This research presents the development of a modified token-based congestion control scheme with adaptive forwarding mechanism (mTBCC) algorithm for addressing congestion problems in opportunistic networks (OppNets). The algorithm addresses the limitations associated with the standard token –based congestion control (TBCC) in terms of its ability to redirect the traffic from more congested nodes of OppNet to congestion-free nodes without necessarily compromising computational time. This is because the TBCC has the tendency to drop significant number of messages when node is full (overflow) in order to control congestion. OppNet was modeled using ONE simulator with Eclipse, which is a java based programming language. The node density was controlled by varying the greatest connected component (GCC) expressed in percentage and the corresponding results were used to evaluate the performance of the proposed approach using (dropped messages and network transit time) as performance metrics. The results showed reduction in dropped messages and network transit time across all scenarios considered. At queue size of 10(QS-10), TBCC had 38592 messages, and mTBCC has 36037 messages, yielding an average improvement of 13.91%, at queue size of 20 (QS-20), TBCC had 30330 messages and mTBCC had 27845 messages resulting in average improvement of 10.78%, at queue size of 30(QS-30) TBCC produced 28356 messages and mTBCC yielded 26767 messages resulting to an average improvement of 5.68% and at queue size of 40(QS-40) , TBCC had 23150 messages while mTBCC had 22197 messages, providing an average improvement of 4.22% respectively for dropped messages. In addition, at 0.5GCC, TBCC had 29401.70 time and mTBCC had 27151.41 time, producing an average improvement of 8.34%, at 0.6GCC, TBCC produced 16319.29 time and mTBCC had 15966.42 time, resulting in average improvement of 2.19%, at 0.7GCC, TBCC yielded 13178.21 time and mTBCC produced 12581.01 time, resulting in an average improvement of 4.61%, and at 0.8GCC, TBCC had 12333.55 time and mTBCC had 11453.23 time, yielding an average improvement of 7.63% for network transit time.
TABLE OF CONTENTS
TITLE PAGE i
TABLE OF CONTENTS viii
LIST OF TABLES xii
LIST OF FIGURES xiv
LIST OF ABBREVIATIONS xvi
CHAPTER ONE: INTRODUCTION
1.1 Background 1
1.2 Problem statement 3
1.3 Motivation 3
1.4 Aim and objectives 4
1.5 Significant of Research 4
1.6 Scope of the Research 5
1.7 Dissertation outline 5
CHAPTER TWO: LITERATURE REVIEW
2.1 Introduction 6
2.2 Review of fundamental concepts 6
2.2.1 Delay tolerant networks 6
126.96.36.199 Delay tolerant network architecture 7
188.8.131.52 Bundle protocol 8
2.2.2 Delay tolerant networks routing protocols 9
184.108.40.206 Flooding strategies 10
220.127.116.11 Forwarding strategies 11
2.2.3 Delay tolerant networks congestion control schemes 13
18.104.22.168 Congestion detection 14
22.214.171.124 Congestion control 14
2.2.4 Opportunistic networks 15
2.2.5 Token based congestion control 18
2.2.6 Adaptive forwarding strategy 19
2.2.7 Opportunistic network environment (ONE) simulator 20
2.2.8 Research performance metrics 21
2.3 Review of similar works 22
CHAPTER THREE: METHODS
3.1 Introduction 32
3.2 Methodology 32
3.3 Replication of the token-based congestion control 33
3.3.1 Initializing token based parameter 33
3.3.2 Token based congestion control scenario settings 34
3.4 Development of a modified token-based congestion control 35
3.4.1 Modified token based congestion control scenario settings 37
3.5 Installation and configuration 38
3.6 Opportunistic network modelling 38
3.7 Simulation model 39
3.8 PRoPHET routing protocol 39
3.9 Visualization 39
3.10 Validation of the congestion control strategies 40
3.11 Performance evaluation 41
3.11.1 Percentage improvement 42
3.11.1 Rrelevant equations 42
CHAPTER FOUR: RESULTS AND DISCUSSION
4.1 Introduction 43
4.2 Results of the token-based congestion control strategy 43
4.3 Results of the modified token-based congestion strategy 49
4.4 Comparison of the results 55
4.4.1 Comparison of mTBCC and TBCC performance for dropped message 55
126.96.36.199 Dropped message percentage improvement 58
4.4.2 Comparison of mTBCC and TBCC performance for NTT 60
188.8.131.52 NTT percentage improvement 63
CHAPTER FIVE: CONCLUSION AND RECOMMENDATIONS
5.1 Summary 65
5.2 Conclusion 65
5.3 Limitation 65
5.4 Significant contribution 66
5.5 Recommendations for further work 66
In the last decade, the deployment of wireless-enabled devices have achieved tremendous growth as the number of mobile wireless device increases giving rise to spontaneous communication between devices. However in many cases the spontaneous network formed between these devices are characterized by partitioning and long periods of disconnectivity (Piórkowski et al., 2008; Zhang et al., 2007). To address these spontaneous network constraints such as intermittent connectivity, high latency and support, delay tolerant applications have emerged using store-carry-and- forward approach to forward messages in Delay Tolerant Networks (Thompson et al., 2010).
