Get help now


Updated January 17, 2019

Download Paper

File format: .pdf, .doc, available for editing


Get help to write your own 100% unique essay

Get custom paper

78 writers are online and ready to chat

This essay has been submitted to us by a student. This is not an example of the work written by our writers.

ENHANCING MOBILITY CAPACITY AND NETWORK PERFORMANCE OF CLIENT-SERVER ARCHITECTURES USING MIPV6 HOST-BASED NETWORK PROTOCOL BY RUPHIN K. BYAMUNGU UNITED STATES INTERNATIONAL UNIVERSITY SUMMER 2018 ENHANCING MOBILITY CAPACITY AND NETWORK PERFORMANCE OF CLIENT-SERVER ARCHITECTURES USING MIPV6 HOST-BASED NETWORK PROTOCOL BY RUPHIN K. BYAMUNGU A Project Report Submitted to the School of Science and Technology in Partial Fulfillment of the Requirement for the Degree of Master of Science in Information Systems and Technology UNITED STATES INTERNATIONAL UNIVERSITY SUMMER 2018 STUDENT’S DECLARATION I, the undersigned, declare that this is my original work and has not been submitted to any other college, institution or university other than the United States International University in Nairobi for academic credit. Signed: __________________________ Date:_______________________ Ruphin Kusinza Byamungu (ID No 649238) This project has been presented for examination with my approval as the appointed supervisor.

Signed: __________________________ Date:_______________________ Dr. Gerald Chege Signed: __________________________ Date:_______________________ Dean, School of Science and Technology (Do not include name) COPYRIGHT No part of this project may be duplicated or transmitted in any form or by any means of electronic compact discs, magnetic any mechanical tapes, recording, photocopying on any information storage and retrieval system without prior written authorization from the author or the United States International University in Nairobi (USIU). © Copyright by Ruphin Kusinza Byamungu, 2018 ? ABSTRACT A huge number of studies have been done supporting seamless mobility networks and mobile technologies over the years. The recent innovations in mobile technologies have unveiled another revolution from the static architectural approach for client-server network relationships to more dynamic, and even mobile approaches.

Due to the special equipment and infrastructure needed to support network mobility management, it is difficult to deploy such networks beyond the local network coverage without interruption of communications. The purpose of the study was to enhance capacity and network performance of client-server architectures using Mobile IPv6 host-based mobility protocol. The research included four specific objectives that were (i) to evaluate MIPv6 technology and client-server mobility problems and proposed a solution framework, (ii) to design and implement client-server architecture using an optimized but also secure MIPv6 solution in a simulated environment, (iii) to evaluate network Quality of Service of the implemented MIPv6 solution for FTP, HTTP and Video Stream services, and (iv) to implement and evaluate client-server Fast Handover MIPv6 solution for better Quality of Service. The methodology used by this research involved Network Modeling and Simulation of the proposed network solution using discrete event simulation approach with Objective Modular Network Testbed in C++ (OMNET++) simulation tool that used INET Framework in extension to implement the solution. To collect data, the research used statistical comparative approach between MIPv6 and Fast Handover MIPv6 technologies where the results data were recorded based on dynamic rate selection of different bitrate values based on FTP, HTTP and Video Stream service models.

This method was used to separately collect and analyze data on network performance of both MIPv6 and Fast Handover MIPv6 using network performance metrics such as Throughput, Packet Error Rate, Packet Loss Rate, Handover and Packet End-to-End delays. The procedure involve multiple processes from preparation, installation and configuration of the system to results extraction and analysis. To analyze the collected data, Statistical quantitative data analysis was used considering the first order statistics such as mean, or average values. This analysis was used on the recorded results in datasets using different data analysis tools such as Wireshark, Ms Excel.

The network design adopted considered the Mobile Node as client and the CN as network server simulated in OMNET++ simulation environment. To make MIPv6 technology more effective and optimized for client-server architectures IP Security and Route Optimization processes were adopted. The adopted design was used for both MIPv6 and FMIPv6 simulation instances. In findings, by measuring performance metrics outputs, a constancy in values for all the metrics demonstrated how important reducing the handover latency was for the overall network performance as Fast Handover MIPv6 implementation process enhanced the overall network performance and ensured a better network QoS than the standard MIPv6. Throughput for instance, performed better in FMIPv6 than in MIPv6 for all services (HTTP, FTP, and Video Stream) through both TCP and UDP protocols. However, implementing MIPv6 did not give solution to unlikely failures in the future.

For specific objective 1, the research concluded that MIPv6 technology provided a convenient approach to implement and solve mobility problems in client-server environment provided that through MIPv6 mobile clients remain connected to the server when moving through different IPv6-based networks. The system designed and implemented in this research successfully addressed objective 2 by ensuring that clients and servers can securely communicate in the mobile environment with optimized packet routing using OMNET++ simulation environment. Network Quality of Service was measure in MIPv6 based on FTP, HTTP and Video Stream service to address specific objective 3. To provide solution of objective 4, network Quality of Service was later configured and evaluated on the same implemented MIPv6 topology. It demonstrated a higher performance level in term of network performance metrics providing better network Quality of Service. In recommendation, to avoid having a client-server network with a single point of failure, future researchers could look for ways to implement MIPv6 technology with more than one HA and with backup techniques for resources for the client to have considerably available Home Network services.

Finally, for further security, it was suggested that in future implementations of MIPv6 technology researchers may find a way to also render the server mobile by providing more capabilities in terms of storage, power capacity and mobility to possibly avoid potential and unnecessary security breaches from attackers, and in the process rendering the entire network mobile. ? ACKNOWLEDGEMENT First and foremost, I wish to express my highest heartfelt gratitude to Almighty God for giving me life, strength and opportunity to go through my academic life and for His incommensurable provision along the journey. To my life-coaches, my parents Jean-Claude Byamungu and Florence M’Mabwire. I could not have reached such a milestone without you. Many Thanks! I am forever grateful to my siblings, who have provided me through moral and emotional support in my life. Without them I may not have been who I am today.

I am equally grateful to the rest of my family members and friends who have supported me along the way. A very special gratitude goes to my supervisor, Dr. Gerald Chege, for the guidance, wisdom and patience demonstrated throughout this project. God Bless! DEDICATION I dedicate this work to my wonderful supporters, my parents Jean-Claude Byamungu and Florence M’Mabwire for their ever-ending support and encouragement which I will forever cherish.? TABLE OF CONTENTS STUDENT’S DECLARATION ii COPYRIGHT iii ABSTRACT iv ACKNOWLEDGEMENT vi DEDICATION vii LIST OF TABLES xiii LIST OF FIGURES xiv LIST OF ABBREVIATIONS xvi CHAPTER 1 1 1.0 INTRODUCTION 1 1.1. Background of the Study 1 1.2. Statement of the Problem 3 1.3. Purpose of the Study 5 1.3.1. General Objective 5 1.3.2. Specific Objectives 5 1.4. Rationale and Significance of the Study 5 1.5. Scope of the Study 6 1.6. Definition of Terms 7 1.7. Chapter Summary 9 1.8. Research Organization 10 CHAPTER 2 11 2.0 LITERATURE REVIEW 11 2.1. Introduction 11 2.2. Mobile IPv6 technology and Client-Server Network Mobility Problems 11 2.2.1. Theoretical Foundation 11 2.2.2. Network Mobility Problems in Client-Server Environment 13 2.3. MIPv6 Network Design and Underlying Technologies 15 2.3.1. IPv4 and IPv6 Review and Comparison 15 2.3.2. IPv4/IPv6 Transition and Coexistence 18 2.3.3. Network Mobility Concept 21 2.3.4. Mobile IPv6 Technology Overview 25 2.3.5. Mobile IPv6 Workflow 28 2.3.6. IPv6 Address Assignment 29 2.3.7. Route Optimization Techniques and Related Security Mechanism 30 2.3.8. Network Security in Mobile IPv6 35 2.3.9. Communication through IPsec 40 2.3.10. Fast Handover Technique in MIPv6 42 2.3.11. MIPv6 Technology Application in Internet of Things 46 2.4. MIPv6 Network Performance Evaluation 48 2.4.1. Network Performance Metrics 48 2.4.2. Performance Evaluation in MIPv6 50 2.4.3. Network Quality of Service Concept in Mobile IPv6 51 2.5. Research Approach 52 2.5.1. Network Factors and Proposed Solution 52 2.6. Chapter Summary 55 CHAPTER 3 57 3.0 RESEARCH METHODOLOGY 57 3.1. Introduction 57 3.2. Research Design 57 3.3. Data Collection Methods 58 3.3.1. Simulation Data Collection Methods 58 3.3.2. Research Instruments 60 3.3.3. Reliability 61 3.3.4. Validity 61 3.4. Research Procedures 62 3.5. Data Analysis Methods 65 3.6. Chapter Summary 66 CHAPTER 4 67 4.0 PROJECT IMPLEMENTATION 67 4.1. Introduction 67 4.2. Analysis 67 4.2.1. Simulation Approach 67 4.2.2. Discrete Event Simulation Approach 68 4.2.3. Overview of Different Simulators 70 4.2.4. Simulation Environment 72 4.2.5. Analysis of the Simulated Networks 74 4.3. Modeling and Design 75 4.3.1. OMNET++ Integrated Development Environment 75 4.3.2. Simulation Modeling 77 4.3.3. Simulation Model Accuracy 80 4.4. Simulation Implementation 81 4.4.1. MIPv6 Technology in Client-Server Architecture 81 4.4.2. Handover Latency Analysis 86 4.5. Testing 87 4.5.1. Overview 87 4.5.2. Testing the Proposed MIPv6 Client-Server Network 88 4.5.3. Fast Handover Mobile IPv6 Implementation 95 4.6. Chapter Summary 98 CHAPTER 5 100 5.0 RESULTS AND FINDINGS 100 5.1. Introduction 100 5.2. MIPv6 and Client-Server Mobility Problems 101 5.3. Design and Implementation of Client-Server MIPv6 102 5.4. Evaluation of Network QoS of Client-Server MIPv6 and FMIPv6 103 5.4.1. Simulated Results Summary 104 5.5. Chapter Summary 116 CHAPTER 6 117 6.0 DISCUSSION, CONCLUSIONS AND RECOMMANDATIONS 117 6.1. Introduction 117 6.2. Summary 117 6.2.1. Purpose of the Study 117 6.2.2. Specific Objectives 118 6.2.3. Research Methodology 118 6.2.4. Major Findings 119 6.3. Discussions 121 6.3.1. MIPv6 and Client-Server Mobility Problems 121 6.3.2. Design and Implementation of Client-Server MIPv6 124 6.3.3. Evaluation of Network QoS of Client-Server MIPv6 and FMIPv6 125 6.4. Conclusions 130 6.4.1. MIPv6 and Client-Server Mobility Problems 131 6.4.2. Design and Implementation of Client-Server MIPv6 132 6.4.3. Evaluation of Network QoS of Client-Server MIPv6 132 6.4.4. Evaluation of Client-Server FMIPv6 for Better QoS 133 6.5. Recommendations for the Project 133 6.6. Recommendations and Future Work 133 References 135 APPENDIX A: MOBILE IPV6 .NED SOURCE CODE 150 APPENDIX B: MOBILE IPV6 .INI SOURCE CODE 153 APPENDIX C: MOBILE IPV6 MODIFIED .INI SOURCE CODE FOR FMIPv6 163 APPENDIX D: OMNET++ SIMULATION ENVIRONMENT SCREENSHOTS AND TCP CHART BASED ON .ELOG 174 LIST OF TABLES Table 1: IPv4/IPv6 Fundamental Comparison 18 Table 2: Comparisons between MIPv6 and MIPv4 26 Table 3: MIPv6 Attacks and Possible Solutions 40 Table 4: Summary of Identified Gaps and Proposed Solution 52 Table 5: Comparisons between Simulation Approaches in Organizational Sciences 70 Table 6: Comparison between OMNET++ and NS2 Simulators 71 Table 7: MIPv6 Initial Network Configuration Parameters 92 Table 8: Traffic Model for Video Streaming Application 94 Table 9: Traffic Model for Web Application 94 Table 10: Traffic Model for FTP Application 95 Table 11: Processed Handover Latency for MIPv6 and FMIPv6 for Video Stream Services 105 ? LIST OF FIGURES Figure 1: Mobile IP/AAA Framework 12 Figure 2: Mobile IP AAA Framework Architecture 13 Figure 3: Manually Configured Tunnel in IPv4/IPv6 Tunneling Mechanism 20 Figure 4: IPv6 Host and IPv4 Host Connected through NAT-PT 20 Figure 5: Dual Stack IPv6/IPv4 Mechanism 21 Figure 6: NEMO Basic Architecture 23 Figure 7: Mobile IPv4 Network and IPv4 Addressing 26 Figure 8: System Architecture of Mixed IPv4/IPv6 Wireless Networks 27 Figure 9: Mobile IPv6 Workflow 29 Figure 10: Return Routability Procedure Messages 34 Figure 11: Forging Binding Update for Information Interception Mechanism) 36 Figure 12: The RHT2 Validation Process 39 Figure 13: MIPv6 Security Process 41 Figure 14: Handover Scenario in MIPv6 43 Figure 15: Fast Mobile IPv6 Handover Process 46 Figure 16: MIPv6 With and without Route Optimization 55 Figure 17: Procedural Steps for MIPv6 Wireless Network Design Using OMNET++ 65 Figure 18: Discrete Event Simulation Model Process 69 Figure 19: OMNET++ Modular Structure 73 Figure 20: Proposed Topology of MIPv6 Network Simulation Model in Client-Server Environment 79 Figure 21: Mobile IPv6 Client-Server Network Model 83 Figure 22: Mobile Node Modular Stucture 85 Figure 23: Handover Delay Results Report 106 Figure 24: UDP End-To-End Delay for Video Stream 107 Figure 25: UDP Video Stream Throughput Implementation Results 108 Figure 26: UDP Video Stream Packet Loss Rate Implementation Results 110 Figure 27: TCP FTP Throughput Implementation Results 111 Figure 28: TCP File Transfer Protocol Packet Error Rate Implementation Results 113 Figure 29: TCP HTTP Throughput Implementation Results 114 Figure 30: TCP HTTP Packet Error Rate Implementation Results 115 ? LIST OF ABBREVIATIONS AAA: Authentication Authorization and Accounting ACL: Access Control List AH IPsec: Authentication Header with Internet Protocol Security AP-ID: Access Point Identification AR: Access Router BAck: Binding Acknowledgement BU: Binding Updates CN: Correspondent Node CNGI-CERNET: China Next Generation Internet/ China Education and Research Network CoA: Care-of Address CoAP: Constrained Application Protocol CoAP-DPMIP: Constrained Application Protocol-Distributed Proxy MIPv6 CoAP-PMIP: Constrained Application Protocol-Proxy Mobile IPv6 DES: Discrete Event Simulation DHCPv6: Dynamic Host Configuration Protocol for IPv6 EDGE: Enhanced Data rates for FSM Evolution ESP IPSec: Encapsulating Security Payload with IPSec FA: Foreign Agent FMIPv6: Fast Handover MIPv6 FN: Foreign Network GPRS: General Packet Radio Service HA : Home Agent HN: Home Network HO: Handover HoA: Home Address ICMPv6: Internet Control Message Protocol version 6 IDS: Intrusion Detection Systems IETF: Internet Engineering Task Force IPX/SPX: Internetwork Packet Exchange/Sequenced Packet Exchange ISP: Internet Service Provider LTE: Long Term Evolution MAN: Metropolitan Area Network ME: Mobile Equipment MIPv6 : Mobile IP version 6 MN : Mobile Node MPLS: Multi-Protocol Label Switching MR: Mobile Router mSCTP: Mobile Stream Control Transmission protocol MT : Mobile Technology NAR: Next Access Router NAT-PT: Network Address Translation – Protocol Translation NCoA: Next Care-of Address NEMO: Network Mobility PSTN: Public Switched Telephone Network RADIUS: Remote Authentication Dial In User Service RFC: Request For Comments (a specification by IETF) RO: Route Optimization RRP: Return Routability Procedure SIIT: Stateless IP/ICMP Translation SNA: Systems Network Architecture WiMAX: Worldwide Interoperability for Microwave Access CHAPTER 1 1.0 INTRODUCTION Background of the Study In today’s Internet and Information Systems resource use, people have struggled integrating the notion of mobile Internet technologies within the very crucial and sustainable technology areas such as the client-server. From the individual to the corporate standpoint, the mobility gap along with the lack of an extended application of the handover and roaming techniques introduces the main problem toward enabling servers as well as their clients to fully and seamlessly transmit information to each other. This study elaborates on different technologies serving as basis toward establishing a sustainable and reliable mobile technology environment with a direct application in client-server setups.

There is a technical coexistence and compatibility between large coverage access networks such as 2.5/2.75G or GPRS/EDGE, 3G or UMTS, and 4G or LTE, with Local Area Networks (LANs) and dedicated short-range communications, making it possible for devices in both large and short-range coverage to exchange information and signals (Céspedes ; Sherman, 2015). Hence resource intensive technologies such as client-server services demonstrate a certain need to consider two specific network architectures that have been deemed to meet their capacity requirements (N. Zhang ; Bao, 2008). These architectures including heterogeneous cellular network where different coverage areas are determined by the transmission power, data throughput, and network density parameters intended for a determined area of coverage, in which way at a wider range, the cellular network becomes a combination of femtocells, microcells, and microcells. A heterogeneous radio access network architecture which also requires an internetworking and interoperability of different radio access technologies such as GPRS, WLAN, WiMAX, and LTE (Ma and Ma, 2014; Phoomikiattisak ; Bhatti, 2013).

According to Zhang ; Bao (2008), most active wireless and mobile networking in the future will have MIP as their common and enabling technology. Therefore, the expectation is that a large number of different technologies and devices running a variety of applications will be deployed, in a loosely couple environment where IP will be playing the role of the unifying architectural environment. Both IPv4 and IPv6 are considered capable of offering significant capabilities into implementing MIP in private and public addressing, but most of the time IPv6 is used instead of IPv4, due to its wide range of address space availability which utilizes a 128-bit address versus 32-bit address in IPv4 and evolved network security approach (Goralski, 2014). Client-server is one of the most prolific technologies today with a significant application in the ever-evolving wireless network, which is enabled by the interconnectivity and mobility potential of the network that is reliable due to its high-speed and sustainable hardware and continually evolving protocols. The explosion of certain mobile applications, based on Internet Protocol such as web or hybrid applications involving protocols including HTTP (web services), FTP, Video Streaming, etc., is the latest example and driving force showing that mobile wireless network is now the focus of technologies such as distributed computing (Fei, Xing, ; Li, 2014), and that to a certain extent would be applied in client-server environment. Users have embraced these technology advances with the proliferation of mobile computers in the form of laptops, palmtops and PDAs at its peak, and as important elements of the current computing environment.

Research reveals another theoretical approach where client-server architecture in a mobile environment is related to its application in mobile multiplayer games where the server stores and processes all the game data sent by all the connected mobile clients. The server therefore, only updates the clients with the particular data they need anytime, anywhere (Opoku, 2012). One of the most prominent mobile technologies (MT) being MIP has been applied in a handful of revolutionized mobile technologies based on the related released IETF standards, though the protocol is of two categories including MIPv4 and MIPv6. The concept is therefore referred to as the pathway to providing home network access extension to users by stretching the home network identity such as the IP address (S. Kim ; Kim, 2006). Some remote access application can then be used in this regard in an attempt to demonstrate the ultimate possibility of integrating distributed technologies of the likes of client-server, that are able to operate on WANs i.e.

across different LANs using MIP, or over a set of communication media such as RF, PSTN, etc. (Park et al., 2013). Statement of the Problem To circumvent the volatility characteristic and coverage limitations of WLANs and services that are transmitted within it, on the perspective of ensuring a consistent IP addressing scheme for its users across different LANs, network managers and other IT operators such as system administrators, and mobile technology technicians would face huddles that through the proposed solution that this research is seeking to realize, it will then be possible to cross-transmit information and network packets from a server in one LAN to a client in a separately different LAN whist baring a unique and unchanged IP address. Although this achievement is highly ambitious, Internet being the main medium through which all communications are transmitted, throughout the years, network and system administrators have strived to open a path to implementing a client-server environment that ensures the mobility and roaming of the IP address of the components due to certain characteristics of the Internet environment such as lack of security of information transmitted, lack of control, large propagation delay, multipath and packet collusion risks, and high transmission error probability which are significantly different from running the same client-server technology within LANs individually (Ning Zhang ; Hong Bao, 2008; Alomari, Sumari; Taghizadeh, 2011). Therefore, the solution proposed by the findings from this research contributes upon encountering much of these obstacles and proposing a friendly environment for mobile nodes and services. Internet Protocol Addressing and node identification in a mobile environment are on the first line of issues qualifying the hesitation and somehow the impossibility of implementing client-server technology within a mobile environment from the individual to the corporate standpoint.

As mentioned by Cisco systems (2012), IP issues ranging from adjusting the host IP address every time it repositions or moves to another LAN (or WLAN), DNS updates taking much longer, to TCP connection breaks and packets transmission security problems. In other words, if a node moves from its network to another without changing its IP address it will be unable to receive its packets, but also if a node changes its IP address it will have to terminate and restart its ongoing connections every time it moves to a new network (with different IP network prefix). These issues make it worrisome to implement technologies such as client-server with sensible applications and services containing personal data, time-sensitive services, etc… However, the mobile technology proposed in this research not only meets node mobility requirements in a wireless network, but also addresses issues related to the distribution and arrangement of IP addresses in such a way that neither the server nor the client get to lose their identification within their home LANs even when the client moves across other networks.

