). Frames collected by these
teams are transformed to signals which inject in the different branches of the users.
These branches are formed by one or two fibers that carry signals bi or unidirectional,
and are passively coupled by optical splitters that allow the union of all the ONT in the
network, without intermediate regeneration of signals (avoiding active elements).

52

These dividers are responsible for receiving information from the OLT and send
all the information to all their outputs. Once the information arrives to ONTs, each will
only be able to process the traffic it belongs, or for which have the access by the
operator, thanks to the AES (Advanced Encryption Standard) security techniques. In
this procedure, TDM (time-division multiplexing) broadcast protocol is used, sending
information to each user in different times.
The OLT has different pre-set time intervals, each one corresponded to a
particular user. So, in function of each temporal segment, the ONT of each user filters
the information addressed to him.
An important aspect to consider is the wavelength () which the OLT transmit
information to ONT. These lengths vary according to if a tree branch or ONT
connection has a single fiber or two fibers.

4.1.4.2. Upstream channel

The upstream channel is the direction of information from the ONT end user to
the OLT operator. In this network, the PON behaves like a point to point.
Each ONT includes the added frames of voice and data (from each user) that
are directed toward the OLT. At this point, the ONT performs the same operation as
the OLT in the downstream channel, i.e., turn the frames into injecting signals through
optical fiber that have been dedicated to the user.
The splitters of each stage are in charge of collecting information from all
corresponding ONTs and multiplex it in a single output fiber towards the operator OLT.
In order to transmit information from different ONT on the same channel, is
necessary (as in the downstream channel) the use of TDMA, so that each ONT sends
the information in different time intervals controlled by the OLT unit.
As for the working wavelengths, it is noteworthy that regardless of whether the
connection of the ONT to the splitter is two or single fiber, the wavelength of the
upstream channel work is always the same. The information sent by the user (voice
and/or data), always travel at
1310 .


53

4.2. Advantages and disadvantages of PON

To understand why FTTx architecture is based on PON network, it is necessary
make a direct comparison between the point-to-point (P2P), passive point-to-
multipoint (PON) and active point-to-multipoint networks (AON).
Below is a comparative table showing advantages and disadvantages of each of
the three types of network configuration cited above, that clearly justify the use of
FTTx PON compared to other configurations.

Network Type Positive Aspects Negative Aspects

Point-to-Point

High capacity High cost of deployment
Active Point-to-
Multipoint (AON)
High capacity
High cost of deployment
High cost of operation and
maintenance
Passive Point-to-
Multipoint (PON)
High capacity
Use of passive elements
(lower investment)
Low cost of operation and
maintenance
Flexibility and scale
All services in a fiber
ITU G.983.3 standardization
Protection methods
required against
sabotages
High impact on
breakdown at central OLT
Table 1: Comparative table of different types of networks

As shown in the table above, PON are the most appropriate to the design of
physical network architecture for FTTx deployments. The fact of having lines
specifically dedicated for user towards to central office (where the OLT is),
considerably reduces the cost of initial network deployment. This is not covered by
point-to-point networks, which despite providing a high bandwidth per user, it's not
worth their high cost of deployment.

As for active networks, the inclusion of active elements increases not only the
cost of deployment of the network itself, but also the operation and maintenance of

54

the same, forcing to manage and centralize it by software and hardware level. PON
reduces these unnecessary costs.
In conclusion, the use of PON architectures means very important advantages
when designing, installation and subsequent maintenance of the network.

4.2.1. Advantages of PON

Many of the PON properties are given by the use of fiber, and of course, of the
passive elements that compose the network, which added to the specific configuration
of a star or tree give it certain advantages over other topologies. This gives to PON two
undoubtedly important advantages: cost savings in implementation and the capacity
and bandwidth of passive optical networks.
However, these advantages are not the only ones, and among other, the most
relevant are listed below:

A PON allows for longer distances between central offices and customer
premises. While with the Digital Subscriber Line (DSL) the maximum
distance between the central office and the customer is only 18000 feet
(approximately 5.5 km), a PON local loop can operate at distances of
over 20 km [16].

There is the possibility of providing every information source in a
different wavelength, avoiding the mix of signals to each other, and
facilitating diffusion from the OLT to the different ONTs. Therefore,
signals voice and data are managed by so-called P-OLT, which operates
in second window wavelengths, and video signals in diffusion managed
by the so-called V-OLT, operating in third window wavelengths. This fact
gives scalability to PON transmission system, given the variety of
wavelengths to use for the same by CWDM / DWDM.

To this, it adds the reduced cost of network deployment in the outside
plant. The use of passive elements in the network supposes a lower cost
of implementation. On the one hand it reduces the cost of installation of

55

active elements, and on the other hand the cost of passive element
itself, which is much lower.

The installation of PON from these elements is much more economic,
and prevents operation and maintenance costs, such as absence of falls
or maintenance of the network feeds.

Finally, it is noteworthy that the high bandwidth allowed by systems
based on PON architectures which can reach the 10 Gbps rate down to
the user. The need to increase the bandwidth and the speed is
nowadays just another justification for the use of PON. This is an
essential support for services such as HD Video, services called "on
demand"...

4.2.2. Disadvantages of PON

Despite the many advantages that have the PON to own intrinsic configuration,
there are some disadvantages associated with it. However, there are not significant
enough to avoid choosing PON as the best possible configuration.
One of the first disadvantages to be considered is that caused by distribution of
information from the OLT to the different ONTs. The fact that a divisor distributes
information from the OLT to all ONTs that are connected to the same stage or
distribution tree, it causes a reduction in network efficiency.
The total capacity is divided into many ONT connected to the splitter, so that
the efficiency of the channel is lower than in a point-to-multipoint link. In addition,
because PON has a preset speed, it is forced to work at that speed but providing
different speeds to the customer service. For example, an ONT that provides 100 Mbps
to the customer is required to work at higher speed rates: 1.25 Gbps or 2.5 Gbps.
Moreover, the fact that all information flow through the same physical channel
increases the likelihood of sniffing on the network, losing security, and forcing to
establish a high level of encryption.


56

Regarding security, PON architecture is sensitive to external sabotage. This
problem is given by the nature of the transmission medium itself. Injection of constant
light to a particular wavelength masks all communication and service tends to fall.
Another important aspect is the fact that a stage or distribution tree, depend
exclusively on a single OLT. A fault in the OLT header supposes a high impact on the
network, since all the ONT and splitters connected to it are affected. However, the
installation of few OLT supposes a cost reduction of network deployment enough
considerable.
The ONTs of PON are quite sensitive to the level drops, and in many cases, the
power budget of the network is quite limited. This budget is directly related to:
The capacity of splitters. A greater number of users, less power reach
everyone from the OLT.

Maximum distance to achieve. The greater the distance between the
OLT and end users, less power will reach the corresponding ONTs.

However, despite the disadvantages mentioned above, the most advantageous
configuration for the deployment of FTTx is PON. Two of the most important
conditions that justify the use of this architecture are:
The economic savings resulting from deploying PON networks regarding
other two configurations (point to point and active optical network).

The flexibility of the network, which allows the usage of a channel by a
large number of users.


57

4.3. PON technologies

Passive Optical Networks are a family of networks (xPON), whose origin is
found in a first network that was defined by FSAN, a group of 7 telecommunication
operators with the aim of unifying the specifications for broadband access to
households. Below is an evolution of PON standards since its inception.

4.3.1. APON

It was the first network that was defined by FSAN (Full Service Access Network).
APON bases its transmission downlink in bursts of cells ATM (Asynchronous Transfer
Mode) with a maximum rate of 155 Mbps shared between the ONT numbers that are
connected. His initial problem was the limitation of 155 Mbps which later was
increased to 622 Mbps.
In each ATM cell is introducing two more cells (PLOAM), responsible of indicate
the addressee of each cell and for maintenance. These networks are referred to as
APON (ATM Passive Optical Network), and are standardized under ITU-T standard
G.983.1.
4.3.2. BPON

Under the same standard, ITU-T G.983.x (ed. 2005), also came up the so-called
BPON (Broadband Passive Optical Network). It emerges as evolution of APON, given
the speed limitation of the same.
BPON networks are also based on ATM cell transmission, but they have the
difference respecting to APON because they can support other broadband standards.
In its first version, BPON networks were defined under a fixed rate of 155 Mbps
transmission for both uplink and downlink. However, later was amended to admit
asymmetric channels:
Downlink: 622 Mbps
Uplink: 155 Mbps

However, despite improvements over APON, they had a high cost of
implementation, as well as various technical limitations. Thus, it has been moving
slowly to solve the problems posed by this technology, which today allows asymmetric
speeds of up to 1.2 Gbps in the following way [17]:

58

Downstream Upstream
155 Mbps 155 Mbps
622 Mbps 155 Mbps
622 Mbps 622 Mbps
1,244 Gbps 155 Mbps
1,244 Gbps 622 Mbps
Table 2: BPON downstream/upstream bit-rate combinations

On the other hand, apart from support different transmission speeds allowing
transmit all the information (upstream and downstream) on 1 or 2 single-mode fibers
(as ITU-T G.652 standard), with a maximum range of 20 km between optical splitter
and ONT, and between ONTs of the same stage.
Wavelengths of work that sets the standard BPON, they vary depending on
whether you use 1 or 2 fibers for each ONT, although for both sets a dedicated
wavelength to video broadcast from the OLT to the ONT, being this different from
those used in the transmission of voice and data. The wavelengths are:

For 1 fiber per ONT, sharing upstream and downstream:

Downstream channel: =1480-1500 nm
Upstream channel: =1260-1360 nm
Video: =1550-1560 nm

Figure 18: Wavelength allocation plan in ITU-T G.983.3


59

For 2 fibers for each ONT, one for upstream and one for downstream:

Video: =1550-1560 nm

BPON networks support a maximum ratio of 32 splitters per OLT, and each
splitter supports up to 64 outputs to ONT users. This brings a total of:

32
64

2.048

4.3.3. EPON

Parallel to the evolution of PON, standardized by the ITU and that have their
origin in the FSAN, there is a new specification made by the working group EFM
(Ethernet in the First Mile), established by IEEE. The intention of the FSM in this sense
was to take advantage of the characteristics of optical fiber technology in PON and
apply them to Ethernet. In this way, they created the standard EPON (Ethernet PON)
under the rule IEEE 802.3ah (ed. 2004) that nowadays is still in development.

The EPON architecture is based on the transport of Ethernet traffic, but
maintaining the characteristics of the IEEE 802.3 specification, and therefore, leave
aside the transfer of ATM cells, in which APON and BPON standards are based and
encapsulates information on Ethernet frames. This allows providing EPON the
following advantages over APON and BPON standards:

It allows working directly to Gbps speeds because of being supported on
Ethernet. This flow is not single-user since it has to be shared among
many users (ONT) as the system has.

Interconnection between EPON stages is simpler.


60

Certain costs are reduced arising from the non-use of ATM and SDH
elements, typical of the above networks.

