U N I V E R S I T Y O F M O R A T U W A 
SRI L A N K A 
DESIGN MODELING AND SIMULATION OF A 
REPEATERLESS OPTICAL FIBER NETWORK FOR 
SRI LANKA 
LIBRART 
PITw5ii7 pi MORATUWA. SRi likwk 
MO«ATU\rVA ^ ™ 
Submit ted in partial fulfi l lment tor the degree of Masters of 
Eng ineer ing in Electronics and T e l e c o m m u n i c a t i o n 
K. P. K a n d a n e a r a c h c h i 
U n i v e r s i t y o f M o r a t u v t a 
80502 
February 2004 
L / M Thesis co(I 
S 0 5 0 2 
Q05O2 
The work presented in this dissertation has not been submitted for the 
fulfillment of any other degree 
(Candidate) (Supervisor) 
DEDICATION 
I dedicate this Dissertation with a lot of respect to my lovely late Mother who directed me 
to achieve the best possible education through a lot of dedication and hard work. 
It is also with reverence and respect that I remember my Father, my school - Thurstan 
College and University of Moratuwa for the guidance given me at all times to achieve my 
goals and aims and providing me with the postgraduate course that I receive. 
CONTENTS 
ACKNOWLEDGEMENTS 
ABSTRACT 
LIST OF FIGURES 
LIST OF TABLES 
ABBREVIATIONS 
1 INTRODUCTION 
1.1 High Speed Network 
1.2 Common Network 
1.3 Optimum Network 
1.4 Objectives 
1.5 Methodology 
1.5.1 Telephone and IP Traffic Forecast by Year 2015 
1.5.2 Network Design 
1.5.3 Network Simulation 
2 LITERATURE REVIEW 
2.1 Traffic Theory for Planning 
2.1.1 Traffic Volume 
2.1.2 Traffic Density 
2.1.3 Calling Rate 
2.1.4 Probability of Loss 
2.1.5 Earlang's B Formula 
2.2 Demand Forecasting 
2.2.1 Telephone Demand Forecasting 
2.2.1.1 Macro-level telephone Demand Forecasting 
2.2.1.2 Extrapolation Method 
2.2.1.3 Method of Comparison with other Countries 
2.3 Traffic Forecasting 
2.3.1 Introduction 
2.3.2 The Gravity Model 
2.4 Economic Indicators 
2.4.1 Gross National Product 
2.4.2 Gross Domestic Product 
2.4.3 Factors of Production 
2.5 WDM Technology 
2.5.1 Optical Transmitters 
2.5.2 Optical Receivers 
2.5.3 Optical Multiplexers and Demultiplexers 
2.5.4 Optical Add Drop Multiplexers 
2.5.5 Amplifiers 
2.6 Optical Amplifiers 
2.6.1 Amplifier Wavelength Bands 
2.6.2 Erbium Doped Fiber Amplifier 18 
2.6.3 Raman Amplifier 19 
2.6.4 Comparison of Raman and Erbium Doped Amplifiers 26 
3. TELEPHONE DEMAND FORECAST 29 
3.1 Introduction 29 
3.2 Income Elastic Model 29 
4. TELEPHONE TRAFFIC FORECAST 33 
4.1 Nodes of the Network 33 
4.2 Traffic Originated from each Node 35 
4.3 Gravity Model 37 
5. INTERNET TRAFFIUC FORECAST 38 
5.1 Traffic Forecast of Internet Dial-up Users 3 8 
5.2 Broadband Users 40 
6. VOICE OVER TRAFFIC FORECAST 43 
6.1 Introduction 43 
6.2 VoIP in Sri Lanka 43 
6.3 Telephone Traffic migration from PSTN to VoIP 43 
6.3.1 Traffic Migration of Business Customers from PSTN to VoIP 44 
6.3.2 International Traffic Migration from PSTN to VoIP 44 
6.3.3 Domestic Traffic Migration from PSTN to VoIP 45 
6.3.4 Traffic Migration Patterns from PSTN to VoIP 45 
7. NETWORK DESIGN 48 
7.1 Network Topology 48 
7.2 Traffic Routing 50 
7.3 Capacity of the Network 50 
7.4 Wavelength allocation and connectivity 50 
7.5 Selection of Wavelengths 53 
7.6 Design Configuration 53 
7.7 Selection of the Fiber 54 
7.8 Selection of Sources and Detectors 55 
7.9 Network Design using Optical Amplifiers 55 
7.10 BER Objective and Design Steps 55 
7.11 Network Design 55 
7.11.1 Selecting a suitable Booster and Pre-Amplifier 55 
7.11.2 Power Budget Calculations 56 
7.11.3 OSNR Calculations of Segments 61 
7.11.4 OSNR Calculations of Optical Line Sections 64 
7.11.5 Estimating Q Factors and preparing of the Performance Budget 64 
7.12 Dispersion Management 67 
8 NETWORK SIMULATION 
8.1 Introduction 
8.2 Simulation Strategy 
8.3 Optimum Results of the Network 
8.4 Simulated Outputs of the Network 
9. SUMMARY AND CONCLUSIONS 
9.1 Telephone Demand and its distribution by year 2015 
9.2 IP Traffic Demand and its distribution by year 2015 
9.3 Network Topology and capacity requirements 
9.4 Designing of a Repeaterless Optical Network 
9.5 Modeling of the Network to ensure desired results 
9.6 Suggestions for future work 
REFERENCES 
APPENDICES 
ACKNOWLEDGEMENTS 
I would like to express my sincere gratitude to my Project Supervisor Prof. (Mrs.) I. J. 
