A NEAR REAL TIME SYSTEM TO 
DETERMINE A COST-EFFECTIVE IDD 
TRAFFIC ROUTING PLAN 
 
 
 
 
 
 
 
 
 
 
MSC IN COMPUTER SCIENCE 
 
 
 
 
 
 
 
 
 
MERENNAGE AROSHA ANTHONY 
CHANDANA FERNANDO 
 
 
 
 
 
 
 
 
 
 
DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING, 
UNIVERSITY OF MORATUWA, 
SRI LANKA. 
DECEMBER, 2007. 
I 
A NEAR REAL TIME SYSTEM TO DETERMINE A 
COST-EFFECTIVE IDD TRAFFIC ROUTING PLAN 
 
 
 
 
 
 
 
MERENNAGE AROSHA ANTHONY CHANDANA FERNANDO 
 
 
 
 
 
 
 
 
 
 
 
 
This dissertation was submitted to the Department of Computer Science and 
Engineering of the University of Moratuwa in partial fulfilment of the 
requirements for the Degree of MSc in Computer Science. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Department of Computer Science and Engineering, 
University of Moratuwa, 
Sri Lanka. 
December, 2007. 
II 
I here by declare, that the work included in this dissertation in whole has 
not been submitted for any other academic qualification at any known 
institution to me. 
 
 
 
…………………………………………… 
Merennage Arosha Anthony Chandana Fernando 
 
 
 
…………………………………………… 
Dr. Sanath Jayasena. (Supervisor) 
III 
ABSTRACT 
 
International telecommunication has become a very important sector in the present day 
communication. IDD traffic is one of the main sources of revenue for national 
telecommunication service providers (Telcos). Hence, it has to be managed properly and 
driven effectively to gain maximum profit. 
Telcos usually deal with multiple international operators (carriers) e.g. SingTel, 
Telstra, in order to use latter’s networks for international traffic. These carriers offer 
different charging structures which vary with the time, volume and the destination. 
Most Telcos incur high costs due to lack of a proper system to analyse and forecast 
an optimised IDD routing plan. 
Therefore, a system which is capable of producing a cost-effective routing plan by 
analysing the current IDD usage pattern and the traffic cost would be much beneficial for 
all Telcos. 
The objective of this research work is to find a suitable methodology for Telcos to 
produce a cost-effective routing plan for the IDD traffic in near real time. 
In our approach, the main tasks are, 
• Forecasting the IDD traffic pattern to each destination 
• Determining the optimised routing plan. 
 
Forecasting the future traffic was based on past traffic, current IDD traffic trend and the 
subscriber growth rate of Telcos. 
A Genetic Algorithm (GA) was used to obtain the optimised routing plan. Dynamic 
behaviour of the problem domain was successfully addressed in the GA structure. The GA 
parameters, population size, number of generations and the GA operations (crossover, 
mutation and reproduction) ratio were determined experimentally. 
 
Experimental results show that our solution is capable of producing a cost-effective routing 
plan in near-real-time which reduces the cost by 30% to 50% against the manual routing 
plan. 
 
Our solution could be used by national Telcos to save significant costs in the international 
traffic and to pass some of that cost savings to their customers. 
 
IV 
ACKNOWLEDGEMENT 
 
I would first like to thank my supervisor; Dr. Sanath Jayasena for his continued and valued 
advice throughout the entire research process. If not for his support, this project would not 
have been a success. I admire his qualities and the ability to listen with patience, when 
working with me. 
  
I would also like to thank my wife Shamalka, my kids Shenuka & Avindya, and my 
parents Milson & Asilda for tolerating me and sacrificing so much during the period of this 
research project and for encouraging me continuously. 
 
Additionally, I cannot forget the help rendered by Professor N. Rajkumar who is a very 
close family friend whose ideas for this research project and advice during the dissertation 
were invaluable to me. His experience on research projects and the plethora of literature 
provided by him were very useful and highly appreciated during the initial stages of the 
project. 
 
Further, I would like to thank Ms. Kusum Liyanage for providing valuable information 
pertaining to International Telecommunication Business Practices. (Interconnect settlement 
procedure and making agreements with international carriers) 
 
In conclusion I wish to acknowledge my friends, Mr. Suranga Ratnapala for helping me 
during the prototype development. Mr. Sanjeewa Ratnayaka for helping me to acquire real 
test data from telecoms, together with Mr. Kassapa Jayasinghe for his assistance in 
drawing figures and illustrations for the dissertation. 
V 
TABLE OF CONTENTS 
 
                Page 
List of Figures                 VII 
List of Tables                  VIII 
List of Abbreviations                 X 
 
CHAPTER 1 – INTRODUCTION       1 
1.1. Background         1 
1.1.1. International Settlements      1 
1.1.2. Planning IDD Routes       2 
1.1.3. Need for a Proper Methodology     2 
1.2. Description of the Research       3 
1.2.1. Traffic Pattern Forecast      3 
1.2.2. Cost-effective Routing Plan      4 
 
CHAPTER 2 - LITERATURE REVIEW      5 
2.1. Interconnect Business in Telecommunications    5 
2.1.1 What is Interconnection?      5 
2.1.2 Why is Interconnection important?     5 
2.1.3 What are the widely accepted Interconnection related principles? 6 
2.1.4 What are the key interconnection issues faced by new entrants? 8 
2.1.5 Other Issues related to Interconnection    11 
2.2 Telco’s Current Practice and Methods on IDD Route Planning  12 
2.3 Dynamic Graphs        13 
2.4 Optimisation Tools        15 
2.5 Genetic Programming       16 
2.6 Genetic Algorithms        17 
 