As transmission speeds are concerned, EPON establishes a symmetric speed
line of 1 Gbps, both for downstream and upstream channels. This technology is also
called GEPON (Gigabit Ethernet Passive Optical Network) due to work at speed of
Gigabit.

It is noteworthy that standardization only allows transmission in downstream
and upstream under only 1 single-mode fiber (according to ITU-T G.652 standard), with
a maximum range of 10 km between splitter and ONT, and between ONTs of the same
stage (there is provision for extending the distance to 20 km in both cases).

The EPON standard establishes a dedicated wavelength for the broadcast of
video from the OLT to the ONTs as the BPON standard, being this different that those
used in voice and data transmission. The wavelengths are:

Video: =1550-1560 nm

EPON supports a maximum ratio of 16 splitters per OLT, and each divider
supports a maximum of 64 outputs to ONT users. This brings a total of:

16
64

1.024

In this case, the standard supports less number of ONT users connected and in
service to the same OLT, regarding to BPON standard. Exactly, the difference between
both would be:

2.048
1.024
2


61

However, despite the EPON standard involves reduce the number of ONTs
connected to the same OLT, the speed ratio is increased in the same proportion
considering BPON in its basic configuration, as follows:

1244
622
2

Finally, it is noteworthy that despite being EPON a standard that allows greater
speed than BPON, it does not reach more since the maximum distance between the
OLT and ONTs keeps the following proportion:

20
10
2

4.3.4. GPON

Today, the more advanced standard on which is still working, it is who is born
from the evolution of the BPON. To work better with changes in communication
technologies and to meet rapidly growing demand, ITU-T created the series of
standards ITU-T G.984.x for Gigabit capacity PON, which were the basis of the standard
GPON (Gigabit PON).
GPON allows transmission of information encapsulated in two technologies:
ATM, just likes in the case of the BPON standard, but improved.

Ethernet or TDM, using for that the GEM (GPON Encapsulation Method)
based on GFP (Generic Framing Procedure).

Improvements that GPON offers respecting all of its previous standards is, in
general, increasing the bandwidth in transmission and providing security to the own
network by protocol level.


62

Thus, GPON allows varied transmission rates in the range of between 622 Mbps
(as his predecessor BPON) to 2,488 Gbps in the downstream channel. Like BPON, this
standard allows data transmission both symmetric and asymmetric where rates of
transmission for each one are:
Symmetric transmission, flow rates between 622 Mbps and 2,488 Gbps
for downstream and upstream channel.

Asymmetric transmission: Different flow rates for downstream and
upstream channel:

Downstream channel: up to 2,488 Gbps.
Upstream channel: up to 1,244 Gbps.

The fact of allow a very high bandwidth, enables the transmission of practically
any multimedia information and supports any operator service. Moreover, given their
full service support (either via ATM or over Ethernet and TDM), makes him global
support multiservice:
Voice transmission.
Ethernet 10/100 Base-T.
ATM Service.
Leased lines.
Wireless extension
Etc.

The wavelengths of work that sets the GPON standard vary depending on
whether you use 1 or 2 fibers for each ONT, although for both sets a dedicated
wavelength for video broadcast from the OLT to the ONTs, being this different from
those used in the voice and data transmission. The wavelengths are the following:
For 1 fiber per ONT, shared for transmission and reception:

Video: =1550 nm


63

For 2 fibers for each ONT, one for transmission and another one for
reception:

Video: =1550 nm

Figure 19: G-PON Access System
[http://policom.dei.polimi.it/education/comunicazioni/lezione-PON.pdf]

GPON supports 128 splitting ratio per OLT, and each splitter supports a
maximum of 64 outputs to ONT users. This brings a total of:

128
64

8.192

Comparatively, the GPON standard allows increasing significantly the number
of ONTs connected to the same header OLT. The proportion respecting to BPON and
EPON standards is:

8.192
1.024
8

64

8.192
2.048
4

The difference, as you can see, is notorious among them, which suggests that
GPON technology allows minimizing the cost of deployment, as it will need less
number of installed OLTs to cover a given area.
Regarding to transmission rate, GPON also increases the data transfer speed,
being the ratio between the three standards the following:

2,488
622
4

2,488
1,422
2

Not only can transfer data faster, but also provides increased reliability since it
has a secure transmission protocol level, introducing fields in ATM/Ethernet for this
mission.
Finally, it is remarkable the increase in the distance between the OLT header
and different ONTs, which is greatly increased at the rate of the following ratio:

60
20
3

60
10
6

Concluding the analytical data comparison between some standards and
others, it can define the GPON standard as a standard up to 4 times faster, allowing up

65

to 6 times the distance between OLT and ONTs, and supporting up to 8 times more
users than the other standards.
4.3.5. Other PON technologies: NG-PON

4.3.5.1. WDM-PON [27]
Thought by many to be the next thing in FTTx, WDM-PON (Wave Division
Multiplexing Passive Optical Network) provides the dedicated bandwidth of a point-to-
point network with the fiber sharing inherent in PON networks. Not a standard today,
WDM PON is not still fully defined but carriers and vendors expect up to 32 subscribers
to be served by a single WDM-PON access fiber.
The WDM-PON architecture is identical to that of EPON and GPON (and their
10G versions) FTTx architectures. However, instead of a splitter, WDM-PON usually has
an Arrayed Waveguide (AWG) filter that separates the wavelengths for individual
delivery to subscriber ONTs.
Advantages of the WDM approach are inherent in the completely separate
downstream wavelengths for each subscriber. This provides more bandwidth to each
subscriber, more security, and better operational control since there is no interference
in the downstream direction among the various wavelengths.
The biggest disadvantage of WDM-PON is its cost (and its maturity as a FTTx
technology). Each subscriber requires its own dedicated transceiver at the OLT, and
this is one of the cost advantages of the other PON technologies lost with WDM PON.
Additionally, the AWG filter is expected to be more expensive than the splitters used
with GPON, EPON, and BPON.

The diagram below shows a typical WDM-PON network:

66

Figure 20: WDM-PON Network

4.3.5.2. 10G-GPON and 10G-EPON

Applications and services such as HD IPTV, video-on-demand (VoD), voice over
IP (VoIP), video conferencing, online gaming and peer-to-peer (P2P) networking make
that in the future will be necessary to implement new technologies from existing
standards. The most recent standards in development are currently under way at the
IEEE P802.3av (10G-EPON) and ITU-T/FSAN (10G-GPON). This NG-PON (Next
Generation Passive Optical Networks) will be briefly explained, since they are very
recent standards, even still in study and very expensive, so it will be applicable for
future projects.
10G-PON (XG-PON) is a broadband transmission standard defined by ITU (ITU-T
Recommendation G.987), that can deliver shared internet speeds of 10 Gbps over
passive optical networks and it was approved on 2010-10-07. Nowadays it is still
completing parts of this standard.
This is the next generation 10 gigabit standard that follows from GPON (Gigabit-
capable PON) and enables its coexistence with installed equipment of the same
network. In addition, 10GPON can be an extension of current GPON and deployment
could be reused.
The diagram below shows a typical 10G GPON Fiber to the Home (FTTH)
network. In 10G GPON, the downstream rate on the fiber is 10 Gbps. Upstream could
be 1.25 (current GPON upstream rate, not shown), 2.5, or 10 Gbps, and the higher the
upstream rate, the higher the cost of the optics in the Optical Network Terminals.
Current GPON networks use ONT optics transmitting at 1.25 Gbps upstream, and this

67

will be by far the cheapest alternative for a 10G GPON ONT transceiver, which is
usually the most expensive component in the ONT. Dynamic Bandwidth Assignment
(DBA), as specified in G.984.3, allows over subscription in the upstream direction, and
this eliminates much of the requirement for higher upstream rates. [30]

Figure 21: 10G-GPON Network Diagram

The number one electronic cost in a PON network is the unshared ONTs,
though the cost of installing the fiber optics generally drives the complete cost
equation. There could easily be thousands of ONTs per single moderate-sized OLT. A
critical element to keeping the cost of the ONT down is an inexpensive optical
transceiver, and 10G GPON transceivers capable of transmitting 10G upstream are
considerably more expensive than 2.5G and 1.25G upstream capable transceivers.
Although upstream has been getting a lot of attention lately, the cost of the ONT gets a
lot of attention by the carriers, so expect 10G GPON upstream rates on the lower end
initially and for quite a while in FTTH networks.
Finally, this standard can be summarized in its two most common ways:
XG-PON1 (Asymmetric 10G-PON): 10 Gbps downstream and 2.5 Gbps
upstream.

XG-PON2 (Symmetric 10G-PON): 10 Gbps downstream and upstream,
but would require more expensive lasers to deliver the upstream
transmission speed.

68

Respecting to 10G-EPON, it refers to 10 Gigabit Ethernet Passive Optical
Network standard about broadband transmission that was defined by IEEE (IEEE
P802.3av). Unlike 10G-PON, this standard was approved in September 2009.
The 802.3av standard places significant emphasis on the coexistence of 1 Gbps
and 10 Gbps EPON systems on the same network. The diagram below shows a 10G-
EPON access network.

Figure 22: 10G-EPON Network Diagram

Downstream on 10G EPON is broadcast Ethernet frames and Idle characters.
Upstream uses Time Division Multiple Access (TDMA) with each ONT transmitting in its
own timeslot, which can vary in size over time to optimize upstream. As burst
transmissions from different ONTs can have different line rates (due to different bit
rate between 1Gbps to 10Gbps), this method is termed dual-rate TDMA.
In the downstream direction, the 1 Gbps and 10 Gbps channels have different
wavelengths, with 1 Gbps transmission into 1480-1500 nm band and 10 Gbps
transmission limited to 1575-1580 nm band. In the upstream direction, the 1 Gbps and
10 Gbps bands overlap. 1 Gbps band spreads from 1260 to 1360 nm and 10 Gbps band
uses from 1260 to 1280 nm band. This allows both upstream channels to share the
same spectrum region characterized by low chromatic dispersion, but it requires the 1
Gbps and 10 Gbps channels to be separated in time domain.


69

As the 10G-PON standard, this standard supports two configurations:
Asymmetric: operating at 10 Gbps in the downstream direction and 1
Gbps in the upstream direction.