Though Mobile Technology (MT) specifications and proposed research solutions put forward in this project, it would be possible to address this issue since MT revolution in networks has endowed us with different approaches amongst which Mobile IPv6, which despite the Mobile Node (MN) relocation between different networks, connectivity at different covered positions is attained continuously with no user intervention. Therefore, Mobile IPv6 provides connectivity to nodes everywhere within home networks or away from home. Nevertheless, in this research, to reach the definite goal of implementing this technology within client-server environment, we exploit and implement IPv6 ability to provide Internet Protocol Security (IPsec) to network nodes and conveys it to the very MN dispatched across different networks through to ensure a secure communication among them, as Moravejosharieh, Modares, and Salleh (2012) stated: “IP Security (IPsec) in the IPv6 protocol can secure Mobile IPv6 more than IPv4”. Therefore, secure communication over the internet (insecure network) is essential and somehow vital to applications such as banking, online purchasing, money transfer, virtual private network (VPN), Electronic Mail sharing, and so on (Moravejosharieh, Modares, ; Salleh, 2012). On the other hand, routing information across different networks would eventually be possible with the technical abilities and research findings and implementation of a MIPv6.

Nevertheless, the technology is fundamentally based on the ability for the MN to bare the same IP address across networks through handover (HO) capabilities. But during the HO process, MIPv6 demonstrate some difficulties in preserving the system’s performance as the MN must attach and detach itself from one FA node to another continuously. However, to solve this kind of issue, based on the research findings, we proposed to apply a procedure that has the ability to speed up the seamless roaming process aiming at enhancing the network performance, effectiveness, and efficiency, driving toward enhancing the QoS of the implemented MIPv6 within client-server environment (Hung & Duong, 2011) which resulted in a Fast Handover Mobile IPv6 (FMIPv6) implementation. Purpose of the Study General Objective The general objective of this study was to enhance capacity and network performance of client-server architectures using Mobile IP version 6, Host-based network mobility protocol. Specific Objectives Evaluate MIPv6 technology and client-server network mobility problems though literature review and propose a solution framework.

Design and implement client-server architecture using an optimized and secure MIPv6 solution in a simulated environment. Evaluate network Quality of Service of the implemented MIPv6 solution for FTP, HTTP and Video Stream services. Implement and evaluate client-server Fast Handover MIPv6 solution for better quality of service. Rationale and Significance of the Study The findings of this study are of great importance because it discloses a particularity of MT use in the world of technology since client-server environment is no longer only considered as a local network technology, but one that has gained more momentum on wider scales, that integrating it with a revolutionary technology such as MIPv6, would not only connect clients and servers together but also allow them to maintain their network identity wherever they go. Using IPv6 protocol is also an important asset since the protocol encompasses internet protocol security, and routing optimization and handover management approach that allow not only a better QoS for the network, but also ensure secure communications between nodes communicating across different networks whilst using the same network identity or prefix. Therefore, for a researcher or an explorer willing to establish a global scale client-server technology environment, finding a solution to seamlessly hold clients in contact with server(s) for access to the system’s infrastructure and services, makes it profitable to the extent that the network is being expanded and securely using other networks’ infrastructure.

This is also deemed to reduce the need for potential conflicts of interest and differences between ISPs in terms of ‘hacking or signal interference’ problems between each other’s network in the race for internet access and competitive advantage, leading to hacking becoming less important than to focus on futuristic technology solutions. To scholars and other researchers, the results of this experience would allow them to gain further knowledge since it explored a complicated and complex notion of IP technology, implemented in much critical environment such as client server, including possibilities of a client and server being mobile. The whole technological idea is based upon MIPv6, and then adds more critical technologies such as IPsec, FTP, HTTP, Video streaming, and route optimization, implementation where clients will have to choose their own location as they move from a network to another with no interruptions. Scope of the Study This study was scoped from an international or global scale to a local and personal level since it uses IPv6 global addressing method to address different network interfaces within the proposed architecture. The study uses one client and one server to examine and extracts performance differences between the proposed MIPv6 and fast handover MIPv6 solutions for different services implemented in the client-server architecture.

For a better preview of handover movement and its downfall and contribution to the overall performance of the network, we limited the study to a total of 2 Access Points, hence two access networks across which the client moves while exchanging information and accessing services form the server. To attain the study objectives, we implemented thereafter, a faster mobility in MIPv6 technology to improve upon the network QoS. After assessing the feasibilities and all probabilities of implementation, we found that this study has the potentiality of being implemented in a real-life experiment, or in a simulation environment. However, we chose to limit the study to a standardized network simulation practice that could translate to the real-life experiment in the scenarios of meeting full architectural, technical and physical requirements (Chen, Lee, ; Chang, 2014). Definition of Terms The following terms enable the understanding of Mobile IP and NEMO technologies, and client-server topology that will be used throughout this research project (Sheikh, Singh, ; Afroz, 2016 ; Al-adhal, Khaled Mahmood, 2012).

Access Router (AR): a router through which a Mobile Node or Mobile Router access Internet. It is designed provides MN access to the network. Binding: is the mapping/association between Mobile Node’s Home Address and its current Care-of Address. It includes exchange of messages (including the Mobile Node’s Home Address and Care-of-Address) between the Mobile Node and Home Agent, and between the Mobile Node and the Correspondent Node that ensure a secure communication of information.

Care-of Address (CoA): a temporary IP address assigned to the Mobile Node by the Access Router considered as the Foreign Agent managing the Foreign Network. Correspondent Node (CN): a node that is independently communicating with the Mobile Node over the internet. Foreign Agent (FA): an entity (router) in a different network than Home Network. It is a Network Mobility-enabling router that helps Mobile Router or Mobile Node to manage mobility and bind to their home network. It serves as default router for registered MNs by tunneling or de-tunneling datagrams from the MN to the HA or from the HA to the MN through its Care-of-Address.

Handover Management: is a phenomenon that addresses concerns about maintaining the MN’s connection as it continues moving and as it changes its point of attachment to the system through another network. Home Address (HoA): is a static IP address that the Mobile Node has been assigned in the Home network, managed by Home Agent. Home Agent (HA): an entity (router) at home network which facilitate a Mobile Node movement. It tunnels packets for delivery to the MN when it is away from home while maintaining the location of the node.

Mobile Network Node (MNN): can be fix or a visiting node to a mobile network. It is treated like a normal node by a mobile network. Mobile Network Node can be reached via its static IP address that is attributed by the Home Network. Mobile Node (MN): is an IP-enabled device with mobility nature, changing it point of attachment as moving across IP networks while keeping track of its. Mobile Router (MR): is a router with all MN’s capacity, but still functioning as a router within the network. Mobility Agent (MA): in mobile networks, it refers to every entity or device that is involved in mobility function (Home and Foreign Agent, Mobile Router, Mobile Node, and so on).

Registration: when the MN is away from its Home Network, it is required to register its Care-of-Address with its Home Agent. Route Optimization: is a process that refers to a direct communication between the MN and CN without any routing help from additional network elements such as the Home Agent. Signaling Overhead: is considered as the amount of signaling messages that are needed to perform mobility management procedure during MN handover process. Tunneling: When the MN is away from home, it will only receive packets if the Home Agent is able to tunnel them to the Mobile Node’s registered Care-of-Address. The packet is encapsulated at the entry point of the tunnel and recapitulated at the exit point of the tunnel.

Chapter Summary This chapter introduced the profound concerns and potential remedy to enabling various existing network architectures such as client-server to integrate with internet protocol mobility. Client-server architectures can be implemented in various kinds of technologies. But for users and clients to stay connected to the server located over the internet even after leaving its current network or gateway, it requires a specific and reliable technology. Mobile Internet Protocol version 6 legitimately responded to the concern based on various technological standards and implementation capabilities of the technology in network architectures. Therefore, RFC 3775 Mobility Support in IPv6 was introduced by IETF in 2004 to practically prove and standardize the MIPv6 technology concept. So, MIPv6 places itself at the idealistic position offering MNs possibility to seamlessly connect and exchange services with the CNs online regardless of their location, i.e.

using different network identifications. The technology as standardized gives network security assurance based on the required IPsec protocol implementation in IPv6 and route optimization process in MIPv6. As the general objective, the specific objectives of this study emphasized on implementing MIPv6 technology in client-server architectures and ensuring optimized and secure routes for network traffic, with a milestone of achieving fast handover technology that would enable a faster and secure transmission of information and exchange of services between clients and the servers. The acceptance level and reliability of the technology enhanced its validity and usefulness by private as well as public or corporate entities. Employing MIPv6 in client-server architecture projects a more flourishing prospect of normal information exchange and better quality of service across services such as FTP, HTTP, as well as Video Streaming using the most important transport protocols, namely TCP and UDP, especially when the handover delay is broken down to a lower margin. Finally, as standardized, the MIPv6 approach implemented in this research was limited to one client and one server to provide the necessary measurements for the comparative evaluation between MIPv6 and Fast Handover MIPv6 technologies.

Therefore, using IPv6 protocol, MIPv6 provides capabilities of potentially implementing the network on a local, as well as global standard based on IETF proposed best practices; hence the network designed in this research project laid out opportunities to implement MIPv6 in a client-server network with one HA and one FA with the potentiality of being expanded and managed globally. Research Organization This research project was organized in 6 chapters. Chapter1 introduced the research study giving insight on important topics such as background of the study, statement of the problem, purpose of the study, rationale and significance of the study, scope of the study, definition of important terms, and a brief chapter summary. Chapter 2 introduced critical review of the literature related to the problem statement established in chapter 1, with subsections aligning with research specific objectives outlined in chapter 1, the proposed solution framework, and the chapter summary. Chapter 3 on the other hand, introduced methods and techniques used in research methodology with emphasis on the research design, data collection methods, research instruments, research procedure, data analysis methods, and a brief chapter summary. Chapter 4 laid out steps taken to design and implement the proposed network model, outlining some important sections such as Analysis of the proposed research concept, modeling and design of the solution, testing of the solution, as well as a brief chapter summary.

Furthermore, Chapter 5 outlined the simulation results and findings organized into a brief introduction, presentation of results as per specific objective, and chapter summary. Finally, Chapter 6 introduced the final phase highlighting research results discussion, conclusions, and recommendations for the project and for future work. ? CHAPTER 2 2.0 LITERATURE REVIEW Introduction This chapter covered a general review of the related literature on implementation and evaluation of MIPv6 technology with an emphasis on the unpredictability of mobile wireless environment that introduces undesirable delays causing the network to perform poorly. The introduction of MIPv6 is also yet to find a satisfactory handover delay time, hence still loss of access to services for a critical period.

The review introduced an approach that addresses the gap with expectations of improving the overall network performance. Multiple security issues are the direct effect of mobile wireless networks. Therefore, in the literature, we found ways to improve network security and provide secure a platform to the node that is usually accessing the much unknown networks. This chapter presented a summary and solution engagements (most technological) in terms of research framework underpinning the study. The review was organized in different sections highlighting the theoretical foundation of the study, review of literature on MIPv6 technology and network mobility in client-server environment, research approach in terms of proposed solution, and the chapter summary.

Mobile IPv6 technology and Client-Server Network Mobility Problems Theoretical Foundation In recent years, Mobile IP has been spread through different levels of application in a diverse number of technology applications and issues. But most of all, the notion has its grassroots from the late 90’s where the Mobile IP working group in connivance with the IETF working group continued to upgrade features and technology parameters with regard to novel requirements from individual to enterprise standpoint. A new charter was then rolled out outlining interactions between MIP and AAA protocols expected to serve manufacturers of mobile equipment, who are interested in incorporating various IETF standard protocols into their mobile products. Hence a milestone had been set to review the use of AAA protocols in MIP to support inter-domain as well as intra-domain mobility and dynamic Home Agent (HA) assignment (C. E.

Perkins, 2002). Subsequently, a new draft describing the Mobile IP technical and security requirements for AAA service has been created for consideration by the AAA working group in connivance with the MIP working group (Lin, Cheng, ; Liao, 2009). The draft used the model presented in Figure 1. Figure 1: Mobile IP/AAA Framework (Source: C. E. Perkins, 2002) When Internet Engineering Task Force (IETF) started focusing and working on the definition of a general AAA infrastructure for the support of roaming operations, it was soon found that the exact same infrastructure could also be useful for true mobile communications, mostly to support MIP Authentication, Authorization and Accounting (Lin et al., 2009).

In the perspective of the future all-Internet-Protocol networks, MIP has widely been adopted for user mobility management purposes, making it a friendly and reference for future mobile technology applications. To ensure a secure access to MIP network, the IETF working group defined and put forward the Mobile IP AAA infrastructure in RFC 2977. The processes of an AAA framework in MIP environment is shown in the Figure 2, where HA and the FA (Figure 1) in the home network and the foreign network, respectively are mobility management agents for the Mobile Node. Signals are exchanged between the three components before packets are delivered to a MN allowing an establishment of routing tables for future packet delivery to the MN (Lin et al., 2009). Figure 2: Mobile IP AAA Framework Architecture (Source: Lin et al., 2009) With ADA: Service Area of Home Agents SDA: Service Area of AAA servers These notions led to the practical improvement of MIP technology, that this research project implemented much improved technological mobility measures (with MIPv6) implemented in client-server architectures, where the mobile node is in charge of mobility management in the network.

In real-life implementation, to prevent security attacks such as eavesdropping, replay, and man-in-the-middle attacks to which the central server is likely to be vulnerable, some security measures such as implementing RADIUS and Diameter protocols are designed to provide a centralized network access management based on the AAA concept (G. Yang, Lei, Wang, Dou, ; Xie, 2014). Network Mobility Problems in Client-Server Environment Wireless Technology Limitations in Client-Server Architectures Wireless technology invention has revolutionized the concept of networks since it offers network entities and users freedom from the constraints of physical and wired network structure. At a relatively low cost, wireless technology and wireless Internet has become wildly available in portable PC, and in handheld phones today. It brings mobile users up to the level of exploiting the technology at their fingertips.

Based on Wireless technology, mobile networking introduced and promised the principle of “anything, anytime, anywhere” to users (Bai ; Williamson, 2004). All types of wireless technology LAN, MAN, WAN, etc…) have contracted numerous limitations in terms of range, mobility, infrastructure and limited capacity of mobile devices in order to expand services reach and allow the client to contact the server without restriction (Xiaorong, Jun, ; Shizhun, 2013). However, mobile technologies such as MIPv6, with all capabilities and security measures that the solution can provide, had it been implemented in a client-server environment, it would not only provide more flexibility and independence to clients, expands the availability of services wherever and whenever possible, but also protect critical services and traffic between the client and the server in the process. Due to the technological capabilities, special equipment and particular infrastructure needed to support network mobility management, it is quite difficult to deploy such networks. This project focuses more on the host-based network protocol (MIPv6) allowing a more independent and self-sufficient environment while nodes are connected to any Internet router or AR that has IPv6 capabilities.

So, we utilized MIPv6 to take advantage of a considerable number of its features. The implementation of this kind of network architecture acquires more substance from the ability of IPv6 technology to handle the IP addressing of a MN (client) as it moves through different networks using the MIPv6 mobility feature (Heydari, Kim, ; Yoo, 2016). So, since the architecture works based on MIPv6, QoS and security features are well managed by the IPv6 provisions and network administration decision making. This means that as MIPv6 technology has the potential to provide a certain level of security and QoS, client-server architectures could well accommodate the technology. Therefore, based on the technology, the network simulation planned by this research work in response to the research problem, we shall be able to simulate, demonstrate, and evaluate how FTP, HTTP, and Video Stream network services would perform between client and the server using technological characteristics of MIPv6.

MIPv6 Network Design and Underlying Technologies IPv4 and IPv6 Review and Comparison In today’s Internet and computer network environments, most communications between end-to-end nodes are using the IP protocol which assigns a unique identification or address to all nodes connected to the Internet and provides the mechanisms to transport data transmitted from a node to another. IPv4 was the first widely deployed Internet protocol standardized about 25 years ago (D. Rudolf, 2009). This protocol suffered and continued suffering several design problems, which tend to restrain the creation of new usages of the Internet. Among the issues surrounding this protocol is the lack of addresses, limited security, to name a few.

The need of addresses has expanded since VoIP clacked the Internet and started using very limited number of addresses (IPv4). The scenario is that VoIP as an example is a voice transmission technology over IP, which transfers the use of IP addresses to mobile phones. Since phone users are growing very fast in numbers (in billions) this address scheme was deemed to be exhausted and then create an unmanageable crisis in IP address allocation to users. Each vehicle in MVN is carrying tens of IP sensors and some multimedia devices, IoT and smarts objects integrated in 5G Networks will also require addresses.

Hence there will always be higher needs for IP addresses (Hyun, Li, Kim, Yoo, & Hong, 2015). The authority managing IPv4 globally is Internet Assigned Numbers Authority (IANA), while the protocol’s address space management is delegated to Regional Internet Registries (RIRs) which are responsible for their allocated regions to assign addresses to end users as well as to local Internet registries such as Internet Service Providers (ISPs) (World Telecommunication /ICT Policy Forum,WT/ICTPF 2013). The protocol is based on a 32-bit logical address which is a total of 4,294,967,296 billion unique addresses consisting of five classes, A, B, C, D and E. In IPv4 every class has 4 octets of 8 bits each. One specific example of IPv4 address is “”.

If the address was translated in binary form, it would have to look like this: “11000000.10101000.00000000.00000001”. From the 3rd of February 2011, the Internet Corporation for Assigned Names and Numbers released the very last block of IPv4 addresses showing the difficulties to allocate or expand IPv4 addresses to new companies in the future (Babatunde ; Al-debagy, 2014 ; Ali, 2012). Nevertheless, some workaround methods were proposed and implemented to encounter the address shortage issue. The most common methods are Network Address Translation (NAT), Classless Inter-Domain Routing (CIDR) and IPv4 Dynamic Host Configuration Protocol, but the issues with NAT and DHCP which presume that users only consume information but do not publish it. However, Internet is about much more than a simple one-way retrieval of information.

That is, users and subscriber to an Internet access service this NAT and DHCP (constantly changing IP address), would not be able to fully enjoy current and future applications dependent on the end-to-end Internet model (D. Rudolf, 2009). Moreover, this protocol enables many security flaws, both on the architectural level and implementation. In fact, private or secret communication over a public medium such as Internet requires some cryptographic services to protect the transmitted data from interception (sniffing), view or modification by a third party in transit. Although IPsec standard is applied for both IPv4 and IPv6, this standard is optional to IPv4. This leads us to the conclusion that IPv6 has practically learned many of the lessons from IPv4 not only in the area of architectural and address scheme improvement, but also in the area of security (Sailan, Hassan, ; Patel, 2009).

On the other hand, IPv6 outperforms IPv4 broadly on many important issues, the main advantage being that IPv6 has a larger address space with about 340 undecillion (2^128) IP addresses which is enough if we estimate that every human gets to use 3 IP address out of 7 billion people living on the earth (340 undecillion – 21 billion) and giving more reasons to migrate to IPv6 (Hyun et al., 2015). The protocol is also best known for better security with not only IPv6 specification mandating that all IPv6-enabled nodes must support the IP Security Protocol, but also including payload encryption and authentication of the source of the communication. Moreover, IPv6 provide better QoS with support for real-time traffic such as VoIP that includes built-in “labeled flows” mechanism similar to the service offered by Multi-Protocol Label Switching (MPLS), where routers can recognize end-to-end flows to which transmitted packets belong. IPv6 facilitates the connection of entities to the network through its auto-configuration mechanism known as “plug-and-play” and called stateless auto-configuration that speeds up network connections mostly in large IPv6 network, and where router provides the network prefix from router advertisement different from the stateful mechanism where DHCP server provides the address (Sailan et al., 2009). Furthermore, we have to emphasize that IPv6 also has its pride in enhancing mobility of users. In fact, the protocol was built with mobility in mind with the objective of allowing a mobile node to move from ink to link while maintaining the same home IP address.

However, IPv6 Neighbor Discovery and auto-configuration mechanisms transparently allow connection to the MN with no needs of any special devices. IPv6 had then to be an optimized protocol embodying the IPv4 best practices by removing obsolete IPv4 characteristics. In the same perspective, IPv6 improves addressing and routing hierarchy with the potential to be extensible, offering support to new options and extensions when needed. In its addressing scheme, IPv6 then uses a 128-bit addressing format represented by 16-bit hexadecimal number fields which are separated by full colons “:”, as the following example shows (Babatunde et al., 2014): 2031:1000:130E:0000:0000:093EC:0000:130B According to Babatunde and Al-debagy (2014), this format makes IPv6 protocol less messy and error-free.

However, this address can be narrowed by compressing the block of zeros to a single zero, and changing successive fields of zeros to double colons “::” as shown here: 2031:1000:130E::093EC:0:130B Table 1 highlights 12 key distinctive characteristics comparing IPv4 to IPv6 protocols. Table 1: IPv4/IPv6 Fundamental Comparison (Source: Babatunde & Al-debagy, 2014) However, despite the comparison and extreme differences and incompatibility between the two networks, the IETF Next Generation Transition Working Group (NGTRANS) has proposed many transition mechanisms to enable seamless and sustainable integration of IPv6 resources into current Networks, IPv4 networks (Punithavathani, D. Shalini & Sankaranarayanan, 2009). IPv4/IPv6 Transition and Coexistence As we already know, IPv6 is the considered the next generation of Internet technology, mostly based on its ability to solve all defects of IPv4 although the replace work seem slow because even the expects are unable to predict when to completely replace. IPv4 network is still stable and consistent protocol for the Internet. The cost to upgrade it to IPv6 is high that only big companies can afford to upgrade IPv4 network to IPv6 network for they are financially powerful and are able to replace or modify their current applications (companies like Google, YouTube, Facebook, Twitter, Amazon, etc.).