CHAPTER 3 - METHODOLOGY OF STUDY     23 
3.1 Traffic Pattern Forecast       23 
3.1.1 Determine the Total Traffic (forecasted) for the Entire Month  23 
3.1.2 Adjust the Traffic for the Rest of the Month    24 
VI 
3.1.3 Allocate Traffic to the Operators with Commitment   25 
3.2 Cost-effective Routing Plan       25 
3.3 Genetic Algorithm (GA)       27 
3.4 The Proposed Solution       28 
 
CHAPTER 4 - RESULTS AND ANALYSIS     31 
4.1 Experimental Data Selection      31 
4.2 Genetic Algorithm Parameter Selection     33 
4.3 Observations and Results       33 
4.3.1 Results for Input Data Set – 1      34 
4.3.2 Results for Input Data Set – 2      38 
4.3.3 Results for Input Data Set – 3      42 
4.3.4 Results for Large Population Sizes     45 
4.4 Findings         46 
4.5 Summary of Results        46 
4.5.1 Summary of Results for Optimisation Point – 1   47 
4.5.2 Summary of Results for Optimisation Point – 2   48 
4.5.3 Summary of Results for Optimisation Point – 3   49 
4.5.4 Summary of Results for Optimisation Point – 4   50 
4.6 GA Structure – Analysis of the Parameter Selection   51 
4.7 Solution Comparison       52 
 
CHAPTER 5 - CONCLUSIONS AND RECOMMENDATIONS   56 
5.1 Conclusions         56 
5.1.1 Future Traffic Prediction      56 
5.1.2 Cost-effective Routing Plan      56  
5.1.3 Results in Near Real Time      57 
5.2 Recommendations        57 
5.3 Future work         59 
 
LIST OF REFERENCES        60
VII 
LIST OF FIGURES 
                Page 
 
Figure 2.1   Interconnection as a Business     11 
Figure 2.2   Generic Genetic Algorithm     18 
Figure 2.3  Mutation Operation      21 
Figure 2.4  Crossover Operation      21 
Figure 2.5  Reproduction Operation     22 
Figure 3.1  Chromosome       27 
Figure 4.1   Cost Reduction with different GA parameters – Check (1) 47 
Figure 4.2   Cost Reduction with different GA parameters – Check (2) 48 
Figure 4.3   Cost Reduction with different GA parameters – Check (3) 49 
Figure 4.4   Cost Reduction with different GA parameters – Check (4) 50 
 
 
 
 
 
 
 
 
 
 
 
VIII 
LIST OF TABLES 
                Page 
 
Table 4.1   Experimental Results Set (1-1)   34 
(Input Data Set 1 - Optimizing point is at the beginning – With the full traffic for the entire 
traffic period) 
 
Table 4.2   Experimental Results Set (1-2)   35 
(Input Data Set 1 - Optimizing point is within the first half – approximately 25% of the 
traffic done in the traffic period) 
 
Table 4.3   Experimental Results Set (1-3)   36 
(Input Data Set 1 - Optimizing point is in the middle of the traffic period – approximately 
50% of the traffic done in the traffic period) 
 
Table 4.4   Experimental Results Set (1-4)   37 
(Input Data Set 1 - Optimizing point is within the latter half – approximately 75% of the 
traffic done in the traffic period) 
 
Table 4.5   Experimental Results Set (2-1)   38 
(Input Data Set 2 - Optimizing point is at the beginning – With the full traffic for the entire 
traffic period) 
 
Table 4.6   Experimental Results Set (2-2)   39 
(Input Data Set 2 - Optimizing point is within the first half – approximately 25% of the 
traffic done in the traffic period) 
 
Table 4.7   Experimental Results Set (2-3)   40 
(Input Data Set 2 - Optimizing point is in the middle of the traffic period – approximately 
50% of the traffic done in the traffic period) 
 
Table 4.8   Experimental Results Set (2-4)   41 
(Input Data Set 2 - Optimizing point is within the latter half – approximately 75% of the 
traffic done in the traffic period) 
IX 
Table 4.9   Experimental Results Set (3-1)   42 
(Input Data Set 3 - Optimizing point is at the beginning – With the full traffic for the entire 
traffic period) 
 
Table 4.10   Experimental Results Set (3-2)   43 
(Input Data Set 3 - Optimizing point is within the first half – approximately 25% of the 
traffic done in the traffic period) 
 
Table 4.11   Experimental Results Set (3-3)   44 
(Input Data Set 3 - Optimizing point is in the middle of the traffic period – approximately 
50% of the traffic done in the traffic period) 
 
Table 4.12   Experimental Results Set (3-4)   45 
(Input Data Set 3 - Optimizing point is within the latter half – approximately 75% of the 
traffic done in the traffic period) 
 
Table 4.13   Experimental Results with Large Population Sizes  45 
(Results obtained with large population sizes) 
 
X 
LIST OF ABBREVIATIONS  
 
 
BTCOM - British Telecom (International Carrier) 
BVSNL - Baharat Sanchar Nigam Limited (International Carrier) 
CCBS  - Customer Care and Billing System 
CDR  - Call Detail Record 
EXDB  - Oracle 10g Express Edition 
GA  - Genetic Algorithms 
GHz  - Giga Hertz 
IDD   - International Direct Dialling 
MB  - Mega Bytes 
MSC  - Mobile Switching Centre 
OSS  - Operational Support System 
PLSQL - Procedural Language/Structured Query Language (Oracle) 
POI  - Point of Interconnect 
PSTN  - Public Switch Telephone Network 
QOS  - Quality of Service 
RIO  - Reference Interconnection Offer 
TRC  - Telecom Regulatory Commission