Symmetric: operating at 10 Gbps data rate in both upstream and
downstream directions.
Asymmetric 10/1G-EPON appear less challenging then the symmetric option, as
this specification relies on fairly mature technology. The upstream transmission is
identical to that of the existing 1G-EPON (as specified in IEEE Std. 802.3ah), and will
rely on field-proven and mass deployed burst-mode optical transceivers. The
downstream transmission, which uses continuous mode optics, will rely on the
maturity of 10 Gbps point-to-point Ethernet devices. [31]
The main driver for 10/10G-EPON is the necessity to provide adequate
downstream and upstream bandwidth to support the MDUs. When deployment
strategy is MDU configuration, one EPON ONT may be connected up to thousand
subscribers. [31]


70

4.3.6. Comparative synthesis among xPON standards

In the previous sections it has made an analytical data comparative obtained by
different standards xPON. These data can be condensed and completed in the
following table:

Features BPON EPON GPON
Standard ITU-T G.983.x IEEE 802.3ah ITU-T G.984.x
Transmission rate
(Mbps)
Down: 155, 622, 1244
Up: 155, 622
Down: 1250
Up: 1250
Down: 1244, 2488
Up: 155, 622, 1244, 2488
Fiber type
Single-mode (ITU-T
G.652)
Single-mode (ITU-T
G.652)
Single-mode (ITU-T
G.652)
Number of fibers per
ONT
1 or 2 1 1 or 2
Operating
wavelengths
For 1 fiber:
Down: 1480-1500 nm
Up: 1260-1360 nm
Video: 1550 nm

For 2 fibers:
Down: 1260-1360 nm
Up: 1260-1360 nm
Video: 1550-1560 nm

Down: 1480-1500 nm
Up: 1260-1460 nm
Video: 1550-560 nm
For 1 fiber:
Down: 1480-1500 nm
Up: 1260-1360 nm
Video: 1550-1560 nm

For 2 fibers:
Down: 1260-1360 nm
Up: 1260-1360 nm
Video: 1550-1560 nm

Max. Number of
splitters per ONT
32 16 128
Max. reach (OLT-ONT) 20 km 10 km (prev. 20 km) 60 km
Max. Distance among
ONTs
20 km 10 km (prev. 20 km) 20 km
Max. insertion losses 0 dB 15/20 dB 15/20/25 dB
Traffic mode between
OLT and ONT
ATM Ethernet ATM, Ethernet, TDM
Transmission
architecture
Symmetric, asymmetric Ethernet (symmetric) Symmetric, asymmetric
Comments
Improvement of APON.
Support for other
services than just ATM-
based

Very popular in Asia.
GE-PON is a special
case of EPON

Upgrade of BPON.
Improved OAM&P.
Dominating in the US

Table 3: Comparative table xPON standards
Chapter 5 Design of a WDM GPON

71

Chapter 5
Design of a WDM GPON

5.1. Considerations of designing and planning of a FTTx Network

Throughout this block, the most interesting and important parameters to take
in account in planning FTTx networks will be described. However it will not carry out a
comprehensive study of the network at all levels, but which only will make the physical
level of the network as a passive configuration and services that this PON will provide
to end users. Therefore, it will execute the design of a pure PON, not including active
network elements.

5.1.1. Layers of an FTTx network [32]

There is no defined standardization of the level structures and functions of FTTx
networks, except physically. Some entities have developed certain characterization of
FTTx networks including structural levels similar to OSI levels. But in the case of this
thesis and because the design includes the development of network infrastructure, it
will opt for the layered structure given by the FTTH Council Europe.
An FTTx network can be considered to have different layers: the passive
infrastructure comprising the ducts, fiber, enclosures and other outside plant; the
active network comprising the electrical equipment; retail services, which provides
internet connectivity and managed services like IPTV; and of course the end users.
Some people also visualize an additional layer, the content layer, lying above the retail
services layer and the end users, which can be exploited commercially by so-called
over the top content providers.

72

Figure 23: FTTx Network layers [32]

This technological structure has implications for the way that an FTTx network
is organized and operated. Lets look at these in more detail:

Passive infrastructure comprises all the physical elements needed to
build the fiber network. This includes physical objects such as the optical
fiber, trenches, ducts and poles on which it is deployed, fiber
enclosures, optical distribution frames, patch panels, splicing shelves
and so on. The organization in charge of this layer will normally be
responsible for network route planning, right-of-way negotiations, and
the civil works to install the fiber.

Active network refers to the electronic network equipment needed to
bring the passive infrastructure alive, as well as the operational support
systems required to commercialize the fiber connectivity. The party in
charge of this layer will design, build and operate the active equipment
part of the network.


73

Retail services come into play once the passive and active layers are in
place. This is the layer where basic internet connectivity and other
managed services such as IPTV are packaged and presented to
consumers and businesses. Besides enabling those services technically,
the company responsible for this layer is also in charge of customer
acquisition, go-to-market strategies, and customer service.

Each network layer has a corresponding function. The network owner is in
charge of the first layer, although they may outsource its construction to a third party.
The network operator owns the active equipment, while the retail services are
provided by the internet service provider (ISP).
The traditional telecom model is based on vertical integration, in which one
entity controls all three layers of the network. This is often the case for incumbent
operators, for example, Orange in France, Telefnica in Spain and Verizon in the
United States.
On the other end of the spectrum is the fully separated ownership of the
different layers, as is the case in some parts of Netherlands where Reggefiber controls
the passive infrastructure, BBNed runs and operates the active network and provides
wholesale access, and various retail service providers package the broadband access
with their services and sell directly to the end-users.

5.1.2. Network elements

For the design of an FTTx network or link, it needs to establish and
accomplishing the following parameters:
Maximum transmission distance
Optical attenuation balance for the system
Fiber type
Attenuation for connectors
Attenuation by union or splice
Maximum reflection (backward)
Connector types
Range of aging or mechanical lifetime

74

Throughout the following sections, it will describe the most important
characteristics to decide the appropriate elements to meet the requirements of link
distance, optical attenuation and capacity required.

5.1.2.1. Optical fiber types

The choice of optical fiber is determined by various parameters, being the most
important and restrictive dispersion and optical attenuation. The advantages of single-
mode fiber (high transmission capacity and low optical attenuation) make this medium
as the most attractive for FTTx networks. So, always single-mode fibers are used on
FTTx networks.
The fact of using single-mode optical fibers in FTTH network allows
homogenizing access to the rest of the network which already uses this type of fiber,
giving it transparency and uniformity. It also simplifies the operation, maintenance and
development. Therefore, it recommends the usage of this type of fiber for FTTH access
network, as recommended by IEEE 1000Base-LX and 10GBASE-LX. It can identify as the
best options the usage of fibers G.652D or G.657A and G.657B for FTTH.
Newer fibers which conform to the G.652.D standards eliminate the water peak
attenuation and allow for full operation of all 20 ITU CWDM channels in metropolitan
networks. ITU-T G.652D defines a full spectrum, low water peak fiber with low PMD
and represents the most comprehensive standard for standard single-mode fibers. [35]
But the private service providers are demanding that the cable manufacturers
adhere to ITU-T G.657 standards. This standard describes two categories of single-
mode optical fiber suitable for use in the access networks.
Category A fibers are suitable for transmission in the O, E, S, C and L-band
(throughout the 1260 to 1625 nm range). Fibers in this category have the same
transmission and interconnection properties as G.652.D fibers with improved bending
loss and tighter dimensional specifications.
Category B fibers are suitable for transmission at 1310, 1550, and 1625 nm for
restricted distances that are associated with in-building transport of signals. These
fibers have different splicing and connection properties than G.652 fibers, but are
capable at very low values of bend radius. [35]


75

5.1.2.2. Number of fibers per user

When designing a network, you have to take into account the current needs
and projections regarding future service requirements referred, such as expansion of
basic services, inclusion of alarms, etc.
The total number of fibers per end user or subscriber is determined by the
degree of utilization of active opto-electrical components that exist on the network.
Most point to point systems are based on the use of two fibers per link, one dedicated
to upstream and the other for downstream. But in this project, and given that it will
use the WDM technology, the uplink and downlink channels will travel through the
same fiber but at different wavelengths. This implies that the number of fibers and the
connectors are cut in half, saving cost and space.

5.1.2.3. Connector types

Connectors that are commonly used for national, regional and municipal
networks to connect each other require higher performance and quality that those
used to connect end users in an FTTx network.
In our FTTx network, it can use different types of connectors: SC, LC, MU ...
depending on the price, performance, life expectancy and usage, although there are
some recommendations for choosing one based on the final network, and are listed
below:
Use SC, LC or MU connectors for WDM single-mode systems that have
one fiber per subscriber.

Use any type of duplex connectors for systems with two single-mode
fibers per subscriber.

Use duplex SC, LC or MT connectors for multi-mode systems with two
fibers per subscriber.


76

5.1.2.3. Type of unions between fibers

Fiber connections which have better performance in terms of losses concerns
are the fusion splices, and are the most commonly used today. No but also tend to be
frequent mechanical splices.
A good fusion should typically have a loss of about 0.1 to 0.15 dB of attenuation
for single-mode optical fiber according to ITU recommendation G.652. The attenuation
produced by the union of fibers is not critical in many systems, but in order to ensure a
lifetime of more durable splice, it is recommended to fusion if attenuation is greater
than 0.3 dB.
Moreover, the maximum recommended attenuation for any kind of fusion or
mechanical splice should not exceed 0.3 dB loss. And it is the fusion joints which offer
better performance and fulfilling, as a rule, such a requirement.
Special care is needed when estimating the attenuation by fusion or splices.
The estimate of this parameter is not always accurate, and although is not required the
metric verification of each union, it is necessary to verify the complete attenuation of
the link to study the contribution of each union to the system.

5.1.3. Retail services

Once an overview of the network elements such as passive infrastructure, this
section will deal with a little more detail the type of services that FTTx network can
offer to its customers, which will help later to have a forecast of bandwidth that meets
all their needs.
Retail services can be divided up according to market segments of this project:
Residential
Business
Public sector


77

5.1.3.1. Residential [32]

Typical residential services include:
Basic telephony (using VOIP)
Internet access
IPTV
A retail package that includes all three elements is termed triple play.
Retail ISPs may also choose to offer other services for customers, such as web
space, online backup, technical advice and so on. These services may be bundled
together with the basic package.
An increasing number of consumers are demanding reliable, high-speed
broadband connectivity in order to access an expanding range of internet activities,
including online shopping, online banking, school homework, access to public services
online, catch-up TV, online gaming, and many more.
Catch-up TV is an example of over-the-top (OTT) internet application, because
it is available to anyone with internet access. This is distinct from broadcast IPTV
services, which are provided by the retail service provider exclusively to its customers.
Many internet applications also require good upstream data rates. Examples
are back-up to an online data vault or online video calling. The network owner may
also wish to offer specialized services, for example:
A housing association-owned network might offer a central system for
booking landlord visits and building maintenance.
A municipal network might offer local TV broadcasts or views from CCTV
cameras.
A power utility might team with the network owner to offer an
integrated system to implement smart grid and/or home automation
technologies.
The high capacity available from FTTH provides a good platform for delivering
new services, but dont get carried away. Getting the basics right (fast, reliable
broadband) is a good way to secure customer loyalty.
Broadband on its own is a profitable product (in fact it is the most profitable
product for many service providers) according to a study on next-generation access
service portfolios by Yankee Group (2009).