Since there is no need to modify applications to provide IPv6 services in IPv4 environment, in universities and in most companies such as ISPs, transition mechanism is more suitable technology in most datacenter. China for instance, has the largest IPv6 network called CNGI-CERNET2 (China Education and Research Network) that has an overwhelming number (over 100) of colleges and universities campus networks, as well as more than 10 research institutes that have connected to the IPv6 network. If all Chinese universities for example acquired the IPv6 technology and deployed it to their users, the transition would go a lot faster in China. But globally, the process suffers from skepticism based on how to make applications in IPv4 smoothly move to IPv6 network without intervention of neither users, nor developers, but also how to insure less changes in datacenters structure, because the less the better. This can only mean that the two protocols can still coexist for a long time (Sha et al., 2017). Only three mechanisms exist to support coexistence and packets transition from IPv4 to IPv6 such as Dual-stack, Tunneling, and Translation mechanisms (Sanguankotchakorn & Jaiton, 2008).

We shall remember that the two protocols are fundamentally incompatible, but through transition mechanisms, devices on both sides will communicate. Therefore, the ultimate goal of transition mechanisms is to provide transparent routing for nodes in IPv6 network so they can communicate with nodes in IPv4 networks. The mechanisms can be explained as follows: First is “Tunneling mechanism” that can be used to connect IPv4 and IPv6 networks that are not directly connected. In tunneling mechanism, IPv6 resources are routed and conveyed to another IPv6 network end but over an IPv4 network. Thus, as shown in Figure 3, IPv6 network resources may need to communicate with another IPv6 network over IPv6 end-to-end, but in the way is IPv4 network that IPv6 information has to traverse.

The typical scenario may be the case of a WAN/bank deployment, where the bank is using IPv6-enabled network across its sites, but the WAN in between the sites support only IPv4. The tunnel is designed to transport or ‘carry’ one protocol, which in this case is IPv6 over another protocol, in this case IPv4. It will then encapsulate IPv6 datagram inside IPv4 network enabling end-to-end communication. In the past, tunnels have also been used to support other protocols such as AppleTalk, IPX/SPX, SNA, etc. (Punithavathani, D. Shalini & Sankaranarayanan, 2009).

Figure 3: Manually Configured Tunnel in IPv4/IPv6 Tunneling Mechanism (Source: Punithavathani, D. Shalini & Sankaranarayanan, 2009) Second, is the “Translation” whose basic function is to translate IP packet. Punithavathani at el. (2009) mentioned that there is a wide range of translation mechanisms but most of mechanisms are based on the Stateless IP/ICMP Translation (SIIT) algorithm that is used as a basis of mechanisms such as Bump in the Stack (BIS) and Network Address Translation-Protocol Translation (NAT-PT). The first includes packets from an IPv4 application flowing into TCP/IPv4 protocol module where the identified packets are translated into IPv6 packets and to be forwarded to the IPv6 protocol module.

The second follows quite a different path involving NAT-PT which is a stateful IPv4/IPv6 translator and whose protocols are at the boundary between IPv4 and IPv6 networks. Each IPv4 node maintains a range of globally routable IP addresses, but when sessions are initiated across the IPv6/IPv6 boundary, theses addresses are assigned to IPv6 nodes which allow native IPv6 nodes and applications to well communicate with native IPv4 nodes and applications, and vice versa based on needs. Figure 4: IPv6 Host and IPv4 Host Connected through NAT-PT (Source: Sanguankotchakorn & Jaiton, 2008) And last is the “Dual-Stack mechanism” which is considered as the foundational and preferred mechanism since it provides the most natural way for IPv6 resources to be deployed where there is no ‘tunneling’ or ‘translation’ that need to be deployed for the end-to-end connectivity. In a Dual-Stack mechanism deployment, both IPv4 and IPv6 are operational on all network components such as routers, switches, firewalls, hosts, servers…). With minimal service disruptions, Dual-Stack technique allows the smoothest transition from IPv4 to IPv6 network environments.

It enables IPv6 in the IPv4 network along with associated features necessary to make IPv6 routable, secure and highly available as shown by the figure bellow. Dual-Stack nodes all understand and run both IPv4 and IPv6 separately (Sookun & Bassoo, 2016). Figure 5: Dual Stack IPv6/IPv4 Mechanism (Source: Ravi, A, & Periyasamy, 2014) Network Mobility Concept In new network architectures (cellular, client-server, etc.), emphasizing on ‘mobility management’ is important for the future of mobile networks. NEMO protocol introduces a new entity (or device) called Mobile Router (MR).

Every mobile network will have at least one MR. However, an MR is similar to a mobile node in Mobile IPv6, but instead of having a single Home Address (HoA), on its own it has one or more IP prefixes as the “Identifier”. It is only after it has established a “bidirectional tunnel” with the Home Agent that the MR distributes its mobile network’s prefixes (namely, Mobile Prefixes) through the tunnel to its Home Agent. Mobile Prefix (MP) of a mobile network is not dripped (leaked) to its AR; hence the AR never knows that it can reach the MP via the MR. But in turn, the HA announces the reachability to the MP. So, packets from and to the mobile network flow through the “bidirectional tunnel” created between the Mobile Router and the Home Agent to their destinations.

Note that mobility is transparent to the MNNs in the moving network (held by the MR) (Devarapalli, Wakikawa, Petrescu, & Thubert, 2005) . Clearly, NEMO architecture as shown by Figure 6 is a set of network devices such as MNNs which refer to an IP-based device operating in a mobile network. In this case, individually they do not need to be aware of the network’s mobility since they are connected to a MR that is assigned a CoA when it roams to another network outside its home network. The CoA is then sent by the MN to its HA for binding. Access Router (AR) is a network device that provides Internet access to the MR. CN is a remote IP device that could directly communicate with the MNNs over any IP network.

Also, FA is an AR located on the visited network that provides mechanisms to assist the MR into give access to home network to complete the mobility and handover process and transmit data. HA is an AR on the home network that enables the MR to roam and seamlessly stay attached to the home network. Home network is the network where MR belongs when it is not roaming, that is when it is still associated with the HA link, and that hold the MR’s IP address pool (Zhu, Wakikawa, & Zhang, 2011a). Since they are all network mobility entities, IP nodes such as MNN, CN, HA, FA and MR are all mobility agent (MA), and since they can perform mobility functions such as handover process, they are all involved in network mobility. Mobility Management solutions involved two principal stages including Information Control that occurs during the handover process constituting the actual mobility cost to process address configuration and Binding Unit, and Exchange Data Frames usually called Tunneling (L. J.

Zhang & Pierre, 2012). IP mobility management protocol is classified into two kinds: “host-based” and “network-based” protocols. In host-based mobility protocols, such as MIPv6, to maintain session in wireless environment, MNs must participate in the mobility-related signaling of the handoff process. It requires the MNs to have the mobility protocol stack with it.

On the other hand, network-based mobility management protocols such as Proxy Mobile IPv6 (PMIPv6) elaborate a different scenario where MNs are not required to participate in the mobility-related signaling since the serving network handles the mobility management on their behalf (Islam, Abdalla, Khalifa, Mahmoud, & Saeed, 2012). However, this research project focuses on host-based mobility management, where results of the performance of individual network entities (clients, server, routers, etc.) will support and formulate particular solution to the concerns in the problem statement. Figure 6: NEMO Basic Architecture (Source: McCarthy, Jakeman, & Edwards, 2009) IP mobility support for Mobile Nodes (MNs) is a standard that was proposed by the IETF to enable a MN to maintain its initial and Home IP address from wherever and whenever it attaches to a new network (taken as Foreign Agent) (Zhu, Wakikawa, & Zhang, 2011b). We can find a lot of mobility-supporting protocols among which Mobile IPv6, which is a host-based solution destined to handle the global mobility of host in the IPv6 networks and Internet as a whole, while PMIPv6 is an evolved version of MIPv6 and a network-based protocol for handling global mobility in the IPv6 network environments. However, this research emphasizes on a host-based version of the protocol based on the substantial demands of the network architecture that a three-tier client-server environment requires while tending to demonstrate the enhancement of its capacity form a legacy network to a MIPv6 network. That is why a modification of the host protocol stack is expected for operation in MIPv6, for instance a MN has to send Binding Updates messages for local registration.

MN must be able to actively send BU and receive Back to be acknowledged and access the network resources when needed. Mobility of IP technology has a leaning on more of a network mobility meaning than the other layers. In addition to network mobility, MIPv6 technology enables mobility of four(4) other layers (Wozniak, 2016). Physical/Link Layer Mobility This layer provides “Fast and seamless mobility”, though it is limited to one single technology, including “2G and 2.5G” cellular networks, or IEEE802.11b (Banerji & Chowdhury, 2013). The physical/link layer’s task will be to actively and severally report current channel conditions and link properties to higher layers, which might then adapt to mobility-experienced problems, particularly, the mobile network’s non-uniformity.

Network Layer Mobility This layer mobility is the only common layer within MIPv6 environment, but it is tricky to do it optimally. This layer handles the IP handover management and support for node mobility. Typically, different from physical connectivity, the main issue is to manage the change in network layer connectivity. Handover issue is addressed latter by the proposed fast handover MIPv6 (Phoomikiattisak & Bhatti, 2015).

Transport Layer Mobility The transport layer as well participate in the mobility management process of the mobile network. It mostly influences metrics related to the transport of packet through transport channels. For instance, the layer will decrease the “packet ratio” attributable to the fallible (imperfect) wireless channels and therefore the user’s quality when it provides reliable end-to-end knowledge delivery. However, each transport protocol (such as mSCTP (Ratola, 2004)) will require own and different solutions for the mobility. Application Layer Even though some existing applications already provide mobility support, we have to note that as for the transport layer, every application will require own and different solutions for the mobility (e.g., FTP, HTTP, SIP).

Application layer takes care of high level mobility tasks. Mobile IPv6 Technology Overview Mobile computing has taken steps forward in the past decade and is more and more widespread. This pushes a needed mobility support for Internet devices. MIPv6 is an Internet Protocol developed as a subset of IPv6 protocol to support mobile connections and mobility of mobile devices. It is an extension and update of the IETF Mobile IP standard designed to authenticate and serve mobile devices (called Mobile Nodes) using IPv6 addresses.

Handling mobility management and providing seamless mobile communications, MIPv6 is thought of opening the Mobile Internet Age. Following the current state and trend of Internet infrastructure, MIPv6 is overwhelmingly needed to provide not only internet and mobility services but also mobile security to mobile devices over the web. MIPv6 holds the general ideas of home network, encapsulation, HA, and CoA from MIPv4. However, MIPv6 presents a different philosophy and a much more-improved design than its predecessor. Indeed, a considerable number of Mobile IPv4 limitations such as the lack of satisfaction from people’s increasing demands, the lack of mobility of the protocol, or later mobility but too difficult to achieve, and other problems such as efficiency, security Triangle Road, etc.

(Sanguankotchakorn & Jaiton, 2008). Hence, as IPv6 inherited the IPv4 technological spectrum, MIPv6 is also a successor of MIPv4, borrowing a lot of concept from MIPv4 such as mobile node, home agent, home address, and foreign agent. However, the critical improvement from MIPv4 in to the standardized MIPv6 include abolishment of the field agent, routing optimization (use of new routing header), an increase in home address options, increased security, and so on. Figure 7 illustrates the architecture and entities used in MIPv4 and its IP addressing process: Figure 7: Mobile IPv4 Network and IPv4 Addressing (Source: Ravi et al., 2014) Therefore, based on information in Table 2, we could conclude that MIPv6 is much advanced and smoothly functioning protocol than its predecessor (MIPv4). Table 2: Comparisons between MIPv6 and MIPv4 (source: A.

Hasan & Al, 2014)) On the other hand, Xie and Narayanan, (2010) conducted some case studies investigating the possibility of marrying IPv4 and IPv6 based on their respective mobility implementation. They investigated all possible handover scenarios in IPv4/IPv6 mixed wireless mobile networks by elaborating mobility solutions for each handoff as shown in Figure7. The architecture shows various autonomous subnets connected to different internet backbone networks with IPv6 and IPv4 networks having separate subnets. However, one subnet is connected to both networks simultaneously.

After investigating different roaming scenarios of a dual-stack MN with a predominant IPv6 based HoA, the investigation lead to a conclusion that the MN follows the mobility solution of the subnet into which it is moving. The investigation also uncovered that the subnets that are connected directly to both IPv4 and IPv6 backbone networks are dual-stack networks. Even though both networks can have AR of FA handling the mobility of the MN, there has to two different CNs, with IPv4 CN connected to the IPv4 network and IPv6 CN connected to the IPv6 network. However, this technology as illustrated in in Figure 8 can be used in many practical mobile environments.

Figure 8 illustrates architectural and technological interaction and coexistence between MIPv4 and MIPv6 technologies based on both IPv6 and IPv6 protocols. Figure 8: System Architecture of Mixed IPv4/IPv6 Wireless Networks (Source: Xie and Narayanan, (2010) Mobile IPv6 Workflow The following procedure explains the flow of operation toward ensuring a well-connected MN within MIPv6 environment. Indeed, MIPv6 offers a way for MNs to seamlessly preserve connectivity while they travel across different areas or access networks. Every mobile node (MNs) is destined to have a home network with a permanent home IP address attributed to the MN. Additionally, in each home network we fin entities such as Home Agent (HA) in charge of tracking MNs as they move from home to different networks. MN received by the network through a broadcasting Access Router (AR), so by the time a MN leaves its home network and moves to the neighbor network, it obtains a new IP address known as Care-of Address (CoA).

The MN is then required to register this new IP address (CoA) with its HA through a Binding Update message which defends its authenticity, authorization, and integrity and is issued over an IPsec protocol opening a secure tunnel of communication between the MN and its HA. Thus, after the binding is received, the HA respond with a binding confirmation so that even as the MN moves to a foreign network, a Correspondent Node (CN) can still maintain communication with the MN using ‘indirect routing’ that is made of packets being relayed by the HA (Al-Kasasbeh, Al-Qutaish, & Al-Sarayreh, 2008). Route optimization is used in order to decrease overhead at the border router. Routing optimization will offer a way for both the MN and CN to forward packets to each other directly without sending or receiving them from the HA. With MIPv6 if there is no security mechanism such as IPsec, the CN does not know which MN sent the BU.

However, the BU is not secret, but it always needs to be sent from a legitimate MN. The network environment and Mobile IPv6 working mechanism are described in Figure 9. Figure 9 illustrates the MIPv6 workflow as previewed and standardized by IETF. Its offers insight on relationships among entities and the kind information they exchange.

Figure 9: Mobile IPv6 Workflow (Source: Modares, Moravejosharieh, Salleh, & Lloret, 2014) IPv6 Address Assignment When a user connects to a MIPv6 wireless network, his or her equipment (MN) will be assigned with an IPv6 address. For example, let’s consider the address below. The user is connected to the MIPv6 network with an assigned IPv6 address 2001:0b:3:0:a:b:c:d. The first part of the IPv6 address is in dark orange color.

This part is the network prefix address. However, the second part of the IPv6 address is in black and is known as the MN’s prefix address. Two kinds of address attribution are then in consideration. The IPv6 address can be formed by either “stateful address configuration mode” or “stateless configuration mode”.

Fist, Stateless address configuration mode assigns IPv6 addresses by combining the network prefix and a predefined MN’s prefix together. On the other hand, stateful address configuration mode follows DHCPv6 protocol. In stateful address configuration, the assigned IP address is the combination of the network address prefix and a temporary generated prefix address. But usually the IPv6 address of a MN is assigned by the “stateless address configuration” (Mrugalski, Wozniak, & Nowicki, 2010). Therefore, stateless address configuration mode appeared undoubtedly more useful for this project’s solution since it provides the MN with mobility and handover decision capabilities, hence host-based mobile network.

Route Optimization Techniques and Related Security Mechanism Mobile IPv6 technology was designed to provide solution to two critical problems at the same time. The first problem was to allow transport layer sessions with TCP, UDP, and SCTP-based transactions to continue even when the MN moves and changes its IP address. Secondly, it allows the MN to be reached through a static IP address called Home Address (HoA). That IP address is considered as the identifier of the MN. As mentioned in Section 2.3.4, the HA’s role is to intercept packets destined to the node and forward them by tunneling them to the MN current IP address (the CoA). The transport protocols use the HoA as a stationary identifier for the MN.

However, tunneling packets through the HA leads to longer paths, hence to a degraded network performance. To solve the performance issue, MIPv6 include Route Optimization that when used, the MN sends its CoA to the CN using BU messages. Anytime RO is used, the CN not only becomes the source of the packets it sends, but also acts as the first router in the packets transfer order, effectively performing packet routing to the MN’s CoA. In fact, the packets are logically routed to the CoA, and then virtually form the CoA to the HoA of the MN. The packet is consumed by the node at the CoA, but since each packet includes a routing header containing the MN’s HoA, the header allows then the MN to select a socket that is associated with its HoA instead of one with the CoA (Nikander, Arkko, Aura, Montenegro, & Nordmark, 2005).

Yousaf, Bauer and Wietfeld (2008) emphasized that RO procedure is initiated by the MN and the message is only successful in case the CN has MIPv6 support. Then, the MN can initiate a BU with the CN, which eliminates the need for HA to communicate and allows a communication over the shortest path via the MN’s CoA, which means that HA is no longer needed and can be disconnected if willing as illustrated in Figure 16. Packet routing from the Correspondent Node (CN) to the Home Agent (HA), right up to the MN can often result in elongated paths. Notwithstanding, Route Optimization (RO) is a narrowed path and direct communication between the MN and the CN where packets do no longer use the intermediary of HA in order to transit. In this research project, we emphasized on the CN storing the MN’s BU information which always contains the mapping between the MN and its current (currently used) CoA, which only works if MN updates the CN with the CoA each time it changes networks. This ensured a well improved network performance for all the implemented services.

Generally, optimizing routes in communications provides certain advantageous scenarios for MIPv6 network such as avoidance of congestion in the HN enabling the use of even lower-performance HA equipment even when the network is supporting millions of MNs. RO also allows the reduction of network load across the entire Internet and reduction of network performance measurements such as latency and jitter, giving greater likelihood of success in terms of network QoS in signaling as packet tunneling is avoided. Finally, the process introduces an improvement of robustness against network problems (partitions, congestion, etc….) since fewer routing path segments and fewer nodes are traversed (Johnson & Arkko, 2011). However, Route Optimization introduces security concerns and threats that need to be taken care of to ensure the MN can safely exchange information with the CN out of HA intervention. In fact, RO Process is realized through BU messages between MN and CN, where the CN creates a Binding Cache that keeps all information on the identity of the MN. The attackers goal here may be to corrupt the CN’s Binding Cache and then to cause packets to divert and be delivered to a wrong address.

This can be very devastating in causing DoS attack at both MN and CN, and even at the address that receives the unwanted packets. The attacker can also intentionally exhaust the resources of the MN, the HA, or the CN by exploiting features of the BU message. In order to analyze and mitigate attacks in MIPv6, we have to understand that some attacks may be more damaging than the others, some may represent acceptable risks, some may even be considered too expensive to be prevented (Nikander et al., 2005). Furthermore, in order to corrupt the Binding Cache, the attacker is obliged to send one or more messages into the network. The attackers are usually known as active, rather than passive attackers in this case, even though passive attackers still pose a threat.

Some potentially damaging attacks may include “Attacks against Address Owners” which also includes “Address Stealing” that usually occurs in case the BUs where not authenticated at all. This mode of attack also includes “Attacks against Secrecy and Integrity” where the attacker may establish spoofed connections with the CN pretending to be the MN by sending spoofed BUs or can even put himself between the MN and CN by sending spoofed BUs to both (man-in-the middle attack). Apart from attacks against the address owners, RO can also create DoS attacks where even the MN itself can be the target of the attack or can be used as vehicle of the attack while the actual target is elsewhere. DoS attack can also be classified into Basic DoS Attacks where by sending spoofed BUs messages, the attacker could be able to direct all packets exchanged between two IP nodes to a random or nonexistent address (es) taking authority to disrupt or stop communication between the nodes.

Another DoS attack is “Flooding”. This one redirects IP traffic to an arbitrary address, and then used to bomb the arbitrary IP address with excessive amounts of packets. The attacker can even spoof acknowledgements to the CN in this case (Zhu et al., 2011b). Therefore, MIPv6 as standardized by IETF, has been designed to prevent or mitigate a number of threats. The intent was to design security measures that were close to that of a static IPv4 based on Internet, with a less excessive cost in terms of packets, processing and delay. The aim is to insure alleged MN is indeed reachable both through its HoA and its CoA.

So to secure RO, security measures were classified into Return Routability Procedure, Quick Bindings Expiration, and “safe creation of State”. Based on its characteristics and its effectiveness in responding to security concerns, as put forward by this research, Return Routability Procedure constituted our focus in securing RO in MIPv6 networks. RR is a mechanism that gives nodes within the network, ability to verify if there is a node on the other end that can respond to packets sent to a specific IP address. The procedure uses checks that yield false positive responses if there is an attacker between the verifier and the verified (the address to be verified) addresses.

So, a basic RRP (Return Routability Procedure) mechanism consists of two checks, which are HoA check and CoA check, which are always included in the message pair: Home Test, Binding Update and Care-of Test, Binding Update. As depicted in Figure 10, Return Routability Procedure is performed as following (Radhakrishnan, Jamil, & Mehfuz, 2008): As mentioned above, RR Procedure provides a way of sharing a common key between the MN and the CN in order to authenticate a Binding Update and to verify the state of the MN (if it is still alive or not) at its claimed CoA (S. S. Hasan & Hassan, 2013). Beforehand, MIPv6 assures that there exists a pre-established security association between the MN and its HA via tunneling of communications by IPsec between MN and HA. The procedure consists of six steps: Step 1 and 2 When ready to start the process of RO, via the HA, MN sends a first message called Home Test Init (HoTI) that includes “Home Init Cookie N0” (called a nonce), to the CN.