Dayawansa (BSc Dip E.E. MSc PhD FIEE) for her guidance, valuable advices and 
encouragements for successfully completing this Postgraduate Research study. 
Also I thank to University of Moratuwa for giving me an opportunity for a postgraduate 
study where I had the opportunity to explore in new technology areas like Repeaterless 
Optical Networks. 
I should express my gratitude to ARTIS Software Corporation for providing me an 
evaluation copy of OptSim Software Tool for simulating and evaluating the designed 
Network. 
Finally I thank to my lovely wife Mrs. J. N. Wickramasinghe and Sri Lanka Telecom for 
providing me necessary information and support for completing this project successfully. 
ABSTRACT 
A reliable and wideband telecommunication network is a vital infrastructure development, 
where wide band services such as ATM, ADSL and IP based services could be supported. In 
Sri Lanka, the requirement of this kind of an optical network is very significant as other 
operators also can share the capacity of the network for transporting their traffic. On the other 
hand the network problems such as excessive BER (Bit Error Rate) etc are experienced after its 
construction. In order to avoid such limitations in the network, the network needs to be 
modeled on appropriate software tools and run with designed network parameters, so that the 
desired BER could be ensured. 
During the initial phase of the study, the total telephone demand by year 2015 was estimated as 
nearly 2 million subscribers. This was estimated through the world trend for telephone 
subscribers together with economic indicators such as GNP and GDP. 
The Nodes of the Network was determined based on the present distribution of customers in 
the County. In this case all the Tertiary Switching Center areas and the Secondary Center 
Areas where the customer base is more than 2.5% of total customers were taken as the main 
nodes of the network. In addition Jaffna and Baticaloa were also taken as nodes considering the 
potential growth of traffic in northern and eastern parts of the Island. 
The Gravity model and Earlang's B formula, traffic tables, were used to find the traffic 
between nodes and the number of circuits between nodes. Based on the traffic distribution 
between nodes, a part of the network was proposed as a fully reliable Ring Network, while 
other nodes are connected through extended links. The IP traffic, which is thought to be the 
major traffic flow in the future, were estimated considering the broadband Internet growth in 
the country. Also the traffic, which are expected to be migrated from traditional PSTN to IP 
Network were identified and estimated to find the total bandwidth requirement of the network 
by year 2015. 
The number of wavelengths in the proposed Network were decided based on the final 
bandwidth requirement. This resulted an island wide network consisting of WDM Ring 
Network having 08 wavelengths that basically covers the southern part of the country and two 
other extensions having a wavelength each to northern and eastern parts of the country. The 
Colombo and the Kandy nodes were selected as Full Fiber Terminal Stations as most of the 
traffic flow between these two nodes. Wavelengths are added and dropped at each branch 
station based on the traffic volumes between these nodes. 
The wavelengths were selected such that the space between adjacent wavelengths is 0.8nm to 
avoid nonlinear effects and cross talks. The G.655 non-zero dispersion fiber was selected to 
mange the dispersion and non-linear effects. DFB and APD are the Source and the Detector 
respectively to suit 1 ong haul t ransmissions h aving n arrow spectral widths and also to meet 
better sensitivity at the receiver. 