78

5.1.3.2. Business [32]

Many large enterprises are already hooked into optical fiber networks because
they have high bandwidth, high reliability and high security requirements that far
exceed those of the residential sector. As a result of their special requirements, these
large businesses are not usually linked directly to the same infrastructure as residential
customers.
However, research has found that there is a considerable market opportunity
to address the underserved lower end of the business market. Small and medium-size
businesses (SMEs) can easily be served from a typical FTTH network. Even if the
network is consumer focused, it is worth speaking to any businesses that happen to be
within the coverage area.
The technical network requirements for businesses may not be substantially
different for those for consumers, but business users may be interested in tailored
service packages with extra features such as uptime guarantees, lower contention,
higher security, and business-grade customer service. Speak to local businesses to
assess their needs.
Cloud services, and application outsourcing, are becoming popular with
businesses. Its understood that these services will work with high speed symmetric
fiber connections only.
Further opportunities may exist to provide services that require features only
fiber can deliver, for example:
Video conferencing and telepresence are further examples of business
services.
High-frequency trading where participants need lowest latency
connections to stock exchanges.
A virtual orchestra where musicians need low latency connections to
their remotely located colleagues (this could be an interesting model for
the professional music recording business).


79

5.1.3.3. Public sector [32]

Dont neglect the public sector when drawing up your FTTH network plan.
Schools, libraries, hospitals, doctors surgeries and local government buildings all
require connectivity and have expanding requirements as more ICT is brought into the
school curriculum, libraries become digital access points, doctors share patient records
electronically, and governments put more public services online. These organizations
can become anchor tenants on the network.


80

5.2. Planning of the WDM-GPON

This chapter is aimed to implement that all methods of design and planning of
an FTTx system that have been detailed in previous chapters may apply to a fictitious
environment but could be perfectly real. It will require all elements of an optical fiber
system for implementing the recommendations and procedures of planning and design
of an FTTx network and that allows building our passive optical network.
To this end, it proposes a model of a valid neutral operator to rent to any cable
operator without requiring them to invest in physical infrastructure. Given the volume
of work involves the design specification of the entire network, and since the design of
the active part of the network depends heavily on transmission equipment, is intended
as an exclusive design of PON, and therefore a passive design of the FTTx network.

5.2.1. Project description

It proposes a model of passive optical infrastructure for a neutral operator that
puts its fiber optic network available to different cable operators in the market. This
model consists of deploying a WDM GPON from a central telecommunications to
different areas where end users will enjoy the services that only optical fiber network
can provide.
The project will see the following information:
Design and planning of the chosen network architecture for deployment
of FTTx network based on WDM PON networks and the feasibility study
system.
Supply all materials and components necessary to deploy a passive
optical infrastructure of an FTTx network, which includes from OLT to
the optical jack (ONT) in the house.
The supply and installation of active equipment will be provided by companies
which provide cable service, as well as the operation, service and active maintenance
of the network.


81

5.2.2. Access network design

To perform an optical infrastructure deployment in the best possible way, is
essential to know the scenario of deployment, its particularities, highlighting the most
important features, and so on.
The area under deployment is a fictitious environment as an expansion of a city
but it could easily belong to a real environment. Here we can clearly see 5 distinct
areas that will need a new generation network for meet their needs:
Hospital area: in that it can find a huge hospital that covers the whole
region and have all the latest technological advances.

Business offices area: it will find 2 buildings with a total of 30 offices
that will give service to several companies.

Residential area: it is a distribution of 20 single-family homes that they
will require a full suite of triple play service.

Mall and school area: next to the residential area it will be located a
school and a shopping center that at least will enjoy access to
broadband internet and HDTV.

Buildings area: Finally it will find another residential area similar to
above but in the form of apartment buildings. In that it will be located 2
buildings employing a total of 22 apartments and they will also need
triple play service.


82

Below it displays a virtual picture of the 5 areas that it will give service through
passive FTTx networks.

Figure 24: Deployment scenario

Having seen the overall design of that it will be our passive optical network, it
will divide the design into 5 sections that correspond to the 5 areas to be designed.


83

5.2.2.1. Hospital area

The area where it will place the hospital is located 3 km from the SPLITTER and
therefore a total of 8 km from the central office. It will arrive to this area a link that
would support GPON up to 1,244 Gbps symmetric and will provide LAN and Wi-Fi
access in all major points of the hospital. This will cover the required services such as
broadband Internet, HDTV, VoIP, telemedicine and diverse applications for real-time
control.

Below is shown a drawing of the area:

Figure 25: Hospital area


84

5.2.2.2. Business offices area

This is an extension of the business area, where it will build two office buildings,
in which each one will have 15 offices, making a total of 30 offices.
Each of these offices will require services such as broadband internet, HD
Videoconference, FTP, VoIP, cloud computing ... with that today and for future
forecasts, 75 Mbps would be enough.
Therefore, he bandwidth required for these 30 offices would be
75
30 2250 , so a FTTB design with a 2.488 Mbps

symmetrical GPON link would be sufficient to cover all needs.

Note below the drawing of the extended business area:

Figure 26: Business area


85

5.2.2.3. Residential area

The residential area will have 20 single-family houses, located 1 km from the
optical SPLITTER and 6 km from the OLT, which will demand last generation triple play
service. These services may include:

Broadcast:
2 HDTV channels: 2 * 15 Mbps = 30 Mbps
1 3D channel: 50 Mbps

Real-time applications:
HD videoconference: 7 Mbps
Online games: 1 Mbps
Home automation, VoIP, audio 2 Mbps

Broadband internet:
Web, mail, P2P 10 Mbps

This will make a total of 100 Mbps per household and therefore it will need a
bandwidth of 100
20 2 to provide this service to all users.

With this in mind, it just needs a GPON link with 2,488 Gbps downstream and
1,244 Gbps upstream.
Next picture shows a schematic of residential area with the 20 households:


86

Figure 27: Residential area


87

5.2.2.4. Mall and school area

It will now describe the zone next to the residential area, where it will locate a
school and a shopping mall. This area is also 1 km from the optical SPLITTER and
therefore 6 km from the central office. Speaking about services, it will be required
broadband internet access from a local area network (LAN) in different classrooms of
the school, apart from VoIP, and broadband Internet with Wi-Fi service that will be
given to the various shops and restaurants in the mall, apart VoIP and HDTV.
The bandwidth of all this depends on the number of stores that want to have
the service, but with a symmetric 1.244 Gbps GPON connection will be enough to
supply the entire area.
The infrastructure development shall be shared from the optical fiber to the
splitter that will be located in the area and it will distribute the traffic to the two
buildings which will be equipped with LAN access at school and other LAN and Wi-Fi in
the mall. This can be seen in the following scheme:

Figure 28: School-Mall area


88

5.2.2.5. Buildings area

Finally is the last area that will serve our passive optical network. This is the
extension of buildings area in which it will be built two buildings: the first with 5 floors
that will house 10 homes, and the second with 6 floors and 12 apartments.
Talking about the services they require, it will be the same as residential users,
i.e. next generation triple play service that it will consist of:

Broadcast:
2 HDTV channels: 2 * 15 Mbps = 30 Mbps
1 3D channel: 50 Mbps

Real-time applications:
HD videoconference: 7 Mbps
Online games: 1 Mbps
Home automation, VoIP, audio 2 Mbps

Broadband internet:
Web, mail, P2P 10 Mbps

Each of the apartments will need 100 Mbps and therefore a total bandwidth of
100
22 2,2 to provide all users.

This can be solved through access to FTTB consist of one GPON link with 2,488
Gbps downstream and 1,244 Gbps upstream. With one splitter the signal will be
divided and it will reach the corresponding MDU, which will distribute the connection
throughout the building via a LAN. The following scheme will see the building area:


89

Figure 29: Buildings area

5.2.3. Choice of network elements

Once it has seen in detail all the scenarios that make up the network and the
services that end users will need, it's time to talk about the elements that will be used
to develop the design. In Figure 26 it can remember the deployment of our GPON
design, viewing in the following figures these areas in detail.

5.2.3.1. OLT

This is one of the most important elements of the network, which is in charge
of distributing the traffic to clients and it performs router functions to provide all the
services demanded by users.
Unlike other equipment on the network, the OLT and the different ONTs are
the only active components, so it will be the operator who chooses these elements.

90

However, as network designer, it can choose from several options to locate the
OLT, each of which has advantages and disadvantages:
Core Central Office:
Long distance coverage.
Easy operation and maintenance.
Low density of users.

Central Office:
Better radio coverage.
Easy operation and maintenance.
Balance of resources (connections, fiber optic cables,
ducts, feed, etc.).

Outdoor Cabinet:
Saving fiber optic cables.
More complex maintenance.
Reduced coverage radio.
Electric power needs.
On-site administration.

With this in mind, the only OLT it needs will be located in the central office
allowing a better radio coverage and much cheaper operation and maintenance costs
than in outdoor cabinet. Furthermore, the central office is more appropriate because
in it are distributed most of the communications of the city.

5.2.3.2. Optical fiber cables

As discussed in paragraph 5.1.2.1., optical fibers that meet the ITU G.652D and
ITU G.657 recommendation are the most appropriate for our design, since they
eliminate the water peak and therefore they can be used throughout the whole band
from 1260 to 1625 nm range, fundamental requirement for this CWDM design.
Therefore, ITU G.652D fibers will be used in all our design, except for residential
area and the last school-mall access, where it will use ITU G657A fibers. This is because
they are compatible with ITU G.652D fibers and in the last part of the network, space
requirements and manipulation of the fiber are much more stringent and therefore

91

these last fibers (ITU G657A) are optimized for the curvatures and this way they offer
fewer losses.
After selecting the optical fiber, the next step is to choose the type of cable that
will support these fibers.
To make a design more economic, it will proceed to the installation in micro-
trenches because it allows reducing the impact of civil works in the installation of new
cables and it can cost 1/3 of a normal channeling. It consists of making a very small
ditch which can varies between 1 and 10 cm wide and 10 to 20 cm in depth depending
on the diameter of the cable to install.

Figure 30: Installation in micro-trenches
[http://www.telnet-ri.es/soluciones/cable-fibra-optica-y-componentes-pasivos/microzanjas-para-la-instalacion-
de-fibra-optica/]

Therefore, ESP-coated optical cables will be used because they are designed for
underground installation of outside plant and are particularly suitable for direct
installation and buried in trenches or micro-trenches. These cables are armed with
steel tape, giving them special characteristics of mechanical strength.

92

Figure 31: ESP fiber optic cable
[http://www.telnet-ri.es/productos/cable-fibra-optica-y-componentes-pasivos/cable-de-tendido/esp/]

Construction details for ESP cable are the following:
1. - Dielectric Fiber-Glass Reinforced Plastic Central Reinforcement Element
(CRE).
2. - PBT Loose Tube Assets, containing optical fibers and passive tubes wired in
S-Z around the E.C.R. and coated with water-blocking material.
3. - Copolymer corrugated steel as a reinforcement element for traction and
rodent protection.
4. - Polyethylene coating.