Simultaneously, the MN sends a CoTI (Care-of-Test Init) message that includes a “Care-of Init Cookie N1” (another nonce), to the CN. Two source IP addresses, of HoTI and CoTI, are HoA and CoA of the MN, respectively. Step 3 and 4 Once the two messages are received (HoTI and CoTI), the CN replies with other two messages: HoT (Home Test) and CoT (Care-of Test), respectively. Before sending the replies, the CN prepare the HoT message which implicate setting a nonce (nonceHI) which is indexed by HI and used in generating the Home Keygen Token “KO”.

Similarly, for CoT, the CN also selects a nonce (nonceCI) which is indexed by CI, for use in generating the Care-of Keygen Token K1. After this step, the CN sends out the Home Test (HoT) message, including three parameters (N0, K0, and HI) all destined to the HoA, and as well, sends out the Care-of Test (CoT) message, including three parameters (N1, K1 and CI). Structurally, the indices (nonceHI and nonceCI) carried in HoT and CoT, are set to remind the CN of which nonce value is used into generating the Keys (K0 and K1), the number 0 and 1 distinguishingly representing Home and Care-of Cookies. In fact, the two generated tokens exchange between the MN and CN are useful to make sure of the aliveness of both HoA and CoA IP addresses.

Step 5 From HoT and CoT, the MN obtained Home and Care-of Keygen Tokens (K0 and K1). The role of these exchanged tokens is to test whether packets that are destined to the claimed addresses are properly and securely routed to the MN; which means that if the MN get these two messages correctly, then the MN is actually at the claimed IP addresses (HoA and CoA). Therefore, when the MN receives K0 and K1, it creates a Binding Key (BK) that is labeled as Kbm, which is generated from a hashing algorithm SHA1 (K0|K1). Where | symbolizes string concatenation. Consequently, Kbm becomes the secret key shared between the MN and the CN.

After a short while, the MN sends a Binding Update message to the CN, which contains HI, CI, and Message Authentication Code (or MAC) containing information on the CN’s address, CoA, the BU message itself, all protected by the Binding Key (Kbm). Step 6 Finally, we have to stressed that to ensure protection of the Binding Key, while the CN as expected receives the BU with MAC using Kbm (as MAC Key), it can rebuild the key (Kbm) dynamically and verify the validity with the sole help of Home Nonce Index (HI) and Care-of Nonce Index (CI). If the BU message is legal, then the CN sends back an acknowledgement with MAC message (message: Binding Acknowledgement) (Kavitha, Sreenivasa Murthy, & Zahoor Ul Huq, 2010). Figure 10 illustrates network security processes in Route Optimization techniques using Return Routability Procedure to secure communications between the MN and its CN. Figure 10: Return Routability Procedure Messages (Source: Kavitha et al., 2010) Network Security in Mobile IPv6 Security Risks and Network Attacks in MIPv6 For all mobility support solutions, it is very important to consider the logic of mobility which practically means that the location of a mobile may change at any given time, thus understand how to secure such plainly dynamic location updates. A wide range of security solutions exist depending upon particular mobility solution proposals.

For instance, HA-based solutions would call for secure communications between the MN and its HA, while routing-based solutions would call for global-routing security (Internet Society IS, 2012), but also how to secure DNS updates and queries. The relevance of this study stops at the level of integrating IPsec protocol across all the network entities serving different kinds of solution as the client (MN) needs of a secure connection to help secure critical communications from being accessed, read, or tempered. IPv6 as well as IPv4 networks are all served by a security technology which is IP Security (IPsec), and but based on literature, we found that it is mandatory for IPv6 protocol, but optional and somehow inexistent for IPv4. So, while the legacy (the existing) network based on IPv4 could not be fully update to support IPsec, IPv6 has relevantly realized all security requirements of IPsec that some scholars pointed out that IPv6 cannot exist without IPsec (Caicedo, Joshi, & Tuladhar, 2009). In comparison with Mobile IPv4, mobile IPv6 complements with new features of IPv6 making the MNs roam more smoothly in New Generation Networks (NGN), hence the mobile network can be accessed anywhere, anytime. Mobile IPv6 nonetheless, present many protocol implementation problems as MNs have to register its current IP address to HA and to other communication nodes such as CNs as they move to other networks away from its HA (Xianhua & Sui, 2011).

The security loopholes presented by this process would provide opportunities for “network hacking” in the location registration and handover processes. Therefore, most potential threats in MIPv6 will likely result in BU message without excluding the fact that the support for nodes’ mobility opens a real path to a certain level of vulnerability to attacks on individual nodes. The binding update message from MNs is adapted in HA or in CNs to create corresponding relations between the MN’s HoA and its CoA which is stored in the Binding Cache (BC). HA as well as CN first have to check in the BC to obtain the destination’s (MN’s) CoA so they can send data packets to other nodes.

Establishing corresponding relations (HA to CoA) which is key to BU will likely affects the security of the HA and CNs transmitting information to MN. Thus, this process is used by attacker to introduce security risks whose successful attacks can be categorized into four: Forged Binding Update Attack (FBUA) Consist of network attackers forging a certain binding update message and set a fake address for the CoA and leaving other parts of the message unchanged, the data packets would then be sent to a forged MN. This means the authentic mobile node would become not addressable. The attackers could also directly be using their personal addresses as message receivers so that data packets sent to the MNs would be intercepted by designated node. So, not only the authentic MN will not be addressable, but also this will make opportunities for attackers to reorient and obtain data packets resulting in information loss (Xiaorong et al., 2013).

Looking from another angle, the attackers could mask themselves and forge BU message to multiple CNs with the CoA set as the victim’s HoA. The data packets sent to MNs by CNs would be redirected to the victim’s node, thus creating a distributed reflection attack (DRA), but on the victim’s side, it might receive plenty of forged data simultaneously resulting in a DoS attack. The fore BU attack is shown in Figure 14. Figure 11: Forging Binding Update for Information Interception Mechanism (Source: Xiaorong et al., 2013) HA Option Attack (HAOA) MIPv6 introduces some new headers in data packets to realize the needed addressable feature of MN when changing its location within its communication links to HA.

Special treatment is then defined for those data packets while attackers also find a way to introduce their own new headers and options to induce hacking risks. As mentioned earlier, the HA “option header” defines 128 bits or 16 bytes to carry an IPv6 address that represents the original address belonging to the “objective option header” and that is processed only by the destination node that will make response in harmony with the IPv6 address defined in the HA address option. Attackers make use of HA option so they can hide the authentic original address and fill the void with their own address (Baig & Adeniye, 2011). Routing Header Attack MIPv6 uses two types of routing headers. The Routing Header of Type 0 (THT0) is meant to designate destination node to allow data packets arrival.

So, the next address illustrated in the routing header would ultimately be the following destination node for subsequent data packet transmission. On the other hand, Routing Header of Type 2 (RHT2) is destined to carry MN’s home address (HoA), allowing MN to correctly implement address registration and notify the corresponding relationship between HoA and the CoA to CNs resulting in the realization of routing optimization without relying on tunnel packet forwarding. Hence, RHT2 contains an IPv6 address that would allow data packets archive authentic destination node. However, attackers take profits of this routing header to hide destination node address using the address in RHT2 and therefore implement address redirection and information interception mechanism. Threats from Dynamic Mobile Prefix Discovery Mechanism MIPv6 agreements planned that MN in foreign network’s links could dynamically be able to obtain information about its home network’s link topology and configuration changes in case the MN is provided prefix discovery mechanism. This makes it possible then for the HoA of a MN to be modified all the time.

This ensures the MN addressable feature. Though, through eavesdropping attack technologies, attackers at communication path between MN and it HA can possibly obtain home link’s network topology and configuration parameters information, so during a normal transmission, they can modify the prefix data message resulting in MN losing its addressable feature. Security in Registration Process To manage and provide security basics to the registration process, an authentication scheme is provided for data packets receivers. The receiver first verifies the legitimacy of BU, then allocates resources and make redirection for subsequent communication. In accordance with network deployment scenarios, network security protection strategies in BU information sent to HA and CNs are as follows: firstly, CN verifies the accessibility of MN’s HoA to ensure the MN’s legitimacy; secondly, CNs verifies the accessibility of MN’s CoA, so as to ensure the CoA validity; and lastly, the final BU of MN carry authentication information that was obtained through the first two steps.

Therefore to prevent attackers from obtaining MN’s authentication information sent by CN, IPsec is used on the first two steps above for encryption transmission (Choudhary & Sekelsky, 2010). Strict verification of the HoA option and routing header validation aims to restrict the attacker’s ability to formulate or hide source addresses, making this operation very important. For instance, RHT2 insures data packets would reach authentic destination node or destination address. Through MN’s strict verification, the legality of routing header in data packet is guaranteed. This is considered as the key to prevent routing header attack which prompts the following verification process (Xiaorong et al., 2013): Figure 12: The RHT2 Validation Process (Source: Xiaorong et al., 2013) Furthermore, considering client-server environment, mobile environment, servers or CNs are more likely to be subject or destination of the majority of attacks and attempts to breach systems.

Therefore, any potential attacks to be detected, there is a need to have an IDS utilized on the server. If an attack is detected to an IP address shared with a client or MN, the client will be placed in “suspicious mode” during its operations where the attacker is assumed to be constantly performing IP scan. If a new attack is detected coming through the CoA assigned to the suspicious client, the client is immediately placed in the “malicious mode”, therefore will no longer be updated with new CoAs, which result in a blacklisted node that is shut down and has to be reset by the network administrator (Xiaorong et al., 2013). Table 4 presents potential attacks that MIPv6 networks can encounter and possible solutions. These attacks can occur in real-world client-server network architectures using MIPv6. Table 3: MIPv6 Attacks and Possible Solutions (Source: Moravejosharieh et al., 2012) Attacks Abbreviation Possible Solution Eavesdropping S Line Encryption ICMPv6 Attacks (through Denial of Service) DoS Access List for ICMPv6 Requests on Network Router.

Man-in-the-Middle MITM Authentication of Control Messages (CM) Manipulation of Binding Cache (through Denial of Service) DoS Authentication of Control Messages (CM) Session Stealing MN-SS CN-SS Authentication of Control Messages (CM) Unauthorized Access UZ Mobile Node (User) Authentication and ACL Communication through IPsec A network security procedure in MIPv6 is applied integrating IPsec into the BU as well as the information sharing processes. First, the MN wants to transmit a data packet to a CN. The MN here will have already provided the HoA as the source address in the packet header. Next, the MN proceeds by sending the BU list to see if it had already sent the BU message to the CN and then it gets the CoA from the entry for the CN.

If there is a match, the MN includes its CoA in the HoA option (RHT2). From there, the packet that has the HoA in its source goes immediately through IPsec. After the headers are encrypted and added, the HA option of type RHT2 is swapped with the “source address” of the packet header. Then once the CN received the data packet, the headers are processed as they appear in the packet (consecutive order). Therefore, this allows IPsec to always see the HoA in the source address of the data packet header. One of the security priorities is to avoid changing the security association every time a new CoA is defined, so this makes it best to use the HoA as the selector in IPsec.

Every time a packet is moving (on the path), the MN’s CoA is used for either source or destination address in the header. The only setback from IPsec is that the protocol does not encrypt the RHT2 header, nor does it encrypt the HoA option which shows the MN’s HoA in the header. But in our solution as we move forward, the Security Parameter Index (SPI) found in the header of the Internet Protocol Encapsulation Security Payload (IP-ESP) header, can be securely used to get access to the HoA which in this case is the source or destination address (Ebalard, 2010). Figure 16 illustrates security process in MIPv6 using IPsec techniques to protect packets transmitted between the MN and its HA. Figure 13: MIPv6 Security Process (Source: Moravejosharieh et al., 2012) Fast Handover Technique in MIPv6 Handover Latency in Mobile IPv6 Handover in mobile technology is the ability for a node or a Mobile Equipment (ME) to move from a wireless AP to another with no disruption of network services, while enjoying the plethora of the all-IP based services. Routing overhead is one of the main issues in realizing a MIP handover.

In NEMO routing overhead consequently makes the use of wireless network resources ineffective due to important and sensitive application packets loss. However, a proposed solution to this problem is the integration of “Fast-Handover” mobility techniques with Mobile IPv6 technology (Islam et al., 2012). MIPv6 is a standard protocol that is designed to manage MN movements between wireless IPv6 networks. Although MIPv6 is proposed to enable network applications and services to seamlessly operate at the required QoS either in a wired or wireless IP network, switching from a network to another create a time delay that the MN uses to identify the next network agent to connect to.

The nodes have to authenticate themselves each time they migrate and change AP, the authentication processes can take as long as two-minute gap before the MN re-authenticates to the network. This Handover latency of MIPv6 is the main cause of packet loss and poor performance, and appears too long for real-time and throughput-sensitive Applications and services such as VoIP, video conference and so on (Zheng & Sarikaya, 2009). Figure 11 illustrates the original handover procedure in the standard MIPv6 where the MN is moving toward a New Access Router requiring association and configuration of a CoA as its new IP address in the FN. Figure 14: Handover Scenario in MIPv6 (Source: Koodli, 2009b) Switching links from one AP to another and IP protocol operations, for a MN, is called handover.

This “handover latency” results from a standard handover procedure in MIPv6 (namely, movement detection, new configure Care-of-Address, and Binding updates). This process is indeed slow as shown in the figure bellow. However, to successfully implement the proposed research solution, we suggest Fast handover technology for MIPv6 (FMIPv6) to actively manage the handover processes and at best, reduce the latency to acceptable and better for real-time and throughput-demanding applications. The protocol helps reduce the handover latency and packet loss inherent with MIPv6 (Khan, 2010 & Koodli, 2009).

Handover latency can vary depending on physical and logical conditions of the simulated network elements, as just from a study by Lai, Sekercioglu, Jordan, and Pitsillides (2006), the latency was of approximately 2.6 seconds, whereas Vassiliou and Zinonos (2010) found a latency of about 3.6 seconds. Furthermore, the study conducted by Kwon, Kim, Bae, and Suh (2010) produced a handover latency of about 5 seconds. Addressing the Existing Handover Latency The focus in this research has been oriented toward predictive handover. Therefore, as previously mentioned, the delay elimination depends on factors such as “IP connectivity” latency, which in turn depends upon the movement detection latency, as well as the new CoA latency or delay.

Once the MN is connected and IP-capable on the new network (subnet) link, it then sends a BU to its HA and one or more CNs. Once its CNs successfully process the BU, the MN can receive packets at the NCoA. But the goal being that the MN has to be able to receive packets directly from CNs at its NCoA. This sharply depends on the BU latency as well as the IP connectivity latency. Therefore, the goal of FMIPv6 is to enable an MN to quickly detect that is has left its previous Access Router or previous network by providing the AP and the underlying network prefix information while the MN is still connected to its PAR or current subnet. Hence, the MN discovers available NAR or AP using link-layer-specific mechanisms, then request subnet information regarding one or more of the discovered AP, in the previous configuration the MN would have to do this when it’s already connected to the NAR.

After resolving an identifier associated with and AP, an MN keeps information called AP-ID or AR-Info. Then when the attachment to an AP with an Access Point ID (AP-ID) is done, the MN already knows the corresponding NAR coordinates including the network prefix, the IP address, and the L2 Address (Hung & Duong, 2011). Different kinds of messages are being sent to completely establish the link (Layer 2 and Layer 3 links), where the “Router Solicitation for Proxy Advertisement” (RtSolPr) message and “Proxy Router Advertisement” (PrRtAdv) message are used to enable MN movement detection. From both messages, the MN elaborates a prospective NCoA when is still attached to its PAR’s link.

Hence, the handover latency due to new network prefix discovery is eliminated. Additionally, this prospective NCoA can be used immediately after the MN is attached to the NAR’s link only when the MN has already received a “Fast Binding Acknowledgment” (FBAck) message from the NAR, prior to its movement. In the event it moves without receiving an FBAck message, the MN can still start using NCoA after announcing its attachment via an “Unsolicited Neighbor Advertisement” (UNA) message. So, NAR will immediately respond to this UNA message in case it wishes to provide a different IP address to be used by the MN.

Operating this way, NCoA configuration latency is then reduced. Also, in order to reduce the Binding Update (BU) latency, FMIPv6 specifies a binding between PCoA and NCoA. When still possible, the MN sends a “Fast Binding Update” (FBU) message to its PAR from its PAR’s link to establish this tunnel with the NAR; otherwise, the MN should send FBU message immediately after detecting attachment to the next network (NAR). However, an FBU message must comprise the “Binding Authorization Data for FMIPv6” (BADF) option in order to make sure that the MN that is able to establish a binding is the only legitimate MN that owns the PCoA. The tunnel will remain active until MN completes the BU with it CNs (Koodli, 2009a). Consequently, the PAR begins tunneling packets from PCoA to the NCoA.