The proposed Network is a Repeaterless Optical Network, where the Power Budget of the 
longest Segment, Kandy - Matara, of 280km was designed without employing a physical 
repeater, which needs power feeding. This was achieved using Raman Amplifiers as line 
repeaters and Erbium Doped Fiber Amplifiers (EDFA) as Boosters and Pre-Amplifiers. The 
Power Budget has been prepared for all other Segments as well based on appropriate 
configurations. Also the BER objective of 10"9 was ensured for the longest Optical Line 
Section of Colombo - Kandy via Matara, in which a couple of express wavelengths are 
assigned for carrying traffic between Colombo and Kandy. The Performance Budget was 
prepared for long Optical Line Sections and the calculated BER was found as better than 10"9. 
This has been further confirmed by the Eye Diagrams after simulating the Network on the 
OptSim Network Simulator developed by ARTIS Software. 
-iii-
• 
LIST OF FIGURES 
Figure 2.1 Offered Traffic and Carried Traffic 6 
Figure 2.2 Factors affecting Demand 8 
Figure 2.3 Economic Index Relating to Telephone Density 10 
Figure 2.4 Traffic flow between Two Office 12 
Figure 2.5 Typical WDM Multiplexer 15 
Figure 2.6 WDM Demultiplexer using Wave Guide Grating Diffraction technique 16 
Figure 2.7 Illustration of the advantage of using Optical Amplifiers 17 
Figure 2.8 Typical Erbium Doped Fiber Amplifier 18 
Figure 2.9 Energy Level Diagram of an Erbium Doper Fiber Amplifier 19 
Figure 2.10 Energy Level Diagram of a Raman Amplifier 19 
Figure 2.11 Gain Vs Frequency difference between the Signal and Pump 20 
Figure 2.12 Configuration of the Forward Pumping Raman Amplifier 20 
Figure 2.13 Configuration of the Backward Pumping Raman Amplifier 21 
Figure 2.14 Hybrid Raman Doped Amplifier 28 
Figure 3.1 World Telephone Demand Trend 30 
Figure 5.1 Growth of Internet Users 39 
Figure 5.2 Growth of Broadband users 41 
Figure 5.3 Growth of Broadband users (log scale) 41 
Figure 7.1 Topology of the Proposed Optical Fiber Network 49 
Figure 7.2 Wavelength allocation and connectivity 52 
Figure 7.3 Schematic Diagram of a WDM Segment 54 
Figure 7.4 Simplest Configuration (Configuration-A) Fiber Link 57 
Figure 7.5 Configuration-B of a Fiber Link 58 
Figure 7.6 Configuration -C of a Fiber Link 58 
Figure 7.7 Configuration-D of a Fiber Link 59 
Figure 7.8 Measured EDF Gain and Noise Figure Vs. Pumping Power 59 
Figure 7.9 Configuration of the proposed Network 70 
Figure 8.1 Output of the Colombo main Fiber via Kurunegala 72 
Figure 8.2 Input of the Kandy main Fiber via Kurunegala 73 
Figure 8.3 Eye diagram of X\ at Kandy FFTS via Kurunegala 74 
Figure 8.4 Eye diagram of A.2 at Kandy FFTS via Kurunegala 74 
Figure 8.5 Eye diagram of A. 3 at Kandy FFTS via Kurunegala 75 
Figure 8.6 Eye diagram of X4 at Kandy FFTS via Kurunegala 75 
Figure 8.7 Eye diagram of X.5 at Kandy FFTS via Kurunegala 76 
Figure 8.8 Eye diagram of X<, at Kandy FFTS via Kurunegala 76 
Figure 8.9 Eye diagram of A. 7 at Kandy FFTS via Kurunegala 77 
Figure 8.10 Eye diagram of A.g at Kandy FFTS via Kurunegala 77 
Figure 8.11 Output of the Colombo main Fiber via Matara 78 
Figure 8.12 Input of the Kandy main Fiber via Matara 79 
Figure 8.13 Eye diagram of X] at Kandy FFTS via Matara 80 
Figure 8.14 Eye diagram of X2 at Kandy FFTS via Matara 80 
Figure 8.15 Eye diagram of X} at Kandy FFTS via Matara 81 
-iv-