Being clear about what type of optical fiber is optimum for our design, and the
appropriate cable for the installation mode of the network, it just needs to know the
number of fibers to use for the deployment and the length of these. It should be noted
that when it talks about optical fiber it will correspond to G652D optical fiber with ESP-
coated cable or G657A optical fiber with ESP-coated for the latest section in the

93

residential area and school-mall access. Therefore, they are going to be described
starting with the OLT and ending in the ONTs of each zone.
The first section of the network will correspond to the optical fiber from the
OLT and connects it to the optical splitter. This fiber will have a longitude of about 5
km.
From the optical splitter will leave 8 optical fibers, 6 of them will distribute all
the information to the respective areas, and the remaining two fibers that are
redundant will go to hospital and the business area, and will only be used in case of
fault.
Two of the fibers will reach the hospital and will have an approximate length of
3 km. The two fibers will be connected to a MDU (Multi-Dwelling Unit) through a
multiplexer, which will let pass the signal through the first fiber, and will only connect
with the second in case of failure of the first. Thus, the hospital will not lose access to
all services. Finally, from the MDU will leave a local area network that will connect with
all the necessary points of the hospital.
Two other fibers of about 1.5 km long will reach the business area and each one
will be connected to an optical splitter to split the signal into other two to reach the
MDUs that will be located in each of the two buildings. This last section will be about
100 meters. In each of the MDUs also it will connect a multiplexer in the same way as
in the case of the hospital. It is recalled that is being designed in this way to ensure the
connection in the event of a problem with the optical fiber. As in the case of the
hospital, from MDU it will serve to the offices throughout local area network.
Now located at the residential area, it will use a fiber optic cable of 1 km (which
will also come from the optical splitter) to reach the area. At this point, it will place a
splitter to split the signal in another 20 to reach all customers in the area. Therefore, it
will need 20 optical fibers of different lengths. The first two cables will have a length of
50 meters and the following will increase its length by 15 meters (since this is the
distance between two consecutive houses). In summary it needs: two cables of 50, 65,
80, 95, 110, 125, 140, 155, 170 and 185 meters, which make the 20 cables needed to
supply all customers. Each of them will arrive to a different home, and will be
connected to their respective ONT.
The school-mall area will be easier. In that will only reach a 1 km optical fiber
which will be connected to a splitter, and so divided into two other signals that will
reach each one a different building. Therefore, it will need two cables, one of 100
meters to reach the school and another of 250 meters to get the mall. On these two
points will be placed two MDUs to supply via LAN and Wi-Fi both the shopping mall
and school.

94

Finally, the last will be the buildings area, as it has seen in previous sections it
will need three optical fibers to supply all customers. The first one will be about 2 km
and the other two that will reach the MDU of its corresponding building will have 50
m. As in previous cases, the MDU will distribute the information to each of the
apartments by copper and coaxial cable.
It is important to consider the reserve of cables. This means that for each of the
optical fibers to be implemented, it always will leave a reserve of about 5 or 10 meters
of cable in case of it needs to move any network element (rack, splitter, ONU ...) it has
a sufficient provision of cable to make the move possible. Obviously, this extension
cable will mean negligible losses.
To facilitate understanding, it has drawn a diagram with all optical fiber cables
and their lengths:

Figure 32: Stages of division


95

5.2.3.3. Connectors

Fiber optic connectors are an important element within the range of passive
devices, allowing the alignment and joining of two optical fibers and in the best
possible conditions. A poor union can be transformed into high losses, and therefore
may cause that the actual network application does not conform to previously design.
In the previous paragraph has been described the type of fiber and cables to be
used for the deployment of passive optical network. It is time now to talk about the
connectors that allow joining the fibers with the other equipment on the network.
It will choose SC (Standard Connector) type connectors because its square
profile allows more easily attach the fibers by a push-pull mechanism that protects the
ceramic ferule. The SC connector is the most popular in both LAN and transport
networks.

Figure 33: SC connector
[http://www.timbercon.com/SC-Connector.html]

It is also important to discuss about the finish of the ferules in the optical
connector. The original fiber connector is based on physical contact (PC connector),
through which the two fibers are connected on a flat surface to avoid air between two
flat surfaces. But nowadays the polishing APC (Angled Physical Contact) is used, in
which the end of the ferule is polished with an angle of 8 in the result that much of
the reflections that do not match the main mode will divert to the coating. In this way
it can get a return loss of -65 dB.

96

In some parts of the design as might be the splitters, it may choose to use
connectors or splices. The splices tend to have better performance in terms of losses,
but hinder the mobility when connecting and disconnecting the fiber. Therefore, in the
design SC connectors with APC polish will be used because that will allow greater ease
of usage in case of changes or testing of the network.
Being clear about the type of connector and the polish which it will use, it can
only find out how many connectors it needs to deploy the network. Viewing the above
scheme (Figure 34) it can deduce that 36 optical fibers will be installed, so it will need
two fiber connectors to connect the ends. This will make a total of 72 SC connectors.

5.2.3.4. Optical Splitters

In order to expand the access network, it will be used optical splitters or
dividers.
A splitter allows the derivation of the optical signal by two or more different
fibers. They are designed to introduce insertion losses approximately equal in all its
output branches and there are several techniques for the manufacture of splitters:
Fused bi-conical taper (FBT).
Planar Lightwave Circuit (PLC).
FBT is an old technology that is based on the fusion of two fibers coupling with
a specific region of the wavelength range that has more losses compared with PLC
technology, although both can be used in PON.
The PLC Splitters (Planar Lightwave Circuits) are based on optical waves
developed on a silicon substrate. The silicon substrate is an element that has very good
optical properties. This technology allows construction of high performance optical
splitter with high density of channels and low size, so that will be considered for the
network access design.
Since the splitters involve a significant loss of power in relation to other
network components, the design of a network must be carefully balanced between:
high branching fibers, distances to customers and powers managed by equipments, so
that meet the main specifications.


97

For the location of the splitters, there are two basic options:
Outdoor Cabinet: in which the splitters are placed in lockers located in
strategic places, which has the advantage of lower costs of deployment
and on-site maintenance required.

Basement: the splitters are placed in rooms or closets
Telecommunications existing near of the user, for example, in a
building. It has the advantage of saving fiber optic cables compared to
the inefficient use of PON ports and requiring on-site maintenance.

For this design the splitters will be located in Outdoor Cabinets mainly for the
reason that the expansion of the network is provided placing splitters at strategic
points, and not indoors where the usage of ports would be limited to a small number
of customers.
Finally, it just needs to decide how many splitters will be placed on the network
and what will be the ratio of division:
For the hospital area, no splitter is needed since the same optical fibers
arrive directly to the MDUs.

For the business area, it needs a couple of splitters with division ratio of
1:2 for each of the two fibers can connect the two buildings.

In the case of the residential area, it will need a 1:32 optical splitter to
encompass a division of 1 to 20. Therefore, this splitter will be the
minimum necessary (it can find 1:2, 1:4, 1:8, 1:16, 1:32, 1:64 even 1:128
division ratios, but it's strongly recommended the minimum ratio
possible to get the minimum losses).

In the mall-school area is needed another 1:2 splitter to distribute the
traffic to the two buildings.
As in the case of the offices, in the buildings area the optical fiber will
need a splitter with division ratio of 1:2 for each of the two fibers that
will reach MDUs.


98

5.2.3.5. ONT and MDU

The ONT (Optical Network Terminal) is the last element of our network. This
charge of optical-electrical conversion oriented to the subscriber obtains the required
services. Must be located in user's home ending the optical fiber and provides the user
interfaces.
As it has talked about the OLT, the ONT is an active device to be selected by the
operator to operate the network. In general, it must be the same manufacturer as the
OLT, although it is working to achieve interoperability between manufacturers, which
would open the market and lower prices.
There are a big variety of ONTs, depending on the services they want to provide
and the interfaces that provides to user. In our case we need ONTs adapted to triple
play service and can provide POTS, Ethernet, HDTV and 3D video.
In the case of buildings such as apartments, offices, schools, malls, hospitals ...
and in general anything that does not mean a single user, it will be used MDUs (Multi
Dwelling Units) to provide more than one user, since the ONT can only serve a
customer.
To complete the equipment of our network, it will just have to see the number
of MDUs and ONTs required:
In the hospital area just needs a MDU to provide service via LAN.

In the offices area, it takes two MDUs, one for each building.

In the residential area, with 20 ONTs (one for each client) will be
enough.

In the school-mall area, it will require two MDUs, one for the school and
one for the mall.

And for the last area, with two MDUs will give service to two apartment
buildings.

This makes a total of 20 ONTs and 7 MDUs to supply all costumers.


99

5.3. Choice of simulation program

Once thought by which technology is going to design the network, the next step
is to choose a simulation program that meets the needs of design and which it allow
covering all aspects.
The simulation using software in the field of telecommunications networks is
now a very useful tool for companies and universities of the sector at the time to
recreate the behavior of a system in which different services and applications are
running in defined conditions the user.
The fact of simulate allows to detect network problems before they are
implemented, test new networking solutions, testing new communications
technologies even including theoretical models that in reality could be difficult to
construct. All this makes it a very useful and powerful tool for the design of
communications networks.
Taking this in mind, it will choose the simulation program Optisystem 9.0 for
the design of our WDM PON network.

5.3.1. Optisystem 9.0 simulation program

Optical communication systems are increasing in complexity very quickly. The
design and analysis of these systems, which typically include nonlinear devices, are
extremely complex and time-intensive, and therefore, these tasks can only be done
efficiently with the help of new advanced software instruments.
OptiSystem 9.0 is an innovative optical communication simulation system that
design, test and optimize virtually any type of optical link at physical layer level, from
video broadcasting analog systems until dorsal intercontinental links.
It is a system-level simulator based on realistic modeling of optical fiber
communication systems. It has a powerful simulation environment and a hierarchical
definition of components and systems. Their capacities can be easily extended by
adding user-defined components, and can be interfaced with a wide range of
complementary tools.


100

A friendly graphical user interface controls the layout of optical components
and network topology, systems, components and graphical presentation. An extensive
library of active and passive components includes realistic parameters dependent on
wavelength. These sweeps parameters are used to investigate the effect of device-
specific data on system performance.

Key Features
The main interface features OptiSystem include:
-Library Components
To be fully effective, the component modules must be able to reproduce the
actual behavior of the device and the purposes specified under selected accuracy and
efficiency. The Component Library includes OptiSystem hundreds of components, all of
which have been validated to deliver care results that are comparable to real
applications.
-Components measured
OptiSystem Component Library lets to manage the parameters that can be
measured in real devices. It also integrates with test measurement equipment from
different suppliers.
-Mixed signal representation
OptiSystem can handle mixed signal formats for electrical and optical signals in
the Component Library. It calculates signals using the appropriate algorithms related
to the simulation efficiency and accuracy cogeneration required.
-Quality and operating algorithms
To predict the system operating, OptiSystem calculates parameters such as BER
and Q Factor using numerical analysis or semi analytical techniques for systems limited
by interference from symbols and noise.
-Advanced visualization tools
These visualization tools produce spectrums OSA, signs with "chirp", eye
diagrams, polarization state, constellation diagrams... It also includes WDM analysis
tools, showing signal strength, gain, noise factor, and optical signal to noise ratio OSNR
per channel.