However, unless the NAR can detect the MN’s presence, setting up a tunnel alone does not ensure that the MN shall receive packet as soon as it is attached to the NAR’s link. Another operation known as Neighbor Discovery that involves neighbor address resolution with Neighbor Solicitation and Neighbor Advertisement, normally result in considerable delay sometimes lasting multiple seconds. For example, one second separates subsequent address resolution retransmissions (by default of one second each) when arriving packets trigger the NAR to send Neighbor Solicitation before the MN attaches. To circumvent this delay, the MN has to announce its attachment immediately with an UNA message that allows the NAR to immediately forward packets to the MN. Conclusively, via tunnel establishment for PCoA and fast advertisement, the protocol properly provides expected forwarding of packets to the MN. The ARs will also have to exchange messages to confirm that the proposed NCoA is acceptable. Fast Binding Acknowledgement (FBack) can be delivered after the NAR considers the NCoA acceptable for use, and when an MN had already sent an FBU from the PAR’s link. There is a level of trust in the relationship between NAR and PAR that the NAR has to rely on before forwarding support for the MN where the NAR will create a forwarding entry for the NCoA subject to approval from the PAR, which the NAR trusts. At the NAR, for traffic like VoIP which are already in progress, handover traffic buffering will be inadequate. Lastly, AR could transfer its network-resident contexts including the network’s QoS, Access Control (AC), and Header compression, in conjunction with handover. After all these operations, the protocol (FMIPv6) provides “Handover Initiate” (HI) message and “Handover Acknowledge” (Hack) message that should be used, and the ARs establish the necessary security association to secure communications(Hung & Duong, 2011). Figure 15: Fast Mobile IPv6 Handover Process (Source: Koodli, 2009b) MIPv6 Technology Application in Internet of Things IoT is projected as the image of the most prolific and important mobile technology of the next generation of Internet. MIPv6 technology as it stands, however, is one of the best contributors for the success of IoT. IoT is an environment where a variety of sensor devices and smart objects are connected to the Internet. For instance, IoT is enabling a particularly new generation of ecosystems in environment such as hospitals, smart cities; and dynamic ecosystems require an interoperable mobility protocol with existing Internet infrastructure, ubiquitous access to Internet, seamless handover and flexible roaming policies. IoT devices are usually constrained devices with low processing power, low memory, communication, and energy capabilities. So, those cited features and capabilities appear as challenges for IoT devices. However, firstly, to solve the battery power problem exhibited by self-dependent and constrained sensor devices, the IETF recently standardized the Constrained Application Protocol (CoAP) (Choi & Koh, 2016 ; Chun, Kim, & Park, 2015). Secondly, a more of a lightweight version of MIPv6 and IPsec for constrained devices have been proposed, which is aware of requirements of the IoT and present the best solution to manage the ubiquitous and dynamic ecosystems in terms of efficiency and security (Jara, Fernandez, Lopez, Zamora, & Skarmeta, 2013). More importantly, we had to focus on one of the critical issues in IoT, that is Mobility Management which at best provides seamless data transmission for mobile sensor devices (Jara, Zamora, & Skarmeta, 2010). So far, some work has already been done on IoT mobility management. Ganz, Li, Barnaghi, and Harai, (2012) proposed a resource mobility scheme for service continuity in IoT networks but using the tunneling between old and new gateways is still inherent to the problem of non-optimization. However, the more inclusive and effective solution has been proposed by Choi and Koh, (2016) suggesting two network-based mobility management schemes based on Proxy Mobile IPv6 (PMIPv6) and CoAP: CoAP-PMIP and CoAP-DPMIP. Practically, the typical applications and services in IoT such as IoT-based healthcare services are design to allow a doctor to monitor through mobile (and portable) sensors the status of moving patients by using the CoAP-based IoT communication so that in case of an emergency, the doctor should be able to provide the appropriate care to the patient. Finally, as for an IoT environment, MIPv6 protocol is considered the most studied and well-known internet protocol to provide IPv6 networks. Due to the enormous overload for mobile node that is involved in all the handover processes, sometimes with very weighty messages and high processing requirement that create some downtime for some IoT-based services, this protocol was not suitable for 6LoWPAN networks. So, in order to make it suitable and favorable for IoT resources, a more lightweight version of MIPv6 and its security protocol (lightweight IPsec) should be carried out (Jara et al., 2013 ; Jara, Fernandez, Lopez, Zamora, & Skarmeta, 2014). MIPv6 Network Performance Evaluation Network Performance Metrics Generally, the notion of network modeling and simulation has long been used by researchers, practitioners and students to analyze the behavior and evaluate performance of computer networks. Since this research project introduces a simulation of mobile network, we considered well-known fundamentals of event-based simulation approach with a discrete-event simulator to evaluate and analyze network behavior. However, it is advisable to consider evaluating simulated network’s performance at different layers, which enables an extraction of accurate information from a simulation (Suárez, Nuño, Granda, & García, 2015). We could present the classification of network performance metrics as follows: Link layer Metrics The most considered measurement at this layer is the network throughput, which is mostly seen as the ratio of bits received to the duration of the transmission. However, throughput can also be calculated using another metric called Nominal Channel Capacity (NCC), which determines the maximum number of bits that can be transmitted per unit of time (Mühleisen, Jennen, & Kirsche, 2010). Layer 2 also includes transmission error measurements among which we can consider two metrics: Bit-Error Rate (BER), which is the ratio of the number of received bits that have been altered during transmission on a communication channel to the number of bits sent and usually critical in connection-oriented services with TCP protocol; and Packet-Error Rate (PER) which represents the ratio of the number of data packets that were incorrectly received to the number of received data packets, both used to evaluate data transmission errors (Abukharis, Alzubi, Alzubi, & Alamri, 2014). However, PER is the most considered because if a bit is an error it is likely that the following bit is an error too in the same data packet. Thus, bit error rate is not independent. Therefore, taking into account the whole packet appears more efficient than measuring bit after another. Furthermore, Packet Loss Rate is an almost similar metric to PER but it focuses of the loss of a complete packet and is frequently used to measure loss of packets in real-time applications (Kukhmay, Glasman, Peregudov, & Logunov, 2006 ; Goudru & Vijayakumar, 2016). Other metrics can be used at this layer such as Signal-to-Interference-plus-Noise Ratio (SINR) in wireless networks where the interference power of other signals is also taken into account. This same metric in wired networks is represented by Signal-to-Noise Ratio to calculate the ratio of the signal power to the background noise, usually caused by the medium and both transmitting ends (Orooji, Soltanmohammadi, & Naraghi-Pour, 2013). Internet Layer Metrics This layer focuses on estimating the performance metrics of two main routing tasks within a network. The first task pertains to path selection process, which tries to determine the best path to a destination node to be adopted by evaluating the cost of the path between source and destination. This task process can be used in both wired and wireless networks. In wired networks, it includes several among which we can cite hop count, which is tasked to find the hope-count routing between two nodes (source and destination); it also includes Round Trip Time (RTT), tasked to measure the round trip delay of unicast probes (signaling) between neighboring nodes (D.-Y. Zhang, Hu, Zhang, & Kang, 2005). In wireless networks, we can still use the aforementioned metrics, but also other metrics such as Expected Transmission Count (ETX) and Expected Transmission Time (ETT), which are respectively used to predict or determine the likely number of retransmissions needed to send a data packet, and to determine the exact time a packet needs to be correctly transmitted over a link. On the other hander, the second metric which pertains to network topology management which establishes how data is transmitted (or forwarded) through the network while controlling the way network nodes are connected with each other. The routing scheme is always affected by the topology that we design the network in, its complexity and scalability, which means that data communications on higher layers will depend on the organization and management of the underlying network. Some commonly used metrics at this layer are betweenness to determine the shortest path between two nodes that likely goes through a specific node, and node degree that calculates the number of nodes that likely depend on a specific node (Sheikhahmadi, Nematbakhsh, & Shokrollahi, 2015). Transport and Application Layer In network performance evaluation, the most used measurement is “throughput”. Throughput can be determined at every layer to analyze the protocol behavior and to estimate their impact on the overall performance, which makes it possible to be used at transport and application layer, even though we had already mentioned it as a link layer metric. Therefore, both TCP and UDP throughputs are regularly used at transport layer. Other metrics, which are time-related such as end-to-end delay and jitter, also exist at the transport layer. In fact, packet end-to-end delay represents the amount of time required to transmit a packet along the path from source to destination, whereas jitter is as the packet delay variation. While both of these metrics can still be analyzed at the network layer, measuring them at the transport layer allows to include several other protocol features related to additional delays introduced by checksum verification or buffering (Madhuri & Reddy, 2016) . Finally, application layer also includes throughput measurement that is also known as “goodput”, introduced to calculate the total number of bits received at the receiver’s application layer divided by the simulation time. (Mühleisen et al., 2010) However, many other metrics can be used at this layer to evaluate the performance of each application protocol. Performance Evaluation in MIPv6 MIPv6 was standardized in extension of the existing protocols in order to support mobile users. The mobility factor that introduces handover delays could be impactful to the holistic concept of network performance where the protocol is charged of intercepting and forwarding packets to a MN and a possibly roaming node. However, seamless roaming of the MN recognizes that users and application do not have to experience loss of connectivity or for the less, any noticeable interruption in traffic. To understand this phenomenon, we have to stress that the impact does not only apply for highly delay-sensitive applications and traffic such as real-time communication traffic (based on UDP), but also for TCP-based applications as their performance is also very critical and sensitive to packet loss, packet errors and reordering. Therefore, as very helpful as MIPv6 could stand, it is imperative that handover process is initiated in such a way that it does not only ensure seamless communication between network nodes, but also and more importantly, minimize the latency and packet loss. Thus, to achieve this, this research project led to the implementation of MIPv6 technology in a client-server architecture using a simulated wireless IEEE 802.11b network in OMNET++ simulation environment. We later presented the enhancement of MIPv6 with regard to the handover latency (Fast Handover MIPv6) in client-server architecture, implementing and testing traffic models such as HTTP, FTP, and Video stream, all based on both TCP and UDP protocols (Johnny Lai et al., 2006). To establish differences in network performance, we considered a set of network performance metrics such as Throughput, Handover Delay, End-to-End Delay, Packet Error Rate and Packet Loss Rate. Network Quality of Service Concept in Mobile IPv6 Quality of service is considered as an ability of network entities and resources (e.g. host, router, application, etc.) to provide some performance level in terms of assurance for consistent network data delivery. It can also be defined as the ability for a network to provide different priorities to application, users, links, data flows, etc., and still guarantee a certain level of performance to (Hussien et al., 2010). In this project we measured the overall performance of the MIPv6 network by examining its QoS parameters such as throughput, handover latency, end-to-end delay, packet loss, packet error rate. Taking aim at the client-server architecture, QoS measurement is performed in multiple ways to indicate how well the server is delivering applications and services to the clients. With the mesmerizing trend in mobile network development, the never-ending innovation in mobile devices (laptop and palmtop computers, PDAs, mobile phones) and mobile technologies (mobile technology generations), including the rapid emergence of new real-time applications, QoS has become a very necessary asset in mobile environment since for instance, MNs will expect to get access to these real-time and multi-media applications such as video conference, Video on Demand (VoD), VoIP, etc., available in its home network (traditional or mobile) when it moves to anther network (over the Internet). So, one of the most revolutionary intentions of the next generation IP network has been to provide better QoS for real-time traffic and applications and provide seamless communication of information that would be transmitted via mobile networks (such MIPv6 networks) infrastructure. Research Approach Network Factors and Proposed Solution Table 4 corroborates the conceptual substance, technical and critical considerations that led to the realization of this research project. It presents major technical and technological differences between the existing client-server architecture concepts with MIPv6 technology. The proposed solution integrated the two phenomena considering the best factors, then introduced an extension of MIPv6 technology with less handover latency and better performance, laying the ground work for the network simulation process which should result in performance evaluation of the two MIPv6 approaches. Table 4: Summary of Identified Gaps and Proposed Solution (Source: Author) Factors Client-server Architecture Mobile IPv6 Proposed Solution: Mobile IPv6 Technology in Client-Server Architecture with Fast Handover Procedure. Bandwidth (based on IEE802.11b and Up) High High High Communication Mode Client/server: Host to Host Peer-to-Peer/client-server: Host to Host Client-Server: Host to Host Compatibility with Legacy Internet Yes Yes Yes Cross-Layer Information Non Yes Yes Error Rate/Packet Loss Medium in WLAN High Medium Handover Latency High Medium Low Network Coverage Range Low High High Network Mobility (Characteristic) No (in WAN)/Yes (in WLAN). Yes Yes (WLAN/WAN) Route Optimization Non Existent Mandatory Mandatory Network Security Low-Moderate Moderate Moderate Network Topology Changes Low High Moderate Scalability Moderate-High Moderate Moderate-High The framework includes specific factors that can influence performance expectations of both client-server and MIPv6 technologies. In both approaches, we only considered a wireless environment since the end-result of the sought solution is entirely based on a mobile wireless network. Despite significant limits that WLAN network encompasses, it still holds important factors that if implemented in MIPv6, it could improve the technology and provide client-server architecture with higher performance level. However, the framework not only includes wireless factors, but also introduces differences and technology improvements of the legacy internet that if implemented in MIPv6 could help reach the ultimate goal of reducing the handover latency, hence improving the network performance. In wireless environment, the proposed factors such as bandwidth, network topology, scalability are all preliminary and decision-based factors that play an important role in the overall network performance. On the other hand, other factor such as network security, mobility, handover latency, packet loss and cross-layer information are directly used and are very influential into establishing performance differences between the existing and the proposed systems. Therefore, the proposed solution suggests a decrease in the handover delay, which is the main and most influential factor that improves the overall network performance in terms of Quality of Service, if reduced. The aim is to combine multiple factors together and find a way to improve them in order to meet the project expectations. The proposed network is mobile with ability to expand in capacity and to longer distances, and can potentially allow any MN that has a home address identification anywhere, anytime to access network resources once it’s provided a CoA by the visited network. The network solution is based on a host-to-host communication mode with reasonably higher bandwidth triggered by low handover latency preferences in simulation procedure. One of the most important network performance metrics was “packet loss rate”, which measured the ability of the network to prevent harmful losses of packets while the mobile is transitioning from its home to the visited network and accessing real-time applications. The narrowed handover latency may reduce the time it takes for the node to resume accessing services, packet loss rate is then likely to decrease, insuring a better network performance. As configured in subsection 4.5.6, network coverage is large and dependent of the wireless network standard that we chose for the implementation as suggested by IETF (Jun, Liyan, & Pierre, 2009). But since it is mobile, the network is to be established and used with no boundaries in space and time as long as the MIPv6-supported Mobile Node is able to contact MIPv6-supported router and access internet as it moves. As the MIPv6 architecture is expected to support changes in network topology, it opens room for a level of reliability and scalability, which consequently introduces some security concerns that the proposed solution addresses. Route optimization is one of the most important factors into ensuring an optimize routing of packets. In client-server environment, it would not exist since the technology is only achievable in Mobile IP. So, to ensure optimization in IPv6 routing, route RO process has been adopted in MIPv6 and can be applied in different extensions such as FMIPv6, PMIPv6, etc. Figure 16 illustrate RO process in MIPv6 where packets sent through the network form the CN or MN to the MN or CN are directed routed without transitioning to the HA. Figure 16: MIPv6 With and without Route Optimization (source: Author) Chapter Summary This chapter presented a general review of the important literature on MIPv6 technology. The very initial concern of integrating mobility in IP technology was to find security measures that could respond to the obvious threats and likely attacks to mobile nodes. In that regard, IETF working group gave ability to manufacturers to incorporate AAA protocols in their mobile products in support of roaming operations. The development of MIP technology had to take into account the legacy internet, which up to that point was IPv4 Internet. As the research project aim was to simulate MIPv6 network technology, we had to first establish technical and technological differences between the legacy IPv4 technology and IPv6 demonstrating how the two protocol can coexist. In fact IPv6 came out correcting some shortcomings from IPv4 and so did MIPv6 from MIPv4, and the two MIP technologies also gave a room for transition and coexistence procedures. MIPv6 presents beneficial prospects for new generation technologies that most of the newly developed and projected technologies such Internet of Things are leading the charge in adopting IPv6 protocol technology and management. To exchange information with any node on Internet, MN always needed to establish a secure tunnel with it HA that conveys the packets to CN. However, tunneling packets through the HA led to longer paths and could create extra signaling noise. Therefore, RO process was implemented to narrow the path with a direct transmission of data between MN and CN. Unfortunately, MIPv6 technology and the RO process introduced security concerns that could be harmful to both HA and FA networks, to the MN and CN as well. Nevertheless, we identified and elaborated security measures to protect both MN registration and RO processes. Even with MIPv6 technology standards as defined by IETF, network mobility still produces an alarming level of handover latency. Therefore for critical areas such as client-server, it is important to preserve the integrity and availability of packets sent by the server. To reduce the delay, Fast Handover technology was introduced in extension to MIPv6. This led us into establishing performance evaluation metrics that would help compare the two implementation approaches performance. Finally, the research approach took into account the advantages and potential that MIPv6 could offer if it was implemented in client-server architecture. Additionally, the improvement of MIPv6 handover conditions would provide mode reliability in services that the server has to offer to client, most important of them being real-time services. CHAPTER 3 3.0 RESEARCH METHODOLOGY Introduction This chapter provides an exposure and discussion of specific points that depict the research methodology used in this study. It emphasizes on the steps that were involved, the data that was collected, how it was collected, the experiments that were conducted, tools used for the experiment, and how data was extracted and analyzed. The chapter structure included Research Design, Data collection method with emphasis on Research Instruments, Reliability and Validity, then we elaborated the Research Procedure, and finally Data Analysis. Research Design According to Cooper and SChindler (2011), research design is perceived as a general blueprint for the collection, measurement and analysis of data, with the focal goal of solving the research problem. It includes the plan of the research, from the well-thought-out and inclusive hypothesis to the final operational application and analysis of discovered data. Thus, this study used a comprehensive description and inclusive experiments of the Mobile Internet Protocol version six characteristics and its technological influences that it could have on client-server architectures suggesting a setup of several LANs with mobility capabilities and ability to manage the moving mobile nodes. This project implements an architecture of client-server based on Mobile IPv6 and proposes upgrades to ensure server services continuity and node mobility management across different networks, seamlessly providing services to clients, and with IPsec protocol implementation and Return Routability security procedure, providing a secure and trusted platform for the network entities. The proposed architecture was used to implement client-server model in MIPv6 environment using OMNET++5.2 with INET framework version 3.6.4 (Al-Rubaye, Aguirre, & Seitz, 2016 ; John Lai, Wu, Varga, Sekercioglu, & Egan, 2002). The simulation environment included different simulation packages and corroborated the technology used that improved MIPv6 operations basic principles by means of handover process, route optimization and tunneling mechanism. OMNET++ 5.2 is an object-oriented, discrete event simulation tool that essentially supports modularity, and is Open Source under its Academic Public License. The project implementation involved a process that enhanced the performance expectations by implementing Fast Handover MIPv6 in extension MIPv6. This new measure allowed us to reduce the handover delay by timely reducing the most impacting parameters in MIPv6 implementation (Koodli, 2009b). The network covered a large geographical area with two different network that handover their services to each other to serve the client. The MN or client was set to be moving across the networks without losing connection with the CN or server and while keeping its original IP address for identification on internet. All the modifications that were applied to archive a shorter handover delay, hence a better network QoS, were set in a different programming language than the one used to define the network itself. However, all the modifications were brought up to adjust the handover impacting parameters to Fast Handover MIPv6 requirements, whilst all the implemented services definition and configuration remained intact. Data Collection Methods Simulation Data Collection Methods As a design research project, this study involved a comparative approach of two logically implemented methods through simulation. The research required a logical experiment using modeling and simulating of MIPv6 and Fast Handover MIPv6 technologies in client-server architecture. Both methods were implemented with security and route optimization processes, which positively responded to the research objectives. Data collection methods involved a dynamic data rate selection method that included different bitrate values. Through this method, eight (8) different bitrates simulation instances were conducted, and a quantitative comparison of performance improvements between MIPv6 and Fast Handover MIPv6 in terms of network QoS was obtained by the application of handover extensions and combinations of various technical schemes. The proposed client-server architecture was tested through the use and test of application services such as File Transfer Protocol (FTP), Hypertext Transfer Protocol (HTTP), and Video Stream. Video Stream and FTP services were configured using a dynamic data rate selection mode (6 Mbps, 9 Mbps, 12 Mbps, 18 Mbps, 24 Mbps, 32 Mbps, 48 Mbps and 54 Mbps) with different bitrates values that helped record and collect result on different simulation instances (Tramarin, Vitturi, & Luvisotto, 2017), whereas HTTP was configured with a fixed bitrate value, all based on IEEE 802.11b. The simulated architecture then provided insights and leading points toward realizing the actual network’s goal. We considered network architecture with potential design expansion in terms of the number of MN, even though we only considered one MN moving between two distant APs in two different subnets between the Home Network (HN) and a Foreign Network (FN) in MIPv6. After a successful MIPv6 implementation, we implemented FMIPv6 on the same network to evaluate the handover latency and service performance of the network with fast handover process. Simulations and network performance of handover latencies could then be compared with respect to the implemented and tested services. The results obtained from the simulation were used to investigate different handover techniques impact on the mobile network performance. The performance measurement of the two handover approaches was based on network protocols including TCP/UDP with metrics such as throughput, packet loss and handover latency, and packet error rate (PER). Therefore, after setting up and configuring MIPv6 network, the next step was to simulate and acquire data in the output of the result file. These results were then kept and compared to the results from a new configure and simulated Fast handover MIPv6 to compare performances of the two networks. Hence, determine how well the handover latency improvement was in Fast MIPv6 as opposed to MIPv6 in terms network performance. Based the implementation of real-time application such as UDP/Video Stream, we were able to differentiate handover latencies and client-server network service performances from both MIPv6 and the proposed fast handover MIPv6 technologies. In addition, Video Stream application helped us measure and collect results on the network throughput, packet