Delay Tolerant Networks (DTNs) are decentralized networks designed to function effectively in extreme environment such as intermittent connectivity, long delays and typically are wireless mobile networks (Coe & Raghavendra, 2010). The DTN applications represent a broad class of networks such as interplanetary internets (Burleigh et al., 2003) , underwater sensor networks (Partan et al., 2007), vehicular ad-hoc networks (Lin et al., 2008; Lu et al., 2008) and opportunistic networks (OppNets) (Zhang et al., 2014), etc. DTN has several attributes contrary to the traditional network. In traditional networks, some assumptions are made such as (Yang et al., 2012):
1. The packet loss ratio is minimal (low error rate)
2. The link between two connecting nodes is always stable
3. The data transfer time between two nodes should not be more than two times round triple time.
The goal of DTN architecture and protocols is to provide delivery of data messages between nodes in extreme environments where traditional network protocols, such as transmission control protocol and internet protocol (TCP/IP) networks and MANET protocols fail. Such environments violate many underlying assumptions of TCP/IP networks, which comprises the existence of end-to-end path between nodes, short delays on communication links and stable connectivity (Coe & Raghavendra, 2010).
The DTN architecture described in (Fall, 2003), uses the so called store-carry-and-forward approach, as contrary to the concept of Internet’s store-and-forward, to deliver messages before forwarding them, however the time elapses while waiting to be forwarded are much smaller when compared to DTNs. The DTN architecture extends the reach of network and facilitates communication among nodes in a challenged environment. It possesses the following characteristics (Yang et al., 2012):
1. High latency
2. Low data speed
3. Intermittent connectivity
4. High error rate
5. Long Queueing time
6. Bandwidth is asymmetry in nature.
Opportunistic networks (OppNets) are characterized by sparse multi-hop variant of ad-hoc networks where nodes utilize any available contact opportunities to exchange and forward messages. It is one types of challenged networks developed from delay tolerant networks, which
is the focus of this research work. The OppNets implement store-carry-and-forward operation to transmit messages, which is distinct from the store and forward mechanism in MANET. The current intermediate node has the capability to store messages during blackout and forward it when connectivity resumes as nodes move (Zhang et al., 2014).
1.2 Problem Statement
Detecting and handling congestion in OppNets is vital and challenging task because of its representative properties. The existing OppNet forwarding algorithms mostly send packets towards specific nodes to maximize delivery ratio and minimize latency. However, as traffic continue to build up in the network, these nodes become congested and drop any incoming traffic. This research proposes development of a modified token-based congestion control scheme with adaptive forwarding to monitor the amount of traffic entering the network to the network capacity and reroute packet from congested node to congestion- free node of OppNet, taken cognizance of least migration cost and the largest available free buffer size from the reference node and drop the message if buffer of all nodes is full.
Traditional Opportunistic Network congestion control mechanisms are mostly directed towards particular node on certain criteria, which eventually overload the node and consequently become congested with time under different applications. They are not adaptable to different opportunistic network scenarios, mainly because they were designed to a specific scenario with special features. Adaptive forwarding techniques often provide results with fairness and efficient resource use. Thus, this approach is especially relevant to opportunistic networks, where connected paths does not always exist between the source node and the destination node. To
sustain the network operating at an acceptable efficiency, nodes in opportunistic networks would need to evaluate the likelihood of congestion, by dissecting the local data. In addition, adaptive forwarding mechanisms provide the power to redirect the traffic from more congested node to congestion free node in the network. In other words, since the opportunistic network environment is dynamic, the opportunistic network node has to be able to adjust in order to cope adequately with the environment. Congestion avoidance is a basic and fundamental task of autonomous opportunistic network nodes. It is especially vital in challenging environments where an end-to-end communication is not always possible to reach a destination.
1.4 Aim and Objectives
The aim of this research is to develop a modified token-based congestion control scheme mTBCC approach to reduce the effects of congestion in opportunistic networks.
The objectives of the research are enumerated as follow:
1. To replicate the TBCC and develop the mTBCC with adaptive forwarding mechanism via opportunistic network environment (ONE) simulator
2. To develop the simulation setup of the mTBCC with adaptive forwarding using ONE simulator
3. To compare the performance of the TBCC-based model and that of the mTBCC-based model on the benchmark of Helsinki area test-bed using dropped message and network transit time as performance metrics.
1.5 Significance of Research
Communication in a challenged network environment occurs intermittently and characterized by high variable latency, making TCP/IP not suitable to be applied. In this situation, OppNets concept provides alternative measure necessary for data transfer. The major significant difference between Internet communication paradigm and OppNets communication is that end-
to-end communication path does not always exist leading to disconnection, high latency, and high error rate in communication. To address this challenge, OppNet uses store-carry-and-forward approach to forward packet from source to destination. OppNet has different routing protocols based on historical knowledge or replication schemes for message successful delivery.
1.6 Scope of the Research
The research presents the development of a modified token-based congestion control scheme with adaptive forwarding mechanism for OppNets. This approach is based on using token based techniques where nodes holding valid token are allowed to inject message into the network and at congestion point, reroute the message to the nearest neighboring node taking cognizance of least migration cost and largest available free buffer space. The algorithm is simulated using ONE simulator and validated using Helsinki area test bed.
1.7 dissertation outline
Chapter One presents the general introduction of the dissertation while Chapter Two presents the comprehensive review of related literature and relevant fundamental concepts about token based, delay tolerant network, delay tolerant network routing, opportunistic network, congestion control, adaptive forwarding strategy and opportunistic network environment simulator and Helsinki simulation area. Chapter Three presents an elaborate procedure for the actualization of modified token- based congestion control scheme with adaptive forwarding algorithm by using the cost function: largest available free buffer size and the least migration cost. The analysis, performance and discussion of the result are presented in chapter four. Chapter Five covers the conclusion and recommendation for areas of further research. Finally, relevant references and Appendixes are presented at the end of this dissertation.
GET THE COMPLETE PROJECT»
Do you need help? Talk to us right now: (+234) 08060082010, 08107932631, 08157509410 (Call/WhatsApp). Email: firstname.lastname@example.org