101

-Data monitors
It can select ports in components for data backup and attach monitors after the
end of the simulation. This allows processing the data after the simulation without
having to recalculate. You can also attach an arbitrary number of displays in the same
port.
- Hierarchical simulation with subsystems
To make a simulation tool flexible and efficient is essential to provide models in
different levels of abstraction, including system, subsystem and component levels.
OptiSystem has a truly hierarchical definition of components and systems, software
tools are used specifically for fiber optic components and integrated at the component
level, and allowing the simulation can be as detailed as desired.
-User defined components
It can add new components and subsystems based on libraries and subsystems
defined by user, or use co-simulation with other tools.
-Multiple layouts
It can create many designs using the same project file allows it to create and
modify designs quickly and efficiently. Each project file can contain multiple versions of
design. Versions of design are calculated and modified separately, but the calculation
results can be combined across different versions, taking into account the comparison
of different designs.
-Reports
A report fully "customized" allows displaying any set of parameters and results
available in the design. The reports produced are organized in reconfigurable
spreadsheets, text, 2D and 3D graphics. This also includes the export of HTML
templates with "layouts" preformatted report.
-Parameter sweeps and optimizations
The simulations can be repeated with an iterated variation of the parameters.
OptiSystem can also optimize any parameters to minimize or maximize any results or
search for objective results. You can combine multiple scans of multiple parameters
and optimizations.


102

- Lists of materials
OptiSystem table provides a cost analysis of the system designed, specified by
the system, "layout" or component. The cost data can be exported to other
applications or spreadsheets.

Applications
OptiSystem design allows automation of virtually any type of optical link in
physical layer, and the analysis of a broad spectrum of optical networks, from long-
range systems to MANs and LANs.
The range of OptiSystem applications includes:
The design of an optical communication system from the component
level to system level in physical layer.
CATV or TDM / WDM network design.
Passive Optical Networks (PON), FTTx.
Optical systems in free space optics (FSO).
Systems of Radio on Fiber (ROF).
Design of SONET / SDH networks.
Transmitter, optical channel, amplifier, and receiver designs.
Estimation of BER and system penalties of different receptor models.
BER for an amplified system and power budget calculation links.


103

5.4. Design of the WDM-GPON and Results

We are now possibly in the most important section of the project, which will
describe the design of the WDM PON.
At this section, the parameters and scenarios of simulation will be shown and
described. The purpose is to show the behavior of links of optical fiber when the signal
goes through all the elements such as optical fiber, splitters, multiplexers and the goal
is to find a good quality of signal in all receivers.
The final goal pursued with this thesis is to evaluate the performance of the
whole system. The parameters used to evaluate this behavior will be the BER and the
Eye Diagram.
The first condition used to assess the performance of the link is generally BER
(Bit Error Rate) at the receiver. Simulations are done trying to approach at the receiver
a minimum BER. Any simulation with a BER<10
-4
at some point of the network will
show that the link is not consistent and reliable.
It has been assumed the use of FEC techniques that will be used for correcting
errors and bringing them from 10
-4
to 10
-12
. These techniques can be for example
Reed-Solomon (255,239) defined at the standard ITU G.709.
The second condition corresponds to the eye diagram. Viewing this diagram it
can describe the quality of the received signal. In the presence of ISI, when the pulse
does not satisfy the Nyquist criterion, the diagram will tend to close vertically. For
error-free transmission in the absence of noise, the eye must be kept some vertical
opening, or otherwise it will exist interference between symbols that will cause errors.
When the eye is not fully closed, the interference between symbols reduces the value
of allowable additive noise. Therefore, the higher vertical opening the greater
immunity to noise.
The following sections will describe the design, displaying the various network
scenarios and analyzing the most important parameters. For that, it will begin with a
comprehensive description of the entire network, and as in previous sections, it will
continue with the detailed analysis of each of the different areas that compose the
design. Finally it will obtain the results of each ONU, which will mean that the network
is viable and therefore its practical implementation would work.


104

5.4.1. WDM PON: Complete design

This is the complete design of our WDM-PON. As reflected in previous sections
it will describe each of the elements with their most significant parameters and which
its function in the design is. To facilitate understanding, it will describe the design from
the OLT to the different ONT's for downstream, and from the ONTs of the different
areas to the OLT for upstream.

It can see the complete design in Figure 34 that is located on the next page.

On the left side of the design you can see the OLT which will transmit
information to different users and it will receive information from the ONT's. All this
information will be transmitted multiplexed at different wavelengths through a single
optical fiber, and then demultiplexed to spread to different areas in downstream, and
multiplexed from different areas in upstream. As explained in previous sections, the
simulation also has been divided in 5 areas.


105

Figure 34: WDM PON Complete design

106

The most important global parameters of the network can be seen in the
following figure:

Figure 35: WDM PON Global parameters

The global bit rate used for the design will be 2.5 Gbps. This bit rate will be used
throughout the design because if the network is working for this bit rate it will work for
all other bit rates as outlined in previous sections because they will be lower than this.
Other important parameters are the sequence length (128 bits) and the
samples per bit (64). These will make a total of 8192 samples and are important
because it needs a large enough sequences for simulate the network at these high bit
rates.
Its also important the iterations of the simulation. In this case, we need 7
iterations for make the design work, because it needs to insert optical delays. These
optical delays are required in some devices when they do not have all signals in the
input ports.

107

Once described the most important global parameters, it is going to describe the
network operation in downstream and upstream.

5.4.1.1. WDM PON in downstream

In downstream, the optical signal will direct from the OLT to the end users
(ONT's).
Returning to Figure 34, the first network element is the optical transmitter
located at the OLT. This laser will broadcast five different wavelengths from 1450 nm
to 1530 nm with a frequency spacing of 20 nm. The transmission power will be 0 dBm
and it will use NRZ modulation.
The next element that is needed is an optical multiplexer, which will multiplex
these five wavelengths to transmit on a single fiber. It will have a bandwidth of 80 nm,
just to cover the five necessary wavelengths. This multiplexer 1:5 (which actually
correspond to a multiplexer 1:8) will have insertion losses of 2 dB.
Given the large losses that the residential area will have due to the great
division that the signal will suffer, it should be placed to place in the OLT an EDFA
amplifier of about 5 dB of gain to ensure good quality signal in the case of major
losses.
In addition, because the OLT will have to handle both the downstream and
upstream traffic, it will insert an optical circulator with insertion losses of 3 dB and
return losses of 65 dB.
Being described the elements of the OLT, the optical signal will be inserted in
the optical fiber which will reach the multiplexer located at 5 km from the central
office. The most important parameters of this fiber, which will carry information in
both directions, are 0.2 dB/km of attenuation, 16.75 ps/nm/km of dispersion and
0.075 ps/nm
2
/km of dispersion slope.


108

Next is attached the spectrum of the signal that has been inserted into the
optical fiber:

Figure 36: WDM PON Downstream spectrum into the first optical fiber

Following the fiber, we can see the SPLITTER block which will contain a
multiplexer and demultiplexer (splitters). Each will be useful in one direction of traffic.
For downstream traffic, the signal from optical fiber will be demultiplexed into 5
different branches, each of them with a different wavelength. This 1:5 splitter will have
starting frequency of 1450 nm, with a frequency spacing of 20 nm. As explained in the
previous multiplexer, the splitter also will be implemented in reality as 1:8 splitter,
therefore it will also have insertion losses of 2 dB.


109

Below is attached the schematic of SPLITTER block:

Figure 37: WDM PON SPLITTER Block

The first branch of the splitter will be directed to the block corresponding to the
HOSPITAL area, the second to the BUSINESS area and so on until the fifth that will be
directed to the BUILDINGS area.

Turning now to the HOSPITAL Block (Figure 38), we find first an optical fiber of
3 km of length, which also will carry information in both directions. Its parameters are
the same as those described above for the other fiber. Note that optical delays that
occur around the optical fiber are required to perform the simulation, since all
network elements need to sign all inputs for make the simulation run. At the initial
time, the value of the input connected to the optical delay will be zero, and it will have
a delay of 1 time interval for the next moments.


110

Figure 38: WDM PON HOSPITAL Block

111

With the Optical Spectrum Analyzer connected to the bidirectional optical fiber,
it can see the spectrum of the signal that will reach the ONU. As we expected, the peak
will be centered in 1450 nm.

Figure 39: WDM PON Downstream spectrum into HOSPITAL Block

The last element of this block will be the Optical Network Unit (ONU) where the
signal will be received and transformed into electrical form to visualize it in the BER
Analyzer. This block is shown in Figure 40 and it will consist in a photodetector PIN
which will transform the optical signal to electrical signal, that will be filtered by a low
pass Bessel filter with a cutoff frequency of 0.75*Bit rate (in this design 2.5Gbps). The
optical transmitter in upstream direction will be explained in the next section.


112

Figure 40: WDM PON ONU Block

Once described the hospital area, we are going to focus on the BUSINESS Block.
In this block it is going to receive the bit rate at 1470 nm. The first element that
it is going to find is a bidirectional optical fiber of 1.5 km in which the parameters will
be the same as the last fiber. Immediately we will find a 1:2 splitter which will divide
the signal into two needed branches to reach the two buildings. The most important
parameter of this splitter is insertion loss. For a splitter with division ratio of 1:2,
typical losses are about 3.5 dB. Since the same division is going to cause a loss of
10log (2) = 3 dB, we will add an insertion loss of 0.5 dB to get 3.5 dB of attenuation in
each branch.
The bidirectional splitter will split the signal into two branches, each connected
to a bidirectional optical fiber of 0.1 km with the same parameters mentioned above.
Finally, each fiber will reach its respective ONU, that in the downstream direction will
be identical to the previous one but it will have some changes in upstream that they
will see in the next section.


113

Figure 41: WDM PON BUSINESS Block

114

Visualizing the Optical Spectrum Analyzer it can see the carrier centered in
1470 nm:

Figure 42: WDM PON Downstream spectrum into BUSINESS Block

The next block to be found is the RESIDENTIAL block, which it will reach a
carrier at 1490 nm to broadcast information throughout the area. The first element we
see in this block is the bidirectional optical fiber of 1 km length, which it will come
connected from the optical SPLITTER seen before and will reach the splitter 1:20 which
will broadcast information to all households. This splitter would correspond in reality
to a splitter with split ratio of 1:32 with a mean insertion loss of 16.5 dB. Since in this
case each branch will have a loss of 10log (20) = 13 dB, it will add an insertion loss of
3.5 dB with a total of 16.5 dB at each branch.


115

Figure 43: WDM PON RESIDENTIAL Block


116

Each of the 20 branches that leave the splitter will be connected to a different
length optical fiber. As explained in previous sections, the two shortest will have a
length of 50 meters and the other will be increased by 15 meters to the longest that
will have a length of 185 meters. The ONU's connected to optical fibers, in the case of
downstream, are identical to those seen previously.
In this case, an optical carrier centered at 1490 nm will reach the RESIDENTIAL
block:

Figure 44: WDM PON Downstream spectrum into RESIDENTIAL Block

We turn now to the block where is going to place the network access that will
serve the mall and the school (SCHOOL-MALL Block). To this part of the network an
optical carrier centered on 1510 nm will reach. As before, the first element is
bidirectional optical fiber connecting the optical SPLITTER with the splitter in the area.
In this case, the fiber will have a distance of 1 km. The optical splitter will have a
division ratio of 1:2 and will be exactly equal to the proposed in the building area
(10log (2) = 3 dB will add 0.5 dB for a total of 3.5 dB insertion loss).