loss rate, and packet end-to-end delay. TCP protocol was used to simulate and collect data form FTP and HTTP network services considering metrics such as network throughput, and packet error rate. All results data were statistically recorded in a dataset and could be exported to other different file formats such as output packets capture (PCAP with Wireshark), output comma separated values (CSV with excel sheets) to perform the comparative data analysis. Finally, the analysis of data outputs was carried out at the end of the simulation in order to confirm that research specific objectives were met. Research Instruments This study projected a deep and understanding use of the network simulation tool, “OMNET++ release 5.2. (Objective Modular Network Testbed in C++)”. OMNET++ tool is a modular, component-based C++ simulation library and framework, focused primarily on building network simulators. It includes different simulation packages and can accommodate a range of other networking frameworks and simulation models that not only are logically implementable but can be brought and implemented in real-life experiments. OMNET++ is a modular, extensible, and component-based C++ simulation library and framework which offers an open source access to the platform under Academic Research version, primarily for building “network” simulators. It offers a discrete event simulation for modeling communication networks, queue networks, IT systems, multiprocessors, hardware architectures, distributed or parallel processes, and other systems. So, the term “network” here includes wireless and wired communication networks, queuing networks, on chip networks, etc. Adjacent domain-specific model frameworks such as support for wireless ad-hoc networks and sensor networks, Internet protocols, performance modeling, photonic networks, etc., are developed as independent projects (Liu & Yang, 2011). Furthermore, OMNET++ throughout all its releases offers an Eclipse-based IDE and a graphical runtime environment, as well as a host for other tools. It also embeds alternative programming languages such as Java, JSON, and C#, database integration, network emulation, real-time simulation, and other functions (Varga & Hornig, 2008; MobiWorld, 2010). OMNET++ has proved to be the optional solution between open-source, the high-priced commercial software tools like OPNET (Fluke Networks FN, 2006), and the research-oriented simulators like Network Simulator-2 (NS-2) (Issariyakul & Hossain, 2012). Additionally, OMNET++’s IPv6 packages are developed in INET Framework libraries, which is an OMNET++ extension and a protocol framework independently developed in 2004. INET IPv6 classes such as, the Routing Table version 6 (Network/IPv6/, and the Neighbor Discovery protocol (Network/ICMPv6/ IPv6NeighborDiscovery) have been modified to allow mobility management, neighbor discovery, and some other functionalities such as node auto-configuration in foreign networks, node’s ability to return home (Returning Home), packet routing and tunneling, fast handover, route optimization, etc. In the extensive version of MIPv6, “xMIPv6” was proposed introducing IPv6 extension header called Mobility Header, with relevant modifications related to packet tunneling to accommodate the tunneling functionality provided by IPv6Tunneling module with IPsec services, and route optimization with Return Routability security procedure (Yousaf, Bauer, & Wietfeld, 2008; Jauhari & Kistijantoro, 2015). Based on various modules and the simulation results output, we were able to amply extract results and analyze the outcome as suggest by the research objectives. Reliability Reliability is a measure of the degree to which a research instrument yields consistent results after repeated trials (Cooper et al., 2011). It refers to the consistency of the research and the extent to which the study can be replicated. Therefore, to ensure a considerable level of reliability, this study was subjected to experienced lecturers and technicians’ contribution in terms of correcting mistakes, guiding throughout the implementation process, advising on tools and technical approaches choice and adjusting the objectives to the process toward achieving the research goal. Additionally, we fact-checked the implementation process and configuration measures, as well as the results obtained with the rest of researchers’ works and literatures in order to verify the consistency of the configure and obtained data with the rest of the research world. Finally, we managed to use a multiple test-retest reliability technique in order to make sure the tools and mechanisms we used had the highest possibility of achieving the goal sought. Validity According to Brotherton (2008) validity is a research accuracy measurement approach that shows whether the research items measure what they are really designed to measure. Based on this imperative motive, to make this study more valid and credible, we stayed on track and updated with external trusted influences such as considering remarks and corrections from the supervisor and others experienced third parties making sure the instruments we used could provide adequate coverage of the study. Research Procedures In this research the procedure involved an experimental process that included network modeling and simulation of MIPv6 network in a client-server environment. The instrument used helped us manage all the required technological procedures such a mobile handover, handover delays, binding updates, route optimization, network QoS measurement, mobile host success test, AP configuration, and MN, HA, FA and CN set up. The experiment had then to integrate MIPv6 technology with client-server architecture, which led us to setting up the server as the current CN and ensure the client remains the MN. The second experiment involved a modification of handover-related parameters that led us to the implementation of faster handover mobile technology technique, i.e. Fast Handover MIPv6, which expectedly influenced the outcome of the extracted simulation results that we compared to the results obtained by the simulation of MIPv6 and drew conclusions based on the displayed performance differences. To be able to objectively test the server’s services and information exchange with its client, we managed to configure and test three types of traffics: HTTP, FTP, and Video Streaming services. Therefore, we used a network simulation tool called (OMNET++) that provided the platform that helped us develop MIPv6 through MIPv6 extension enabler framework provided in INET Framework to enable the logical simulation and testing of the proposed design in Section 4.3. This combination helped us create modules that were used to simulate MIPv6 technological and efficient requirements and enabled service sharing and access between the server and the mobile client. Based on the current successful MIPv6 implementation, we modified the system to integrate all necessary Fast Handover MIPv6 requirements in order to enhance the whole network performance. As illustrated in Figure 17, the implementation procedure consisted of multiple interdependent modules subset to the overall OMNET++ implementation. It gives a holistic idea of the technical requirements and procedure, and network implementation model that we considered to carry out the expected results as stated by the research objectives. Statistical methods as used in this research project, were involved in carrying out the study based on successive processes including planning, network design and topology, data collection, data analysis, drawing meaningful interpretation and reporting of the research findings. Therefore, as shown in figure 17 the whole procedure took a chronological succession of events involving network development and initialization (configuration). At the start, OMNET++ simulation environment required system preparation that involved installing Microsoft Visual Studio (OS choice was optional) that supported the C++ language that OMNET++ was developed in. The installation of OMNET ++ followed a step by step procedure as instructed in the installation guide (Varga & OpenSim, 2016). The installed OMNET++ platform then opened a possibility of installing INET Framework extension which includes various modules including xMIPv6 that we used to implement the project solutions. Two programming languages are available in xMIPv6 module: Network Definition and Network Initialization programming languages. NED language allows us to develop the MIPv6 network architecture and visualization source code. INI language, on the other hand, provided platform to develop a code that contains all the required configuration parameters responding to the fixed objectives as aligned in Figure 17 under omnetpp.ini branch. Apart from parameters related to application configuration, the rest of the parameters were involved in MN mobility management, hence contributed into the handover delay process. These parameters were used later to improve the handover process by obtaining a shorter latency while improving on the general network performance. After network development and configuration processes, we ran the codes based on defined data collection methods and recorded results in results datasets that we later analyzed using result analysis methods defined in section 3.5. Figure 17 illustrates of events and modular configurations of the solution implemented in order to achieve research objectives. The simulation library represents various components that we used to design (in NED files) and configure (in INI files) the network. All steps needed to implement the MIPv6 client-server network and its extensions (such as FMIPv6) as designed in OMNET++ are therefore demonstrated. ? Figure 17: Procedural Steps for MIPv6 Wireless Network Design Using OMNET++ (Source: Author) Data Analysis Methods To analyze the collected data in result output, we used statistical quantitative data analysis approach employing the first order statistics such as average, or mean values that were displayed in the output results. With the ability of the simulation model used, we started by analyzing the captured traffic traces to visualize the network attributes using Wireshark network analysis tool. It gave general perspective on the application level process as it mapped network events to modeling constructs with the exact time, source and destination IP addresses for every event that occurred during the simulation time. With this approach, after the preliminary analysis, the modeler may decide to disregard processes or events that appear less important for the study in question (Z. Ali & Bhaskar, 2016). The collected data, in the simulation output was analyzed using two different tools (Wireshark, MS Excel) and was based on the designed network topology and underlying technology specifications. The methods used were subsequent to the statistical approach chosen to be used in data analysis of the simulation results and were configured in the INI language configuration file. OMNET++/INET Framework allowed us to gather the results of the simulation in an analysis file. The simulation configuration was set to produce three types of output during simulation related to the way we defined them. OMNET++ uses a set of built-in types of data such as output scalars, output vectors, output histograms, which ate statistically recorded and can be exported to different other file formats such as output packets capture (PCAP with Wireshark), output comma separated values (CSV with excel sheets) (Sarkar & Membarth, 2015). Therefore, different types of data (depending on the technological measures and configuration parameters performed) are gathered ready for output display, plotting, and analysis. However, Handover delays that we measured by uniquely using the real-time application (video stream) was simulated with UDP protocol and analyzed with PCAP data output format in Wireshark which help target handover start and end times. We used the same protocol to measure video stream throughput and packet loss rate, and end-to-end delay in terms of QoS of the application. To measure HTTP and FTP QoS performance, we used TCP to measure Throughput, and Packet Error Rate metrics of the network. So, apart from handover delay analysis, the other metrics such as throughput, packet loss rate, end-to-end delay, as well as packet error rate were analyzed by exporting the recorded results scalars, vectors, or histograms to SCV files for comparative analysis between the MIPv6 and fast handover MIPv6 implementation in client-server architectures. Chapter Summary This chapter introduced and laid out the research methodology that led to the implementation of the proposed solution providing a path to the realization of the predefined objectives. The chapter proposed a research design that introduced all the technical and logical needs to design, setup and configure the platform needed to implement the proposed solution in Chapter 2. This process was followed by the chosen data collection methods that involved the use of both connectionless and connection-oriented transport protocols (UDP/TCP) that helped implement network services such as Video Stream, HTTP and FTP leading to the recording and extraction of results that we finally analyzed using statistical quantitative methods with results outputs in scalars, vectors, as well as histograms. The research procedures followed in this study to implement MIPv6 and Fast Handover MIPv6 was depicted in a diagram detailing different steps taken from the network design to implementation and results extraction as we used OMNET++ simulation tool and INET Framework in extension. Finally, all results recordings were configured as mean values. This helped us record more implicit values and ensured efficiency in memory management. We also singled out particular statistics such as Received Packet, Packet Sent, to be recorded considering their importance in the network QoS measurement. All these statistics were then kept in the “results” file in OMNET++ that we created in the INET’s xMIPv6 model directory. The results file recorded and generated three different types of files including scalars (.sca), vectors (.vec) and event log files (.elog), all saved in data analysis (.anf) file in OMET++. The obtained data in output was then exported to excel sheet files to be analyzed as a combination that established the statistical and quantitative differences in network performance and Quality of Service preferences between the proposed MIPv6 and Fast Handover MIPv6 technology in client-server architectures. CHAPTER 4 4.0 PROJECT IMPLEMENTATION Introduction The project aimed to propose a practical approach for mobile internet protocol implementation by employing a rather significant improvement of a client-server architecture with structural and modular expansion and improvement of critical network parameters that resulted in better network performance and improved QoS. In the academic world, to measure and verify performance of such mechanism, we can conduct either simulations or experiments, or even both. But in this case, we chose to employ the simulation solution. Moving forward, in the testing subsection, we describe the main assumption-based solutions in simulation models and simulation parameters used to evaluate performance of the proposed MIPv6 and the Fast MIPv6 solutions in client-server architectures. The statistics and numerical results obtained from simulations are presented and discussed in chapter six of this research study. This chapter expands and tackles the proposed concept via successive sections such as Analysis of the proposed concept, modeling and design of the solution, testing of the solution, as well as a brief chapter summary. Analysis Simulation Approach Simulation processes was performed for the realization of the project objectives highlighted in chapter 1 of this project and involved a particular simulation approach. There are three different simulation approaches in organizational sciences, which are: discrete event simulation (DES), system dynamics simulation (SD), as well as agent-based simulation (ABS) (Maidstone, 2012). This study focused on discrete event simulation approach that we used to implement the solution using OMNET++ simulation environment, which is a widely used discrete event simulation tool. As we explained in Chapter 2 of this study, the successful implementation, but also technology insufficiencies and underperformances of IPv4 as well as Mobile IPv4, has been a catalyst for the implementation of IPv6 as well as MIPv6 that does not only expand the network and mobile user reachability, but also introduces a set of variations and extensions that are used to improve handover latency and consequently, the overall network performance, specifically the Network’s QoS. Nonetheless, handover latency in MIPv6 has not yet provided the highly needed level of latency achievement that would allow the mobile network to reliably provide services to real-time applications such as Video Conference, etc. (Shih & Chen, 2010). The proposed solution focused on the implementation of MIPv6 that we improved into Fast Handover MIPv6 technology extension, all implemented in a client-server environment. The idea behind this protocol is to reduce the CoA acquisition and configuration latency with ability to perform these tasks in advance to the actual handover while the MN is still connected to its current AR. Therefore, this chapter focuses on modeling, design and implementation (simulation) of two different network concepts in client-server architecture. The proposed network is the simulation model of MIPv6 architecture that we improved and obtained a better performing Fast Handover MIPv6 in client-server network architecture. Discrete Event Simulation Approach Being considered as the widely used simulation technique in Operational Research, DES models a process by ways of a series of discrete events. This process means that simulated entities and network units ate thought of as moving between different states based on time (Maidstone, 2012). In definition, a DES is described by Albrecht (2010) who stipulated that: “Discrete event simulation utilizes a mathematical/logical model of a physical system that portrays state changes at precise points in simulated time. Both the nature of the state change and the time at which the change occurs, mandate precise description” Typically, DES model and technique is thought of as networks of queues and servers (network entities), and can therefore be used to discover if it would be advisable to add more entities (client, servers, etc…) to a service node, or if needed, restructure the system even more radically. For instance, in the situation depicted in Figure 18, the designer or modeler may find that after running the model, delays ensue at the final process activity. Adding more time to the system process may be the foreseen solution; however it is not the only solution. A better fixe may be to individually reduce activities time, or more radically, it may be the case that in the system some activities are less important and could less affect the system if completely removed. So, in DES modeling software, all approaches cited above can be simulated and in the event that test statistics are used, the approaches can be compared to each other (Fajar, Sarno, & Fauzan, 2018). Figure 18 illustrates the modular proportion of the implementation process of a discrete event simulation model. This model was then applied in this project as we used a simulation environment that is practically based on it. Figure 18: Discrete Event Simulation Model Process (Albrecht, 2010) The process definitions depicted above differ from each other on the basis that as event is a change in a particular object state, activity is seen as the state of that object over a period of time. On the other hand, object activity is described as the state of an object between two successive events describing consecutive state changes for that particular object; and last is the process described as the succession of states of the object over a time span (Lytchkina, 2009). However, we would precise that the three models can be compared to each other, and in some instances, they can be run complementarily. Table 5 presents the differences that can be established between the three simulation models comparing their conditions for use and main characteristics. Table 5 illustrates how DES essentially describes how in a system, events trigger changes in variables, while SD describes how variables cause change in each other over time. The last simulation approach is Agent-Based, illustrating a system where agents react to one another and the environment. Table 5: Comparisons between Simulation Approaches in Organizational Sciences (Source: Siebers, MacAl, Garnett, Buxton, & Pidd, 2010) Overview of Different Simulators Countless simulation tools such as OPNET, NS3, NS3, OMNET++, etc., have been developed for computer and telecommunication networks modeling and simulation. However, his research’s established objectives required a specific environment that could potentially respond to them as expected. Only OMNET++ and NS2 simulator clearly demonstrate ability to help implement the solution and advanced motives as needed (Hanumanthappa & Annaiah, 2006). Therefore, we briefly compared OMNET++ and NS2 simulation tools in Table 6 to establish a case on the choice we made to implement the solution. Table 6: Comparison between OMNET++ and NS2 Simulators (Source: Arvind, 2016; Bilal, Khan, & Othman, 2012) Features NS2 OMNET++ Language Supported Object Orient, written in C++/OTLC Object-Oriented, event driven simulator written in C++. Support Hierarchical Models Complex models are broken down in modules and sub-modules Models are flat. Not possible to create subnets or implement complex protocol as a composition of several independent units. GUI Support Poor Good Time Taken for Download and Installation Moderate Very less time. Easily available Time Taken to Learn Moderate Long Flexibility Designed as a TCP/IP network simulator. So, impossible to simulate other things other than packet-switching networks and protocols with it. Is flexible. Can simulate anything that can be mapped to components that communicate by passing messages. Platform Unix, Linux, Windows, Cygwin Linux, Window, MAC OS Availability of Analysis Tool It has analysis tool It has analysis tool Create Trace Files It creates trace files It creates trace files Scalability No scalable simulations possible. Cannot simulate large network topologies. Scalable simulation possible. Can simulate very large scale network topologies based on the computer memory space. Therefore, based on those characteristics, our focus aligned with OMNET++. The network was then developed using INET Framework 3.6.4 in OMNET++ 5.2 simulator, which allowed us to consider designing the whole network, define communication mechanisms and indispensable protocols on the simulation environment, and finally evaluate the results of the simulation. In the perspective of mobile technology implementation with IPv6 protocol variations and handovers, ranging from simple client-server architecture to a globally implemented mobile IPv6 client-server architecture, INET Framework in its current setup provides a range of network characteristics necessary to implement different mobility scenarios. However, a particular INET Framework sub-module known as extensible Mobile IPv6 (xMIPv6) was used in this research. Details of xMIPv6 are found in Yousaf et al. (2008) Simulation Environment The simulation was successfully carried out using OMNET++v5.2 and INET Framework v3.6.4. As mentioned in Chapter 2, OMNET++ is an extensible, modular, discrete-event, and object-oriented C++ network simulation environment that encompasses libraries and frameworks. As depicted by Figure 19 presents an architectural modular formation for simulation models in OMNET++. Models are gathered from components called modules, which can be connected to each other to form more enhanced and somehow complicated modules called compound modules (Xian, Shi, & Huang, 2008). However, OMNET++ by itself is not a simulator. However, its environment provides the tools and the infrastructure to perform particular simulations. On the other hand, the framework that allowed the complete implementation of this project, namely INET, is a framework and an open-source library that works in extension of OMNET++ simulation environment. It is provided the infrastructure, kernel and library by OMNET++ to facilitate projects designing and validating various protocols and applications. Some of the protocols incorporated in INET framework include models for the Internet Protocol stack (IP, TCP, UDP, SCTP, etc…), link layer protocols, and mobility models. They are all developed as application services, messages, or protocols (Jauhari & Kistijantoro, 2015). Figure 19 demonstrate the modular flexibility and scalability of OMNET++ environment. A Module can have one or multiple sub-modules, which can also have sub-modules on their part. The figure illustrates the openness of OMNET++ environment and solidifies its differences with other environment such as NS2. Figure 19: OMNET++ Modular Structure (Source: Varga & Hornig, 2008) Analysis of the Simulated Networks As mentioned in previous points, IPv6 and MIPv6 simulation in OMNET++ consist mainly of xMIPv6 simulation model for OMNET++ (Yousaf et al., 2008). From OMNET++ 3.2 up to OMNET++5.3 (latest release), xMIPv6 was considered as the main modeling domain of the INET environment framework. However, as developers are adding and improving modules, xMIPv6 has been attributed more distributed capabilities helping to improve not only MIPv6 network performance, but also improve performance of the basic IPv6 networks. At this point in time, INET framework contains several simulation models for wired and wireless connectivity in both standard and mobile environment. The simulation model consisted of the basic standard IPv6, an optimized MIPv6, and its enhancement in terms of Fast MIPv6 technologies in client-server architectures. The IPv6 client-server network is composed of one server and two client computers, one client being wired to the network via Ethernet link, and the other two clients connects to the network using wireless link. The network simulation of IPv6 client-server allowed the implementation and testing of transport protocols (TCP, UDP, Telnet and SCTP) within an IPv6 network with the assumption of having clients remotely connecting to the server. The results of this particular simulation included a continuous nodes’ Ping process through a Ping Application, TCP protocol analysis through services such as Telnet access, HTTP and FTP application, and Video Streaming services through UDP protocol. The analysis emphasized on network metrics like traffic Bit Error Rate (BER), packets throughput and end to end delay between the server and its clients, which in this case are influential factors of the network Quality of Service and helped us consider the differences between network performances of the non-mobile IPv6 network and mobile IPv6 network and its extension (Fast MIPv6 in this case). The mobile network simulation model took the aim at particularly a basic MIPv6 setup in client-server environment. MIPv6 network here is composed of two distinct and distant networks, which Home Network consisting of the HA and the yet to move MN, and the Foreign Network (Visited Network). The model also involves a remote node called Correspondent Node, and that set up as “Server” accessed over IPv6 Internet (Aristomenopoulos, Bouras, & Stamos, 2008). However, to achieve the ultimate goal of this research project, network performance comparisons have been established between the designed MIPv6 and the proposed Fast Handover MIPv6 networks by using OMNET++/INET Framework results analysis module to measure performances with different traffic source categories and services such as FTP, Video Streaming, as well as HTTP based on either TCP or UDP traffic. The results of the simulation comprise of However, using the simulation modules, we developed an extended capture of network statistics through the generation of network traffics that are displayed in a Wireshark file, which to some extent helps to monitor traffic and manage network performance. We used TCP protocol for this regard. All protocol implementations used for this project were also described and satisfactorily explained. Modeling and Design The process included model architecture design limited at one client and one server interacting with the rest of MIPv6 architecture components. The simulation model was based on OMNET++ simulation tool and INET Framework. Through the architecture developed, we were able to implement the project solutions ranging from an initial MIPv6 network