Figure 8.16 Eye diagram of A.4 at Kandy FFTS via Matara 81 
Figure 8.17 Eye diagram of X5 at Kandy FFTS via Matara 82 
Figure 8.18 Eye diagram of at Kandy FFTS via Matara 82 
Figure 8.19 Eye diagram of X.7 at Kandy FFTS via Matara 83 
Figure 8.20 Eye diagram of X% at Kandy FFTS via Matara 83 
Figure 8.21 Output of the Kurunegala Fiber 84 
Figure 8.22 Input of the Anuradhapura Fiber 85 
Figure 8.23 Eye Diagram at Anuradhapura Station 85 
Figure 8.24 Output of the Anuradhapura Fiber 86 
Figure 8.25 Input of the Jaffna Fiber 86 
Figure 8.26 Eye Diagram at Jaffna Station 87 
Figure 8.27 Output of the Kandy Fiber 88 
Figure 8.28 Input of the Baticaloa Fiber 88 
Figure 8.29 Eye Diagram at Baticaloa Station 89 
LIST OF TABLES 
Table 2.1 Comparison of Raman and Erbium Doped Fiber Amplifiers 27 
Table 3.1 Calculation of Total Telephone Demand at the end of each Year 31 
Table 4.1 Distribution of Telephone Customers in year 2001 34 
Table 4.2 Nodes of the network and its traffic distribution 36 
Table 5.1 Growth of Internet users and its forecast 38 
Table 5.2 Distribution of Internet users by year 2015 and Bandwidth requirement 39 
Table 5.3 Growth of Broadband users 40 
Table 5.4 Forecast of Broadband users 42 
4. Table 5.5 Distribution of Broadband users by year 2015 and Bandwidth requirement 42 
Table 6.1 Traffic migration patterns from traditional PSTN to VoIP 45 
Table 6.2 International and a portion of Domestic traffic as VoIP 46 
Table 7.1 Capacity Requirement of each Segment 50 
Table 7.2 Wavelength Requirement of the Network 51 
Table 7.3 Wavelengths in the Network 53 
Table 7.4 Parameters of G655 Non-Zero Dispersion Shifted Fiber 54 
Table 7.5 Typical Parameters of a Booster, Pre-Amplifier and Raman Amplifier 56 
Table 7.6 Distances between Nodes of the Network 57 
Table 7.7 Network Segments and its particular Configuration 60 
Table 7.8 Power Budgets of Segments 61 
Table 7.9 Parameters required for SNR Calculations 62 
Table 7.10 SNR Calculations of Segments 63 
Table 7.11 Performance Budget 67 
Table 7.12 Dispersion Coefficients of each wavelength 68 
Table 7.13 Lengths of DCF to compensate dispersion of each Segment and OLS 69 
V 
-vi-
v 
ABBREVIATIONS 
ADP Avalanche Photo Diode 
ADSL Asymmetric Digital Subscriber Line 
ASE Amplified Spontaneous Emission 
BER Bit Error Rate 
BS Branch Station 
CR Calling Rate 
DCF Dispersion Cut-off Fiber 
DFB Distribution Feed Back 
DSF Dispersion Shifted Fiber 
DSL Digital Subscriber Line 
EDFA Erbium Doped Fiber Amplifier 
EOL End of Life 
ES Extension Station 
FFTS Full Fiber Terminal Station 
FWM Four Wave Mixing 
GDP Gross Domestic Product 
GNP Gross National Product 
IP Internet Protocol 
ITU International Telecommunication Union 
NF Noise Figure 
NI Net Income 
NZ-DSF Non Zero Dispersion Shifted Fiber 
OADM Optical Add Drop Multiplexer 
OLS Optical Line Section 
OSNR Optical Signal to Noise Ratio 
PC Personal Computer 
PCM Pulse Code Modulation 
PMD Polarization Mode Dispersion 
PSTN Public Switched Telephone Network 
RA Raman Amplifier 
SLT Sri Lanka Telecom 
SNR Signal to Noise Ratio 
SPM Self Phase Modulation 
SRS Stimulated Raman Scattering 
SSC Secondary Switching Center 
STM Synchronous Transport Mode 
-vii-
TDM 
TRC 
USB 
VOIP 
WDM 
-viii-
Time Division Multiplexing 
Telecommunication Regulatory Commission 
Universal Serial Bus 
Voice Over Internet Protocol 
Wavelength Division Multiplexing