117

Figure 45: WDM PON SHOOL-MALL Block


118

Finally we need a bidirectional optical fiber of 100 meters to reach the first
ONU which will provide service to school and another of 250 meters to arrive at the
second ONU which will offer service to the shopping centre. Below is attached the
carrier centered at 1510 nm which will reach the area:

Figure 46: WDM PON Downstream spectrum into SCHOOL-MALL Block

The last block corresponds to the access network located in the area of
buildings (BUILDINGS Block). The optical carrier that will reach this block is the latest
from the optical SPLITTER and that is centered on 1530 nm. This SPLITTER will be
connected to a splitter with 1:2 splitting ratio using a bidirectional optical fiber of 2 km
of length and the same parameters as the previous. The 1:2 splitter will also be
identical to those above, with a total insertion loss of 3.5 dB. To this splitter will be
connected two fibers of 50 meters long that will connect with the corresponding ONU.
Each will serve a different building.


119

Figure 47: WDM PON BUILDINGS Block


120

The optical spectrum analyzer shows the optical carrier that it was expected
and it is centered on 1530 nm:

Figure 48: WDM PON Downstream spectrum into BUILDINGS Block

5.4.1.2. WDM PON in upstream

In upstream direction, the optical signal will travel from each of the end user
(ONT's) to the OLT. Therefore, we will review the various blocks that define the five
areas and end up explaining the OLT as the receiver. Note that the signal in upstream
will pass through the same elements and therefore they will have the same
characteristics and parameters.
It is also important to note that in each area is going to be transmitted in a
different wavelength. Therefore, TDMA will be implemented in each of the areas for
each user transmits in a determined instant of time. This will avoid that two users
might collide when they wish to transmit information at the same time.
Returning to Figure 34, the first block to be analyzed will be the HOSPITAL
Block. In it we will find the ONU that in this case will transmit at 1270 nm with a power

121

of 0 dBm, NRZ modulation and 2.5 Gbps bit rate. It can see the transmitter of the
hospital ONU in Figure 40. The optical signal will cross the same elements of the block,
but this time in upstream.
The next section consists on the BUSINESS Block (Figure 41). In it we can see
two ONU's which will transmit at 1290 nm wavelength. The parameters of these ONU's
will be the same: 0 dBm of optical power, NRZ modulation and bit rate per default.

Figure 49: WDM PON Business, Residential, School-Mall and Buildings ONU Block

In this case, as it has more than one transmitter operating on the same
wavelength, it must take into account the time division multiple access (TDMA).
Therefore it will use a Dynamic Select Y which it will allow to pass the signal only at a
determined time instant and the rest will be zero. Since we have only two ONU's, we
will use the following formula to define the time interval of each one:
Timeslot * (1/Bit rate) * Sequence length / 2 + Time window / 2
where Timeslot will be 0 in the first ONU and 1 in the second.
The Figure 50 shows the signal transmitted by the two ONU's in the time
domain and therefore proves that the signals transmitted by the various ONU's do not
overlap in time:

122

Figure 50: WDM PON Optical Time Domain Visualizer in BUSINESS Block


123

We turn now to analyze the RESIDENTIAL block (43). In that we will find 20
ONU's corresponding to the 20 homes and which will transmit data with a default bit
rate and a wavelength of 1310 nm. These ONU's will have the same look as before and
will also transmit with a power of 0 dBm and NRZ modulation.

However, to adjust the interval of time that has to transmit each ONU, we must
change the formula in the Dynamic Select Y by the following:

TimeSlot * (1/Bit rate) * Sequence length/20 + Time window/20

where TimeSlot will value 0 for the first ONU, 1 for the second, and so on until
the twentieth ONU which will value 19.

The following figure shows the first three time intervals for the first three
ONU's. Obviously, the others will fulfill the formula above and therefore will transmit
at time intervals unique to each:


124

Figure 51: WDM PON Optical Time Domain Visualizer in RESIDENTIAL Block


125

The block for the school and the mall (SCHOOL-MALL Block shown in Figure 45)
will have features similar to business block described above. The ONU's will broadcast
at a wavelength of 1330 nm and the transmitter will have the same parameters as
above.

Respect to TDMA, we will use other time the described formula in the business
block: Timeslot * (1/Bit rate) * Sequence length / 2 + Time window / 2, where again
TimeSlot will be 0 for the first ONU and 1 for the second.

In Figure 52 is shown a representation of the signal in time domain for the two
ONU's:


126

Figure 52: WDM PON Optical Time Domain Visualizer in SCHOOL-MALL Block


127

We are now in the last block design: the BUILDINGS Block shown in Figure 47.
The two transmitters of this block will have the same characteristics as the
transmitters of the previous block, except that they will send information in a
wavelength of 1350 nm. In addition, the same formula is used to get that the
information transmitted by users do not overlap, i.e., they also will use TDMA.

The following figure shows again the temporal representation of the optical
signal of the transmitters of each block:


128

Figure 53: WDM PON Optical Time Domain Visualizer in BUILDINGS Block


129

With all the signals from the different blocks will be multiplexed the five
wavelengths using an optical multiplexer 5:1 (8 to 1 in reality) which is within the
SPLITTER block seen in Figure 37. This multiplexer will have insertion losses of 2 dB
and a starting frequency of 1270 nm with a frequency spacing of 20 nm. With these
parameters will get multiplexing the five upstream wavelengths that can be seen in the
following figure:

Figure 54: WDM PON Upstream spectrum into the optical fiber

This signal consists of the 5 upstream optical carriers that will be sent through
bidirectional optical fiber of 5 km connecting the SPLITTER Block with the OLT. This
fiber is the same as that described above for downstream, so it will get the same
parameters.
Finally the signal will arrive at the OLT, where it will cross by a bidirectional
optical circulator and will be sent to the optical receiver.


130

Figure 55: WDM PON Upstream spectrum into the OLT receiver

The first element of the receiver will be the Buffer Selector, which will be used
to select only the latest iteration of the simulation, which will be the one with the
correct results. After this element, the signal will pass through the PIN Photodetector
where it will be converted to electrical domain and as in other cases will be filtered
through a Low Pass Bessel Filter with a cutoff frequency of 0.75 * Bit Rate. To end, this
signal will be regenerated in order to be displayed on the BER Analyzer.


131

5.4.2. WDM PON: Results

Having explained the whole design, it just needs to get the results that will
decide if the design works or not, and therefore whether it will be feasible to
implement in a practical case.
The results that are going to expose are the eye diagram, the quality factor, the
probability of error (BER) and the decision threshold as a function of the width of the
bit. Lets see first downstream results and later upstream results.
In the case of the ONU located in the HOSPITAL Block, it has been obtained the
following results:

Figure 56: WDM PON HOSPITAL ONU Results


132

In the BUSINESS block, the path that the signal covers to two ONU's is exactly
the same since the last fiber has the same dimensions. So the results are identical:

Figure 57: WDM PON BUSINESS ONU Results


133

The RESIDENTIAL Block will have 20 ONU's. As expose the results of the 20
ONU's will require a lot of space, we are going to put the results of the closest ONU
(located at 50 meters from the splitter) and the farthest ONU (situated at 185 meters
from the splitter).

Figure 58: WDM PON RESIDENTIAL ONU1 Results


134

Figure 59: WDM PON RESIDENTIAL ONU2 Results

The results are very similar, although as it expected, the results of the ONU
placed on the shortest fiber will be a bit better than the results of the ONU at the end
of the longest fiber.


135

Turning now to the SCHOOL-MALL Block, the results for the two ONU's are:

Figure 60: WDM PON SCHOOL-MALL ONU1 Results


136

Figure 61: WDM PON SCHOOL-MALL ONU2 Results

As in the previous case, the only difference between the two paths is the last
section of optical fiber, which in one case will be 100 meters and in the other 250
meters. Therefore, the results are very similar.


137

The last block in downstream is BUILDINGS Block. As in the case of the business
block, the paths that the signal covers are identical, so both ONU's will offer the same
results:

Figure 62: WDM PON BUILDINGS ONU Results


138

As for the uplink, the results obtained in the OLT are:

Figure 63: WDM PON OLT ONU Results

With all this demonstrates the quality of design in both upstream and
downstream.

Chapter 6 - Summary and Conclusions

139

Chapter 6
Summary and Conclusions

In this concluding section is pretended, on the one hand to summarize the
sections on which it has worked on the thesis, and on the other hand to show the
conclusions of the designed network and which could be the lines that could be
followed for future network improvement.
Throughout the project it has undertaken a number of general guidelines
whose development has resulted in a thorough understanding of a passive optical
network. These basically are:
Theoretical detail and physical basis of the operation of optical
networks.

Planning and design of a PON.

Simulation and obtained results by the designed network.

The detailed theory of optical networks is considered essential for correct
understanding of FTTx networks, as well as necessary to perform a particular design of
passive optical network on which such networks are built. When making any design, it
is essential to know the technology with which it is based as the final solution will
depend on the needs that arise and the features and topology of each network type.
Thus, in Chapter 2 has been detailed the physical basis of optical technology
such as operating principles, composition and geometry, fiber types and the most
important structural and transmission parameters as attenuation, dispersion and
nonlinearities, and it has concluded with the explanation of the advantages and
disadvantages that may have the usage of this technology. Previous understanding of
the methodology of transmission through optical fiber has allowed knowing the
advantages of these systems at the physical level and the limitations that have optical
fiber networks, to take into account in the physical network design phase.


140

Chapter 3 deals with the types of FTTx networks that exist, which are defined
taking into account as far as the optical fiber reaches in the access network. Once
defined these types, it has talked about the point-to-point and point-to-multipoint
configurations. It is important to consider these concepts to define what type of
network it wants to design. In this case FTTB and FTTH networks have been used for
access to our different areas, with a global point-to-multipoint configuration allowing
us to save in optic wiring.
In Chapter 4 has discussed the PON. The first step has been to describe the
operating principles of a network of its kind for both the uplink and downlink, taking
into account all the elements of the passive network. It has made a comparison with
point-to-point networks and active optical networks (AON), concluding PON design
optimal for its exploit of bandwidth and its lower implementation costs. Having
defined the PON as the best option, advantages and disadvantages of these types of
networks have mentioned and it have defined all the standards that exist today (from
APON to next generation networks such as 10G-PON or WDM-PON). Finally it have
done a comparison between these technologies, choosing WDM-PON as the best
option because of its greater speed, reach and number of users supported.
Once we know the structure and characteristics of these networks, the last and
most important chapter has been dedicated to the design, simulation and obtaining
the results of the proposed WDM-PON network.
In the first section has discussed the design considerations of this kind of
network. First, it has defined a 4 levels network standard where we have focused on
the first level (passive Infrastructure) as it is one that covers the design in which the
project is based.
After explaining the services that the different areas will require, it has passed
to the network planning. In this section a description of the project has made,
physically describing each of the areas and analyzing the bandwidth required to meet
the needs of all customers in each area. Having described each area in detail, it has
taken into account all the elements that are necessary for the implementation of the
network, such as the OLT and the ONU's (to be chosen by the operator to manage the
network), optical fiber standards and cable types to use, the optimal connectors for
this passive network and the size and location of optical splitters, arguing in each case
why the election and how many will be needed. Also it has taken into account the
reserves of the cables, that will allow to make movements of the elements of the
network in case of any change or expansion, and provide redundant links to critical
areas such as business area and the hospital area, to ensure as far as possible that the
proper functioning of the service in case of failure or any power outage is guaranteed.