implementation from which we modified handover-related parameters to fasten the handover process based on technological requirements of Fast Handover Mobile IPv6. The protocols are all implemented with potential adoption at personal, corporate and scale. OMNET++ Integrated Development Environment As mentioned earlier, OMNET++ IDE is an Eclipse based platform with C++ programming language, platform editors and views. Using OMNET++, models can be created and configured using languages such as NED and INI programming languages, kept in respective files. Apart from that, the environment allows batch executions, and simulation results analysis through statistical estimates and calculations using Scalars, Vectors and Histograms for pictorial and graphical displays of the simulation results (Varga & Hornig, 2008). The platform encompasses a considerable number of modular components (Mayer & Gamer, 2008), where the main ones include: NED Editor: helps create and generate NED files that can be edited in both graphical and source code editing mode. Through the graphical user interface, the chosen topology can be built by creating compound modules, sub-modules, channels and their characteristics, and other component types. Visual and non-visual properties can be edited both by direct source code commands, and from the Properties View. The text mode used is based on Eclipse C++ editing and offers context-aware completion of keywords and module types, automatic indentation, syntax highlighting and source code validation as the user is typing. INI File Editor: used for configuration and validation of network parameters of the simulation model (topology) both in graphical and source code editing mode. The editor examines the NED declarations and configurations and can provide automatic generation of most useful NED parameters. The user will be prompted to provide the desired values. It offers a ‘Problem View’ that displays all the errors, warning. It also offers info messages and documentation. Simulation Launcher: allows the simulation to be directly executed form the IDE as normal C/C++ program, as standalone applications (under Tkenv or Cmdenv modes) or as batches of simulations where every run differs in defined module parameters. The Tkenv environment uses automatic animation method which animates the flow of messages and reflects the state changes of the nodes, module output window method which displays textual debugging or tracing information, generated by the individual modules or module groups, and object inspector methods displays the state or the content of every individual object. Scave and SQLite (For Result Analysis): OMNET++ tools, used to visualize simulation results saved into vector, scalar or histogram files. They allow the user to group the data into various datasets and perform various processing, filtering and plotting operations. Various chart and graph types are available and can be plotted based on the created dataset. Simulation Modeling The simulation of IPv6, MIPv6 and FMIPv6 client-server architectures in OMNET++ consists essentially of xMIPv6 Simulation Model for OMNET++ (Yousaf et al., 2008). This model is developed in INET/Framework and allows the implementation of MIPv6 protocol. It also provides extension measures for MIPv6 in its parameter configuration file program with respect to the requirement of the implemented technology. Therefore, since INET Framework releases do not provide a direct Fast Handover implementation, MIPv6 parameters modification approach with respect to Fast Handover technology requirements in handover process set up, was used to respond to the project objectives in Chapter 2. Different parameters in MIPv6 implementation are presented in Table 7. Network Topology and Simulation Setup To evaluate the proposed approach, three network topologies were simulated. From a standard client-server MIPv6 network, we built up an optimized and fast handover MIPv6. Some of the simulation parameters were used in more than one scenario, as will explicitly be described in the Section 4.5. The most significant and impacting values related to the simulation environment, and to the handover process improvement are summarized in the Table 7. In fact, the topology of the modeled mobile network solutions has two wireless ARs and a gateway connecting to the Internet via an IPv6 router. Based on the C++ and NED/INI coded algorithms for router solicitation advertisement, and IPv6 flat IPv6 configuration mode capabilities of IPv6 routers, the gateways and network routers are able to automatically provide network and subnetwork IP addresses MIPv6-supported nodes that can be found in the network coverage area (Mrugalski et al., 2010). Furthermore, this research focused on two different aspects of the mobile network topology that we explained in Chapter 2, which included handover latency and route optimization parameters. Reason being that these two aspects favor the aim of the project by enhancing the QoS in terms of network performance of MIPv6 achieving a faster and reliable handover mobility (FMIPv6) for three particular services (FTP, HTTP and Video Streaming) based on both UDP and TCP protocols in client-server environment. Additionally, based on security risks introduced by both of the aspects (Handover and RO), we proposed and implemented required security measures that protected communications between the client and the HA, and the client and the server (or CN). MIPv6 Network Topology The project ojectives expands expectations of IPv6 implementation capabilities to a mobile network, allowing network entities to communicate and exhange information in a mobile fashion. In chapter 2, there is a well an thorough explanation of MIPv6 and its coming to life. This topology encopasses some similarities, but also some major and decisive differences with MIPv4. Therefore, based on the differences in parameters recorded in the literature, this topology implements mobile network parameters such as route optimization, route solicitation and advertisement, layer 2 and layer 3 handovers, and others. As we explain in the next section, the model and modular implementation of the network exhubit the working configuration and expectations of evry entitiy on the proposed topology for MIPv6 client-server network in Figure 20. Figure 20 illustrates the process model the proposed solution. It describes all necessary modules and processes that lead us to the implementation and improvement of MIPv6 handover, hence improvement of the network QoS. With one Home Network, one Foreign Network, one client and one server, it illustrates Handover process, securtity through IPSec (using packet tunneling method) protocol and Route Optimization process. Figure 20: Proposed Topology of MIPv6 Network Simulation Model in Client-Server Environment (Source: Author) MIPv6 topology included one HN, one FN, one client and one server for the purpose of implementing and testing working network architecture. As the client moves from it HN to the FN, it establishes an IPv6 communication tunnel with the HA to inform of its attachment to a different network (FN) by sending BU message carrying its CoA. At this point, for the packets to be routed between the server and the client, the HA is used to relay the two entities. But after a certain period of time, the client send another BU message to the Server to establish a direct communication and start exchanging information without relying on the HA. The decision creates the route optimization process with extra security measure dependent of mobility extension headers in IPv6 that in carried out by RR procedure. So, until the client leaves the FN, it will be using the optimized route. Additionally, this topology and its logical implementation cover the final and most significant milestone of all the scientific and literature contributions that this project has unfolded. In actual or real-life MIPv6 network implementation, to be able to introduce and test the handover process or its improvement i.e. enhance network performance and QoS for the users (clients), the topology considers not only one FA router, but as many IPv6 routers as possible supporting handovers and allowing network scalability. Different modules including “IPv6 Interface Table”, “IPv6 Network Layer”, “IPv6 Link Layer”, etc., contribute on the network mobility management and can be subject to modifications related to the handover latency, hence can considerable enhance network performance in terms QoS (Yousaf et al., 2008). All these modules were implemented and helped achieve the end-goal of the research based on protocols such as UDP, TCP, and Telnet. So, these modules are well involved in the overall improvement of MIPv6 in order to obtain a Fast Handover MIPv6, then ameliorating the overall QoS. The realistic impact of the modifications starts taking effect from the very first FA or AR that the MN visits in the process. Simulation Model Accuracy The accuracy of the simulation model is determined by a set of characteristics, topology and technical expectations. The most important driving factor for network performance evaluation in the proposed solution is the handover delay. After simulation and test, the results show the appreciable accuracy and consistent behavior of the simulation model in terms of the handover delay for both MIPv6 and Fast Handover Mobile IPv6 as compared to the handover delay measured in the real test bed by Yousaf et al. (2008). The accuracy of the simulation framework (INET) has been ensured by modeling messages based on the actual message formats, a careful selection and implementation of message/event timers and timeouts and by careful and close conformance and strict adherence to the relevant standards and norms as of the IETF best practices. Therefore, the accuracy of the xMIPv6 simulation model gives a ton of confidence and makes it ideal for building other MIPv6 related mobility management protocols (such as FMIPv6) with a considerable degree of confidence in its accuracy. Furthermore, to realize the expected research objectives, we considered that the effect of handover delay on higher layer protocols such as UDP and TCP needed to be compared for both simulation models (MIPv6 and FMIPv6). Apart from Yousaf et al. (2008), we made other comparison with previous simulation project such as the one done by Pieterse, Wolhuter and Mitton (2013) to conform our technology adoption and parametrization results with their findings. Simulation Implementation MIPv6 Technology in Client-Server Architecture Mobile IPv6 is the essential object of this study. The perspective of implementing the technology encouraged us into simulating a well-designed and function IPv6 client-server above. So, in this chapter, we describe the simulation models used to evaluate the performance of the MIPv6 as well as FMIPv6 technologies in client-server environments. The numerical results from the simulations are presented and discussed in chapter 5 of this project. As designed earlier, the client-server network setup with MIPv6 technology can be seen in Figure 21. As standardized by the IETF, the MIPv6 network, as designed and developed in this project consists essentially of one HA (router), one FA (router), one client, and one server. We used a scalable programming methodology should allow, whenever necessary, to increase the number of clients and/or servers to the needs. We could possibly proceed with implementation that takes aim at the handover between two FN applying the same configurations and modules as used in this project, then end up with simulation performance of the same proportion as the one that considers only the HN and the next FN in OMNET++ (León, Hernández-Serrano, & Soriano, 2010 ; Branco, Pinto, & Pigatto, 2017) . The architecture has eight nodes with 5 IP nodes and 3 non-IP nodes in total in the simulation area designed in Figure 21. The nodes consist of one MN labeled as client represented by a laptop icon. Three routers represented by router icons. One router on the left labeled as “Home_Agent” is the Home Agent that connects the client to the network and internet through an access point “AP_Home” and one router on the right labeled as “FA” is the Foreign Agent that through binding and association processes connects the client to the network and internet through an access point “AP_FA”. We also have a hub that connects the Correspondent Node to the Gateway Ipv6 router. Lastly, the Correspondent Node represented by the desktop image. To visualize network activities and signaling we used the “visualizer” module in OMNET++ that allowed the display of signal power coverage range of wireless nodes, data packets exchange as well as the MN trajectory, as it is explained in Section 4.5. Figure 21 pictures the architectural implementation of MIPv6 and the structural requirements to implement the proposed solution. As described in the topology sub-section, one client, one server and one AR were part of the architecture for implementation and testing. Results coming out of this architecture allowed us to establish the required differences between MIPv6 and fast handover MIPv6, after multiple required runs and test and after extracting the underlying results. Figure 21: Mobile IPv6 Client-Server Network Model (source: Author) Home Agent The HA is similar to a standard IPv6 router and was developed from the Router6 model in INET. NED Parameters for the HA include “isMobileNode” and “isHomeAgent”. In case “isMobileNode” parameter is set to “True”, the router6 acts as a MN, and if the “isHomeAgent” parameter is set to true, the router acts as the HA in the network architecture. However, both parameters cannot be set to true. To hold the binding data between the HA and the MNs, the HA is set to carry a single instance of the “BindingCache” module. Client As the project title suggests, the host node (client) is responsible of mobility and handover decision. This implies that the MN module bears all the mobility capabilities and handover decision sub-modules to accomplish its mission while moving across networks. The client, which is a mobile wireless Mobile Node in this case, is a wireless compound node and an IPv6 and MIPv6 enabled node. The MN has a WLAN management interface of type Ieee80211NicSTA and a mobility module that handles the mobility of the MN between different networks. In case of the MN, the isMobileNode parameter always has to be set to “true”, while both the isHomeAgent and isRouter parameters are set to false. The MN consists of a BindingUpdateList that hold the bindings information made between the MN, CN and HA. In Figure 22 the MN (WirelessHost) module, Binding List (buList) and Binding cache submodules can be seen in “xmipv6support” module. The “mipv6support” module is connected to a modified IPv6 module. Thus, the IPv6 module is modified in order to receive, recognize and forward mobility packets to and from “mipv6support” module (Yousaf et al., 2008). Figure 22 details the structural and modular constitution of the MN highlighting connections and interdependences between modules and sub-modules that are involved in the network mobility and handover decision taken by the MN. The MN has to be defined as supporting MIPv6 for it to be recognized by other network entities. The most important component of the network layer module is the “NeighborDiscovery” sub-module since it manages all the mobility and network abilities needed by the MN to perform handovers. Figure 22: Mobile Node Modular Stucture Server The server being the static wired correspondent node this case, is an IPv6 host with MIPv6 support and contains a Binding Cache which gets updated with every BU received from the MN. All the mobility parameters mentioned in the Correspondent Node are set to false. The Correspondent Node (server) thus includes a “BindingCache” messaging submodule that is set to capture all the binding information enabling Route Optimization between the MN and the CN (Le & Chang, 2010). For the purpose of testing the technology and fulfilling implementation objectives, we limited the network to one client and one server with the prospect of scaling their numbers as it pleases. This network is made of a combination of wireless media between the client and both the Home and Foreign Network APs (AP_Home, and AP_FA), while the rest of the architecture uses Ethernet links. The network IP address configuration included IPv6 flat network configurator, where we chose to adopt “stateless auto-configuration”. Flat configurator mode is processed based on the IPv6 Network Layer module’s source code, which automatically attributed IPv6 address to network entities and to nodes visiting the network form outside. For wireless medium to take effect, it need an appropriate configuration that requires a radio medium manager that laid the ground work for the mobile node and the APs to exchange association information and data traffic. The server is connected to the IPv6 internet through the “IPv6Router” via Ethernet medium interface and so are the HA and FA. From the perspective of the server network, we prepositioned a non-IP device (in this case a Hub) between the server and the network gateway (IPv6Router). Handover Latency Analysis A client is unable to receive IP packets on its first association point with the AP until the handover process finishes. The time-period between the reception (and transmission) of the last IP packet by the MN through the old connection (with the previous AP) in HN and the first packet through the new connection (with the new AP) in FN is the handover latency. Handover latency is affected by a process made of several components (Amin, Bakar, Abdullah, & Khokhar, 2011): Link Layer Establishment Delay (DL2): is the required time by the network nodes physical interface to establish a new association with the visiting client. This is the L2 handover between AP linked to different access routers. Movement Detection (DRD): This is the time required for the MN to receive wireless beacons from the new AP, after disconnecting from the old AP or the old access network. Duplicate Address Detection (DDAD): handled by the network router. It indicates time required to recognize the uniqueness of the mobile IPv6address within the home network. Binding Update/Registration Delay (DREG): is the time elapsed between the sending of the Binding Update from the MN to the HA and the transmission/reception of the first packet through the new AR (FA). Therefore, it becomes possible to mathematically find the overall handover process using the delay components identified above as: DMIPv6 =DL2 + DRD + DDAD + DREG According to Amin et. al (2011), we can still break the delays down to: DMIPv6 = (TPRB + TAUTH + TRASS) + (TRSOL + TRADV) + DDAD + (THBU +THBA + 2THOTI + 2THOT + TCBU + TCBA) Where: At L2 we have: Probe (TPRB), Authentication (TAUTH), and Re-Association (TRASS) delays. At Route Discovery, we have: Router Solicitation (TRSOL) and Router Advertisement TRADV delays. Finally, BU and Back delays with HA, 2THOTI, 2THOT: HoTi and HoT process and TCBU, TCBA: BU and BAck with CN. Testing Overview IETF has standardized IPv6 protocol and its extension techniques that allowed researchers in network management and innovation to focus on subjects patterning to SDN (Software Defined Networks) development as a breakthrough in providing proof of concept via implementation of network projects through simulation environments (C. Perkins et al., 2011 ; Waehlisch, Koodli, Fairhurst, & Liu, 2014). Therefore, a simulation environment is much need to test the solutions proposed in this project’s objectives before taking it to the proof-of-concept level that involves network or prototype design, equipment and entity parameters, defining tests, and examining test results. These factors are then used to clearly respond to the objectives defined in this project. The testing procedure for this research was conducted in OMNET++ using IINET Framework and involved an initial implementation of IPv6 protocol through a simple Client-Server architecture which lead to the understanding and implementation of mobile IPv6 and its extension protocols based on xMIPv6 that offers a user-friendly programming environment and different parameters that we modified in order to achieve the research goals. Therefore, all the general parameters were used by every module, every protocol, or application created and defined set of parameters for a particular NED network. To measure the network QoS, we considered a predefined QoS parameter that is based on the network’s MAC protocol in xMIPv6 (Yousaf, Müller, & Wietfeld, 2010). However, in the network definition, we considered an Ethernet line of 100 Mbps, and if necessary to use fiber optics line with 1Gbps of bitrates. To test the TCP network, INET Framework provides a library that contains TCP application definition for both the client and the server within the network. Therefore, for this proposed network, on the client side, we used Telnet Application to test remote access of the server by the client using just the server’s IPv6 address or using the name of the node that was resolved by the DNS characteristics defined in “IPv6Address” Module in INET Framework. As explained in Section 4.4, in INET framework, IP addresses are defined using the “IPv6FlatConfigurator” sub-module. However, this configuration can be personalized by creating an INI (network configuration) file with static addresses in the INI environment, but since we opted for IPv6 Address auto-configuration mode for this research, using the flat configurator module was the obvious and most relevant choice. Flat configurator module implement a global address resolution process that attribute automatically attributes IPv6 addresses to nodes based on their notation (name given to node) (OMNeT++, 2012). Testing the Proposed MIPv6 Client-Server Network To test the network, multiple categories of mobile network parameters, multiple assumptions were taken to conform and prove the concept with regard to real-life implementation requirements. The network simulation is performed on a rectangular flat terrain in of 1111.8151 meters x 891.405 meters. For a HA and one FA, the network has only two APs located in different location respective to that of the HA and FA routers as shown in Figure 21. The simulation was performed based on different assumptions including the client (user) behavior, network characteristics, as well as radio transmission characteristics. Client Behavior As the network design suggests, only one client was considered using the network resources as a single user, represented by a sing MN. The initial location of the client on the simulated surface is within the Home Network at the point (218, 125), as it moves to any potential Foreign Network on its path. The client is moving at a speed of V=10m/s from West to East and East to West as it follows a rectangular path. It is assumed that the client that is using a MN is engaged in sending files, requesting HTTP pages, as well as watching a video. The simulated video traffic stream starts at t0=5 seconds with a size of 20 Mega Bytes for tests of separate 8 bitrates including 6, 9, 12, 18, 24, 36, 48, and 54 Mbps run separately. As of HTTP and FTP traffics, the simulation of the services starts at t0=5 and t0=3 seconds respectively, with traffic of 350 Bytes upload and 1 Mega Byte download and a file size of 12 Mega Bytes. The client movement pattern can be assumed as a typical movement for a pedestrian as used in previous publications such as (Yousaf et al., 2008) and (Jakubiak & Koucheryavy, 2007). Additionally, if the client enters a new network (new AP), it has to stay within this network for as long as the handover process lasts. Otherwise, the possible handover will never reach its completion and the client will then lose its internet connectivity. Furthermore, data traffic parameters and types are also arbitrary selected to be able to assess the network performance. Network Characteristics The transmission range (200 meters in this case) of the client node is determined by the power of the node transmitters, the receiver sensitivity, the thermal noise, the SNIR threshold, as well as the pass loss coefficient. This process elaborates on how to calculate the wireless transmission range (Matzer & Levy, 2004): The received (Pr) can be expressed as: Pr(dbm)= 10log10 (Pt)+ GT+ GR+ 20log10 (?/4?d0)+10? log?(d0/d)+ Pshadow loss + Pfading Where Pt = Power Transmitted (in milliwatts), GT=Power Gain of the transmitting antenna (in dB), GR= Power Gain of the receiving antenna (in dB), d0= Reference distance (at which the path loss inherits free space path loss), ? = Wavelength (in meter), ? = Path loss exponent (ranges between 2 to 5), PShadow loss = Power due to Shadowing (in dB), and PFading = Power due to Fading (in dB) Then, based on the receiver sensitivity we calculate the MN transmission range. In our case, Gt and Gr are zero and so are shadow power and fading power. So, the variable “d” can be solved with the defined “Pr” as -82dBm in the simulation parameters to get the range. More of these parameters are detailed in the radio transmission characteristics assumption. Both transmission rate and transmission are set to be the same as used by Yousaf et al. (2008) as the routers transmit at an Ethernet transmission rate of 100 Mbps. This value can arbitrary be selected it is sufficient for network traffic demands and does not incur any packet losses. The MN is able to participate in the network as soon as it is in the antenna’s transmission range. Since most of the simulation parameters in this project are in agreement with Yousaf et al. (2008) and Pieterse et al. (2013), our simulation results should relatively be comparable to their results. Radio Transmission Characteristics Transmitters’ power was set to 2.0mW. It specifies the total amount of power of radio frequency energy output by the transmitter antenna. Receiver sensitivity that we set to -82mW evaluates radio frequency signal by indicating how faint or faded the signal can be successfully received or accepted by the receiver antenna (AP). The simulation included considering other parameters such as the path loss (FSPL) coefficient that was set to 2, and that defines the loss of signal encountered due to obstacles such as inside a building, densely populated areas, etc. over distance. The SNIR threshold equals to 4dB. SNIR threshold is the Signal to Noise and Interference Ratio, also known as SNR threshold (Haraty, Mohammadi, & Mehbodnia, 2014). These setting also align with Yousaf et al. (2008). Configuration of General Parameters As we stated in chapter two of this research project, MIPv6 is a node-based mobility technology that allows the mobile node to make network and mobility decisions as it moves across different networks. This is basically the first lines of code within the “.ini” simulation parameter file as shown in table 7. The most important and impacting parameters are those patterning to the mobility of the node. They impact the simulation behavior and results on record. Therefore, according to the parameters defined in the code, the client has been assigned with a rectangular mobility: “RectangleMobility” model. This practically means that client 0 moves in a rectangular fashion from the home access network to the foreign access network as it may decide to return home or move to the net access network. So, we defined (195m, 125m) and (945m, 180m) as specified in the INI file source code. The MAC layer of the network uses IEEE802.11b with 54Mbps of bitrate and a mobility speed of 10 meters per second which is approximately 36 km/h. Neighbor discovery module is one of the most influential metrics of the handover process and is determined by the interval in the range between the minimum and maximum router advertisement intervals which can relatively be modified and which we configured and attributed the standardized values of minIntervalBetweenRAs, which is equal to 1 second and maxIntervalBetweenRAs equals 3 seconds in MIPv6 implementation. These values can still be reference in (Yousaf et al., 2008). In the configuration of wireless settings, wireless channel scan processes were activated with a probe delay of 0.1 second with minimum and maximum channel scanning time set to minChannelTime= 0.15 second and maxChannelTime= 0.3 second, respectively as default values in OMNET++/INET Framework environment, respective of IETF RFC 6275. Both authentication and association time-outs were set to 5 seconds. The simulation time was set to 120 seconds that was extended to the simulation of services (FTP, HTTP, and Video Streaming). Simulation parameters are listed in table 7. Table 7: MIPv6 Initial Network Configuration Parameters (Source: Author) Attributes Values *.total_mn 1 *.total_cn 1 **.neighbourDiscovery.minIntervalBetweenRAs 1.0s **.neighbourDiscovery.maxIntervalBetweenRAs 3.0s **.radio.transmitter.power 2.0mW *.radioMedium.mediumLimitCache.minReceptionPower -82dBm *.radioMedium.mediumLimitCache.minInterferencePower -82dBm **.client0.