141

Once clear the distribution network and all the elements needed for its
construction, it is important to use a simulation program that meets the needs of
design, covering all aspects. In this case, it has opted for the OptiSystem 9.0 simulation
software, since it meets all our requirements.
The last section focuses on the final design of the network and the results that
show that the network works. The images show what has the design been and an
explanation of all the important elements and parameters that have taken into
account to make it work. To make more understandable the design, the explanation
has been separated into two parts, downstream and upstream, and as in other
sections, the scheme has been divided in 5 areas that define the entire design. Finally,
it has attached the obtained results in each of the ONT's, being these the eye diagram,
the quality factor, the BER and the decision threshold for the minimum BER.
These results have showed a BER of about 10
-12
for the worst case that will be
of the farthest house in the residential area. This shows the high reliability that will
have in each of its points the implement of this network.
The choice of optical fiber as transmission medium allows guaranteeing
longevity of the network during the coming years. The logical and physical design of
the same establishes optimal conditions for the current implementation of the system
but also offer the possibility of future expansion.
The fact of using the Gigabit PON technology on CWDM enables nowadays the
possibility to connect a large number of users. In addition, the specified equipment will
be valid for migration to future standards as 10GPON, DWDM and other technologies
that improve the performance of optical fiber and provide the best possible
performance.
Other features that have the design is the high degree of availability and
network capacity, the ability to integrate multiple systems and services and the
possibility of interconnecting these networks with other wide area networks and
adding local networks by creating LAN.
At present, there would have no sense to design a municipal network that is
not expandable in the future to offer FTTH connections to each user. This thesis has
taken this into account and have raised a fiber network that bring the fiber to the
maximum number of users so in medium to long term the current infrastructure is
reusable for future use.


142

The deployment of next generation networks over FTTH networks is becoming
more common and the emergence of companies such as neutral carrier model are an
important solution for the evolution of them. Companies that are launched in this field
are increasingly common, with important views of the future at technological and
economic level.

Glossary

143

GLOSSARY

AES Advanced Encryption Standard
APC Angled Physical Contact
APON ATM Passive Optical Network
AON Active Optical Network
ATM Asynchronous Transfer Mode
AWG Arrayed Waveguide Grating
BER Bit Error Rate
BPON Broadband Passive Optical Network
CATV Cable TV
CCTV Closed-Circuit TV
CPM Cross-Phase Modulation
CWDM Coarse Wavelength Division Multiplexing
DSL Digital Subscriber Line
DSF Dispersion-Shifted Fiber
DWDM Dense Wavelength Division Multiplexing
EDFA Erbium-Doped Fiber Amplifier
EPON Ethernet Passive Optical Network
FBT Fused Bi-conic Tapered
FEC Forward Error Correction
FSAN Full Service Access Network
FTP File Transfer Protocol
FTTB Fiber To The Building
FTTC Fiber To The Curb
FTTH Fiber To The Home
FTTN Fiber To The Neighborhood
FTTO Fiber To The Office
FTTP Fiber To The Premises
FTTU Fiber To The User
FTTx Fiber To The x
FWM Four Wave Mixing
GEM GPON Encapsulation Mode
GFP Generic Framing Procedure
GEPON Gigabit Ethernet Passive Optical Network
GPON Gigabit Passive Optical Network
HDTV High-Definition TV
IEEE Institute of Electrical and Electronics Engineers
IPTV Internet Protocol TV
ISP Internet Service Provider
Glossary

144

ITU-T International Telecommunications Union - Telecommunication Standardization
Sector
LAN Local Area Network
LASER Light Amplification by Stimulated Emission of Radiation
LC Little Connector
LED Light-Emitting Diode
MAN Metropolitan Area Network
MDU Multi-Dwelling Unit
MTU Multi-Tenant Unit
MU Miniature Unit
NG-PON Next Generation Passive Optical Network
NRZ Non-Return to Zero
NZDSF Non-Zero Dispersion-Shifted Fiber
OAM Operations, Administration and Maintenance
ODN Optical Distribution Network
OLT Optical Line Termination
OPGW OPtical Ground Wire
ONU Optical Network Unit
ONT Optical Network Termination
PE PolyEthylene
PLC Planar Light Circuit
PLOAM Physical Layer OAM
PMD Polarization Mode Dispersion
PON Passive Optical Network
POTS Plain Old Telephone Service
PVC PolyVinyl Chloride
P2P Peer-to-Peer
SBS Simulated Brillouin Scattering
SC Standard Connector
SMF Single-Mode Fiber
SPM Self-Phase Modulation
SRS Simulated Raman Scattering
TDM Time Division Multiplex
TDMA Time Division Multiple Access
VOIP Voice-Over-Internet Protocol
V-OLT Video Optical Line Termination
WDM Wavelength Division Multiplexing
WDM-PON Wavelength Division Multiplexing Passive Optical Network
xPON x Passive Optical Network
XG-PON 10 Gigabit Passive Optical Network
10G-EPON 10 Gigabit Ethernet Passive Optical Network
References and Bibliography

145

REFERENCES AND BIBLIOGRAPHY

[1] Martn Sanz, Jos (1996). Comunicaciones pticas. Editorial Paraninfo.
[2] Bates, Regis J (2001). Optical Switching and Networking Handbook. New York:
McGraw-Hill.
[3] Martnez Herrero, R; Mejas Arias, P (1999). ptica geomtrica. Ed. Sntesis.

[4] "Shoaling, Refraction, and Diffraction of Waves". University of Delaware Center
for Applied Coastal Research. [cit. 27/12/2010]
http://www.coastal.udel.edu/ngs/waves.html.

[5] Escofet Soteras, Jaume; Milln, Mara Sagrario (2004). ptica geomtrica.
Editorial Ariel.
[6] Janer, Carlos (2009). Dispositivos para comunicaciones pticas. Dinel.

[7] Keiser, Gerd (1991). Optical Fiber Communications. New York: McGraw-Hill.
[8] Tomasi, Wayne (2003). Sistemas de comunicaciones electrnicas. Madrid:
Prentice Hall.

[9] Hecht, Jeff (2002). Understanding fiber optics. Prentice Hall.

[10] S, Nilsson-Gistoik (2002). Optical fiber theory for communication networks.
Ericsson Networks Technolgies AB.

[11] Encyclopedia of Laser Physics and Technology. Optical Fiber Communications.
[cit. 01/02/2011]
http://www.rp-photonics.com/optical_fiber_communications.html.

[12] Gumaste, Ashwin; Antony, Tony (2002). DWDM Network Designs and
Engineering Solutions. Cisco Systems.

[13] Prat, Josep (2008). Next-Generation FTTH Passive Optical Networks. Springer
Science + Business Media B.V.

[14] Keiser, Gerd (2006). FTTX concepts and applications. John Wiley & Sons, Inc.


146

[15] White Paper (2008). AON vs. PON A comparison of two optical access network
technologies and the different impact on operations. KEYMILE International
GmbH.

[16] Kramer, Glen; Mukherjee, Biswanath; Pesavento, Gerry (2000). Ethernet PON
(ePON): Design and Analysis of an Optical Access Network. University of
California.

[17] Lam, Cedric F (2007). Passive optical networks: principles and practice. Elsevier
Inc.

[18] International Telecommunication Union (2001). ATM-PON requirements and
managed entities for the network view. ITU-T Recommendation Q.834.2.

[19] International Telecommunication Union (2001). Broadband optical access
systems based on Passive Optical Networks (PON). ITU-T Recommendation
G.983.1.

[20] International Telecommunication Union (2001). ONT management and control
interface specification for B-PON. ITU-T Recommendation G.983.2.

[21] International Telecommunication Union (2001). A broadband optical access
system with increased service capability by wavelength allocation. ITU-T
Recommendation G.983.3.

[22] International Telecommunication Union (2008). Gigabit-capable passive optical
networks (GPON): General characteristics. ITU-T Recommendation G.984.1.

[23] International Telecommunication Union (2003). Gigabit-capable Passive Optical
Networks (GPON): Physical Media Dependent (PMD) layer specification. ITU-T

Networks (G-PON): Transmission convergence layer specification. ITU-T

Networks (G-PON): ONT management and control interface specification. ITU-T


147

Networks (G-PON): Enhancement band. ITU-T Recommendation G.984.5.

[27] The Product Group LLC. The essentials of WDM PON. [cit. 18/03/2011]
http://www.fttxtra.com/ftth/wdm-pon/the-essentials-of-wdm-pon/.

[28] International Telecommunication Union (2010). 10-Gigabit-capable passive
optical networks (XG-PON): General requirements. ITU-T Recommendation
G.987.1.

[29] International Telecommunication Union (2010). 10-Gigabit-capable passive
optical networks (XG-PON): Physical media dependent (PMD) layer
specification. ITU-T Recommendation G.987.2.

[30] The Product Group LLC. 10G GPON Tutorial. [cit. 20/03/2011]
http://www.fttxtra.com/ftth/10g-gpon-brief-overview/.

[31] Sameer Ashfaq Malik (2008). 10G EPON- Unleashing the Bandwidth Potential.
ZTE White Papers.

[32] FTTH Council Europe (2011). FTTH Business Guide. [cit. 30/03/2011]
http://www.ftthcouncil.eu/documents/studies/FTTH-Business-Guide-2011-
2ndE.pdf/.

[33] The Institute of Electrical and Electronics Engineers (2004). 1000Base-LX.
IEEE 802.3 Recommendation.

[34] The Institute of Electrical and Electronics Engineers (2004). 10Gbase-LX.
IEEE 802.3 Recommendation.

[35] Telecom and networking communications today. OF Cable: Fiber on a rollout.
[cit. 09/04/2011] http://www.communicationstoday.co.in/oct2007/of-cable-
fiber-on-a-rollout-3159-154.html/.

[36] ITU-T Manual (2009). Optical fibres, cables and systems. International
Telecommunication Union.

Design of Passive Optical Network

Uploaded by

Copyright:

Available Formats

Design of Passive Optical Network

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Design of Passive Optical Network

Uploaded by

Copyright:

Available Formats

What is the main topic of discussion in the document?

What technologies are discussed in relation to fiber optic networks?

BRNO UNIVERSITY OF TECHNOLOGY

FACULTY OF ELECTRICAL ENGINEERING

the value of wavelength in a vacuum:

= refractive index of the first medium,

30 2250 , so a FTTB design with a 2.488 Mbps

20 2 to provide this service to all users.

22 2,2 to provide all users.

You might also like