**.mgmt.accessPointAddress “10:AA:00:00:00:01” **.wlan*.mgmt.numAuthSteps 4 **.wlan*.bitrate 54 Mbps **.mac.qosStation TRUE **.rtsThreshold 1000B **.AP*.wlan*.mgmt.beaconInterval 0.1s **.AP_Home.wlan*.mgmt.ssid “HOME-NET” **.wlan*.agent.activeScan TRUE **.wlan*.agent.default_ssid “” **.wlan*.agent.channelsToScan “1 2” **.wlan*.agent.probeDelay 0.1s **.wlan*.agent.minChannelTime 0.15s **.wlan*.agent.maxChannelTime 0.3s **.wlan*.agent.authenticationTimeout 5s **.wlan*.agent.associationTimeout 5s **.mobility.constraintAreaMinZ 0m **.mobility.constraintAreaMaxZ 0m **.client0.mobilityType “RectangleMobility” **.client0.mobility.constraintAreaMinX 195m **.client0.mobility.constraintAreaMinY 125m **.client0.mobility.constraintAreaMaxX 945m **.client0.mobility.constraintAreaMaxY 180m **.client0.mobility.speed 10mps Furthermore, network traffic has been generated between the client and the server with three types of services HTTP, FTP and Video. Every service is defined with network characteristics and traffic model that runs between the client and the server as listed in Tables 8, 9 and 10. Table 8: Traffic Model for Video Streaming Application (Source: Author) Application Traffic Model Value Simulation Time 120 Seconds Start Time (Information Exchange) 3 Seconds (after launching the simulation) Server Port 3088 Video Size 25 Megabits Send Message Interval 10 Milliseconds Packet Length 1000 Bytes Table 9: Traffic Model for Web Application (Source: Author) Application Traffic Model Value Simulation Time 120 Seconds Start Time 4 Seconds Server Port 80 Number of Request Per Session 1 Page Request maximum Size Truncated in 350 Bytes and 20 Bytes Table 10: Traffic Model for FTP Application (Source: Author) Application Traffic Model Value Simulation Time 120 Seconds Service Start Time 3.5 Seconds Server Port 21 File Size 20 Mega Bytes Fast Handover Mobile IPv6 Implementation The overall concern in this project is as to how to avoid the handover latency and enable real-time applications such as video stream to be transmitted between the client and the server while reducing of the off-time that ultimately converts into loss of packets and low quality of service. Based on the technology standards and implementation procedures as standardized by IETF in RFC 5268 on mobile IPv6 fast handovers for 802.11 networks, the overall implementation could be carried out using test bed implementation, or a simulation (Koodli, 2010). As of this research project, we implemented the solution using the discrete event simulator, OMNET++. For the FMIPv6 implementation which is an extension of MIPv6, the handover duration is shortened from the initial Fast Binding update that the client (MN) initiate to connect to the FN, hence to access HA resources. At a higher degree, FMIPv6 and its functionalities relies on L2 triggers, hence on L2 handover, in order to execute L3 process in a faster way (Vassiliou & Zinonos, 2010). The aim of the technology is to allow an MN to quickly configure its NCoA before it moves and connects to a new network, and to use the NCoA immediately upon connecting to the new network (FA). In other word, the FMIPv6 protocol aims to eliminate the existing latency that occurs during the BU procedure that is initiated by the MN, by providing a bi-directional tunnel between the old and new networks while the BU procedures are being performed between the MN (client), the HA and the server. However, OMNET++ does not provision for a direct implementation of FMIPv6in its current releases. Nevertheless, the simulator contains almost all the parameters that when implementing FMIPv6 through test bed or in other simulators, they affect the network performance as much as modifying them right in the .ini file for parameter configuration in OMNET++ provided by xMIPv6. After all, the aim is to create a fast detection movement at layer three with L2 involvement, hence diminishing the time it takes for the client to be granted internet access by the FA network. As explained earlier, MIPv6 implementation introduced a handover latency essentially caused by three major components: movement detection, duplicate address detection, as well as delay interval for a router to reply with a router advertisement upon a receipt of a router solicitation. Thus, the implementation of FMIPv6 came into substantially reduce or completely eliminate one or all of these components in order to reduce the overall handover for a better network performance with better QoS (S.-J. Yang & Chen, 2010). The parameters modified in our experimental evaluation are all part of the general configuration of the simulation and services, as they are designed to impact on different services exchanged between the client and the server. The “Beacon Interval” is the time between successive beacons signals sent to the MN by the APs. HA and the FA Min and Max RouterAdvInterval signal refer to the minimum and maximum values of the range specified for randomly sending unsolicited router advertisements to the MN. The overall delay, hence the network performance, also depends on the DAD procedure that is counted twice as the networks’ routers (HA and FA) verifies the uniqueness of the Client’s IP address. However, independently of the consistency of the simulated network parameters, due to the differences in access networks, hardware, network traffic, and implementation versions, there cannot be a single value for the overall MIPv6 handover delay. Referring to some real implementation of the technology in the literature, we found related values for MIPv6 handovers that vary from 3.6 seconds (Vassiliou & Zinonos, 2010) and between 3 and 5 seconds (Y. Kim, Kwon, Bae, & Suh, 2005), which as précised in chapter two is intolerable for real-time applications. Hence, to archive FMIPv6 handover MIPv6 handover latency can be shortened by direct manipulation of a number of parameters in the MIPv6 implementation. FMIPv6 is designed to eliminate the delays associated with the MN’s movement detection and CoA testing (DAD), and the overall time introduced by the CoA configuration process. It is designed to allow MN to anticipate movement in its network layer mobility. After all, link layer triggers are required for the anticipation and whole handover initiation. The triggers are delivered to network layer modules as specific events that report changes in respect to the link and physical layers conditions between the MN and the APs. FMIPv6 solution manages to reduce BU/Registration delay but in our research, we focused on the other three delay components including DL2, DRD, and DDAD. So it can be concluded that handover procedure consists of just Link Layer or L2, and Network Layer or L3 (Çetin & Çetin, 2015). Modifying L2 Delays L2 means the MN is establishing a connection with and new network’s AP. For the purpose of improving the handover delay, L2 delay is carried out in some FMIPv6 methods which eliminate handover latency in some cases, or can limit it to L2 handover latency (Çetin & Çetin, 2015). In OMNET implementation, these methods contain L2 triggers introduced in xMIPv6 source code in order to perform MIPv6 extension (FMIPv6 in this case) in INET if required. As L2 delays are highly dependent on the physical medium, they therefore exhibit great variations. Furthermore, since the probing, or scanning delay is the most prevalent during an L2 handover, we believe that it merits special attention as affirmed in (Johnny Lai et al., 2006). In fact, on its own, the probe delay maintains 90% of the total L2 handover delay (Fu & Atiquzzaman, 2005; Zaidi, Bhar, Ouni, & Tourki, 2011). In the propose solution, we shortened the wireless beacon interval, probe delay and scanning time to values bellow 0.1 s (or 100ms), respectively, in an effort to reduce DL2. Modifying Router Advertisements The metrics that helps the MN identify that it has changed subnets are Router Solicitations (RSol) and Router Advertisements (RAs). They provide the MN with the necessary information for the creation of the NCoA to establish communication with its HA and Server. In MIPv6 implementation, routers send unsolicited RAs messages to advertise its presence to other nodes. For better performance, networks require faster movement detection by modifying RAs values (MaxRtrAdvInterval and MinRtrAdvInterval). Therefore, we should necessarily allow a quicker sending of RAs more frequently than the 3 seconds establish in the standard MIPv6. In traditional IPv6, the values for RA intervals are usually in the order of 3 to 5 seconds, but for Mobile IPv6 these values need to be significantly lower (Vassiliou & Zinonos, 2010). In this project, we reduced RA intervals to 0.01-0.05 seconds in an effort to deduce their effect on DRD delay. Modifying Duplicate Address Detection This metric is one of the most effective in affecting handover delay since the MN has to bare a unique IP address while travelling across networks. Indeed, when the MN reaches and discovers a new access network and a new router and creates a new CoA by guarantying its uniqueness. In fact, it tries to find out if the given CoA address is unique or not in use by any other node in the network. Handled by the “NeighborDiscovery” module in INET, DAD offers some room for QoS improvement. In INET, the value emitted by this metric is of 1second. To manage the fast handover implementation, we modified the emitted value in the source code by attributing 0 second to the DAD as noted in RFC 5568 and RFC 4862 (Koodli, 2009b). Chapter Summary This chapter presented a thorough analysis of all required methods and tools useful for a swift implementation of MIPv6 and then thoroughly presented the analysis of all required elements that helped improve the network performance in implementing Fast Handover MIPv6 in client-server architecture using OMNET++. Multiple simulators such as OPNET, NS2, OMNET++, etc. could be considered for the implementation of the proposed solution but OMNET++ presented more prominence in terms of support for hierarchical models, good Graphic User Interface, time taken to lean, its flexibility and scalability, and so on. The chapter later presented modeling, design and configuration of the proposed MIPv6 client-server architecture designed in OMNET++. To realize the implementation in OMNET++, INET Framework was added in extension which needed xMIPv6 models. The architecture contained one mobile client and one server with three routers (one for HA, one for FA and one as IPv6 Internet Gateway) that connected the client to the internet and ultimately to the server through WLAN interfaces using APs. Three services were implemented and tested with the use of TCP and UDP transport protocol. The process of providing a faster MIPv6 required radical modifications within some modules and changes in network configuration parameters values, where the most influencing values pertained to the handover delay which encompasses multiple factors (L2 Delay, Route Discovery, Duplicate Address Detection, etc.). All the services (FTP, HTTP and Video Stream) were implemented with the same traffic model irrespective of what handover technology we implemented. That was to allow the accuracy and verification of the implemented technologies. The next chapter will present results and findings from the empirical analysis and design of the proposed solution. ? CHAPTER 5 5.0 RESULTS AND FINDINGS Introduction This chapter presents the results of the simulation and tests of MIPv6 and Fast Handover MIPv6 technology solutions in client-server architecture driven by the research objectives. Multiple tests were conducted differently and multiple times to ensure the accuracy of the obtained results that are saved in the result files created in OMNET++ simulation environment. Data collection was based on one client communicating with one server using video stream application through UDP protocol, and HTTP and FTP applications through TCP protocol for both MIPv6 and Fast Handover MIPv6 in extension. While HTTP services were simulated using a single data rate (54Mbps-bitrate) as defined in subsection 4.5.6, FTP and Video Stream services were run multiple times using data rate variations in terms of bitrates values (6 Mbps, 9 Mbps, 12 Mbps, 18 Mbps, 24 Mbps, 36 Mbps, 48 Mbps, and 54 Mbps) to test and compare the recorded results in tables and plotted graphs. Considered as the most important application of this study in terms of handover delay management, Video Stream’s QoS performance measurements displayed on the graphs, demonstrated possibilities of improving handover latency conditions in MIPv6 implementation by reducing the delay, therefore reducing network packet loss. With this same application we also measured the network throughput, and packet end-to-end delay metrics. On the other hand, using TCP (with HTTP, and FTP) protocol, we measured TCP/Throughput and Packet Error Rate metrics for MIPv6 and Fast Handover MIPv6 implementation in client-server architecture to measure and compare performance differences. Chapter 5 presents results based on research specific objectives and a brief chapter summary. To show off the overall network performance, we considered a mix of both specific objective 3 and 4 in order to realize the comparative statistical conclusion between MIPv6 and the proposed Fast Handover MIPv6 solutions in same client-server architecture. MIPv6 and Client-Server Mobility Problems Based on the specific objective 1 of this study, we managed to gather and analyze the most important literature pieces that were critical to the fulfilment of the project’s purpose. This is addressed in chapter 2, which introduced the foundation of Internet Protocol Mobility. Mobile IP technology was then spread through different applications and used in innovation and technology inventions of the future. However, the nature of the technology projected diverse security risks technology limitations that IETF practitioners decided to integrate AAA protocols with MIP to serve manufacturers of mobile equipment who would decide to incorporate IETF standard protocols into their products. AAA infrastructure was used to support roaming operations, and also to secure mobile communications by establishing levels of security of information in Mobile IP environment as defined by IETF standards (Lin et al., 2009). Mobile IP technology was adapted to IPv4, but this protocol did not sustain the ever-growing number of mobile users and had some technical limitations that IPv6 later came to resolve. In that regard, Punithavathani, D. Shalini and Sankaranarayanan (2009) explained differences, coexistence and correlations between the legacy IPv4 and the all-IPv6 protocols as three mechanisms were used: dual-stack, packet tunneling, and translation mechanisms. The notion of NEMO and MN behavior inspired the instauration of MIPv4 and later MIPv6 which enhanced capacity and independence of MNs in the mobility decision making process while moving across different networks (Al-Kasasbeh et al., 2008). According to the findings in literature, MIPv6 as a technology, is used in a panoply of applications where we can cite Internet of Things (Choi & Koh, 2016 ; Chun, Kim, & Park, 2015), Lightweight Wireless Personal Area Network (Wireless Sensor Networks), Mobile Ad Hoc Networks where MNs can move from one MANET to another, and in other IP-based technology inventions for the future (Jung & Peradilla, 2013). Furthermore, handover latency was found to be one of the main driving forces in network performance evaluation (Islam et al., 2012). Therefore, solving the issue of a handover latency introduced by MIPv6 by reducing delay period was poised to have positive effect on the overall performance of the network. This means that by introducing FMIPv6 technology, we could potentially improve on the overall network QoS. However, both mobile technology solutions have introduced technological approaches such as Binding Update and Route Optimization, which resulted in multiple security concerns. But the implementation of MIPv6 as standardized by IETF provided technologies to implement in order to insure secure communications in the mobile environment (Chen et al., 2014). Therefore, IP Security (through packet tunneling ) and Return Routability Procedure were introduced to ensure proper authentication and identification of both the MN and the CN participating in the network communications (Ebalard, 2010). Mobility can be introduced in different other technologies, such as client-server. Client-server architectures were found limited to a LAN/WLAN that communicate with other networks over the Internet based on a non-roaming concept that could introduce much long down-time when the client want to connect to a network from an uncovered area. The client could be restricted from services even when he is in an imperative need to access them on the server. Therefore, in addition to proposing the introduction of MIPv6 technology foundation in client-server environment, the proposed solution suggested a decrease in handover delay time, hence improve the overall network performance in terms of QoS. To meet the intended goal, multiple factors, based on client-server and MIPv6 technologies, were introduced and improved in the proposed framework. The network is mobile, based on a host-to-host communication mode with reasonably higher bandwidth triggered by the exploration of low handover latency in result. One of the most important network performance metrics in the process of reducing handover latency was packet loss rate necessary for real-time applications such as Video Stream, VoIP, etc. This metric is low as long as the handover latency is low as well, hence good overall network performance. Finally, as chosen, all the factors appeared to have an impact on the realization of the end goal and on the overall network performance in both MIPv6 and Fast handover MIPv6 approaches (Tripathi, Radhakrishnan, & Lather, 2012). Design and Implementation of Client-Server MIPv6 The second specific objective as stated in chapter 1 of this study was about design and implementation of the proposed MIPv6 solution in client-server architecture that depended on specifications and parameters characterization that would lead to the intended solution and serve as ground work into recording simulation results that would later be analyzed. Therefore, the implemented solution included design and simulation of MIPv6 technology that we later improved by modifying or nullifying handover time-related parameters with the aim of improving on the overall network performance. As illustrated by the design in Figure 20, the network included a mobile client that is destined to keep communicating with its Home Agent, and that when moving from one network to another, instantly get internet connection every time it reaches an IPv6 access network coverage and start exchange (send/receive) information with the remote server over the IPv6 Internet. In the simulation environment (OMNET++ simulator), the implementation included a wide terrain as shown in Figure 21, on which the network was designed and configured with respect to actual network’s design constraints which helped prove the concept and validity of the intended solutions. Based on the network design in section 4.3, the implementation process invoked a list of parameters as shown in Table 7 that were used to configure the network in the INI programing language files through which MIPv6 was implemented and which lead to the improvement of the handover process that drew the path to the expected results as recorded in the results file. Precisely, for a faster handover process, we modified characteristics related to L2 handover delay, Route Advertisement Delay, and Duplicate Address Detection delay in the general configuration (Amin et al., 2011). In Tables 9, 10, and 11 three traffic models were developed and configured in terms of services to the client. Each service (HTTP, FTP and Video Stream) was simulated and tested with the same parameter values for both MIPv6 and FMIPv6. Evaluation of Network QoS of Client-Server MIPv6 and FMIPv6 As stated in the specific objective 3 and 4 of this research project, the designed and simulated solutions (MIPv6 and FMIPv6) in client-server needed to produce results that would help us verify and prove the concept. Results were then extracted using statistical quantitative method with emphasis on the average or mean values in the simulation results datasets (Loehnert, 2010 ; Iványi, 2005). As implemented in this research MIPv6 technology captured all network requirements in terms of handover delay, network services to clients, and network performance measurements that helped establish fundamental differences between MIPv6 standard network and its extension, in this case Fast Handover MIPv6 solution. The network considered a minimum of one client and one server, but the configuration provided was open for expansion and scalability possibility in terms of number of clients and servers that would depend on network planning and needs. As mentioned in Section 3.3 in research methodology Chapter, the results collected from the simulation of the proposed solution responded to functional and technical requirement of MIPv6 and Faster Handover MIPv6 as suggested in the research proposed framework in Section 2.5. Therefore, as recorded, Throughput, Packet Loss Rate, Handover delay and End-to-End delay metrics were used to measure and evaluate the overall network performance of Video Stream services using UPD protocol, whilst via TCP protocol, network Throughput, and Packet Error Rate (PER) metrics used to measure performance of FTP and HTTP network service performances. Finally, the comparative approach that was considered provided differences in performance between the proposed MIPv6 and Fast Handover MIPv6 technology solutions in client-server network architectures. Thus, the results collected form the simulation were presented. Simulated Results Summary The simulated network results bellow exhibit profound differences not only in adopted technology and handover process, but also in choice of services and applications used per protocol. All the simulation instances were subject to a total and repetitive simulation limit time of 120 seconds, while for rest of the simulation measurements and principles, all details are provided in Section 4.5.3. So, in this section we presented the recorded and collected results from simulated MIPv6 and Fast Handover MIPv6 network technologies, applied in client-server environment. UDP Handover Delay for Video Stream Services As suggested by Koodli (2009b), in its movement across networks, there is a period during which the MN (client) is unable to send or receive packets due to link switching delays and IPv6 protocol operations (configuration). Therefore, based on the dynamic data rate selection method used in this research with bitrates values of 6, 9, 12, 18, 24, 36, 48, and 54 Mbps, network performance in output showed a decrease in handover latency as shown in Table 11. Table 11: Processed Handover Latency for MIPv6 and FMIPv6 for Video Stream Services Bitrates Last Packet Time (Sec) Next Packet Time (Sec) MIPv6 HO Latency (Sec) FHO Last Packet Time (Sec) FHO Next Packet Time (Sec) Fast MIPv6 HO Latency (Sec) 6 Mbps 32.463552 36.547304 4.083752 32.59532 35.158772 2.563451 9 Mbps 32.463076 37.226589 4.763513 32.59485 35.1776 2.582751 12 Mbps 32.462836 36.035364 3.572528 32.59462 35.737004 3.142387 18 Mbps 32.462596 37.51496 5.052364 32.59438 35.64694 3.052563 24 Mbps 32.46248 36.054156 3.591676 32.59427 35.26686 2.672595 36 Mbps 32.46236 36.994072 4.531712 32.59415 35.396016 2.801871 48 Mbps 32.4623 36.85418 4.39188 32.59409 35.555688 2.961599 54 Mbps 32.462284 37.414884 4.9526 32.59407 35.855968 3.261899 As recorded in the PCAP file, the differences established between the last time a packet was received by a client and the next time the client receive a new packet after handover process provided path to calculating the differences in time value and lead to a considered client’s handover latency. The long sought fast handover delay demonstrated a difference that ran between 2 and 3.2 seconds as opposed to the standard MIPv6 whose handover latency ranged from 3 to 5 seconds with eight bitrates values run separately. To interpret the results data displayed in Table 11, the obtained numbers were plotted using a spreadsheet (Microsoft Excel) for data analysis that we used to compute the data as shown in Figure 23. Figure 23 is a graph capturing handover delay values from the MIPv6 and FMIPv6 simulation instances, where we established the difference between the last time a packet was received by a client and the next time the client receives a new packet after handover process. Figure 23: Handover Delay Results Report Since the network had only one user (client), the differences were established in terms of computed bitrates values run separately during the simulation as of 6 Mbps the handover delay was 4.08 seconds for MIPv6 and 2.56 seconds for Fast Mobile IPv6 simulation. With 9 Mbps we obtained 4.76 seconds for MIPv6 and 2.58 for FMIPv6 in handover latency. But for 12 Mbps, we observed a slight decrease in handover delay for both MIPv6 and FMIPv6 with 3.57 seconds, and 3.14 seconds, respectively. However, 18Mbps-bitrate introduced a larger handover delay of about 5.05 seconds for MIPv6 but FMPv6 recovered with a decay down to 3.05 seconds in handover latency. At 24 Mbps, the handover delay was recorded at 3.59 seconds for MIPv6, while for FMIPv6 we had 2.67 seconds. We observed 4.53 second-latency for 36 Mbps for MIPv6 and 2.8 seconds for FMIPv6. We could observe almost the same gap for 48 Mbps where the MIPv6 handover latency was 4.39 seconds and 2.96 seconds for FMIPv6 implementation. Lastly, we observe another rise in handover delay as to 54Mbps bitrates, the delay was of 4.95 seconds for MIPv6 and 3.26 seconds for FMIPv6. UDP Packet End-To-End Delay for Video Stream Services Figure 24 demonstrates the difference in packet end-to-end delay between the implemented MIPv6 and in extension, FMIPv6 with a clearly better network performance and ultimately better QoS since the handover latency is reduced. Figure 24: UDP End-To-End Delay for Video Stream The graph illustrates a relatively low packet end-to-end delays as they vary based on the introduced simulation method based on different bitrates values. At 6Mbps the end-to-end delay was 0.001985737 seconds for MIPv6 and a slightly low delay of 0.001903426 seconds for Fast Mobile IPv6. With 9 Mbps we had 0.001477538 seconds for MIPv6 and 0.001399753 seconds for FMIPv6. For 12 Mbps, we observed a slight decrease in delay for both MIPv6 with 0.001233889 seconds and FMIPv6 with 0.001155962 second delay time. Steadily, the delay decreased by almost the same proportion since 18 Mbps produced an end-to-end delay of exactly 0.000984387 seconds for the implemented MIPv6 and 0.000903983 seconds for FMPv6. At 24 Mbps, the end-to-end delay was recorded to 0.000862317 seconds, while for FMIPv6 we had 0.000778113 seconds. At 36 Mbps, the simulation introduced a 0.000738048 second end-to-end delay time for MIPv6 and 0.000655351 seconds for FMIPv6. For 48 Mbps however, the MIPv6 end-to-end delay was of 0.000677186 seconds and 0.000592667 seconds for FMIPv6 implementation. Finally, the handover delay was 0.000658089 seconds for MIPv6 and 0.000572876 seconds for FMIPv6 at 54 Mbps. UDP Throughput for Video Stream Services Throughput is one of the most important Quality of Service and network performance metrics for any kinds of networks at all layers. It therefore made sense to measure throughput in this project in order to determine performance level differences between MIPv6 and the proposed FMIPv6 technologies implementation in client-server architecture. Figure 25 illustrates the overall performance evaluation results for QoS/Throughput metric with differences established between MIPv6 and FMIPv6 technologies implementation. Figure 25: UDP Video Stream Throughput Implementation Results A for the outcome from the other metrics, throughput measurements were established with different simulation instances and configuration parameters between MIPv6 and FMIPv6 technologies with the results in output suggesting that at 6 Mbps the throughput was of about 0.734 Mbps for MIPv6 while it was up to 0.745 Mbps in FMIPv6 implementation. For a 9Mbps, the throughput was of about 0.732 Mbps for MIPv6 implementation, and 0.747 Mbps for FMIPv6. With 12Mbps of bitrates, the simulation produced a throughput of 0.739 Mbps and 0.7419 Mbps for MIPv6 and Fast MIPv6, respectively. For 18 Mbps, the throughput went a little lower for MIPv6 with 0.729 Mbps and kept the pace for FMIPv6 with 0.742 Mbps. At 24 Mbps, the throughput was of 0.7389 Mbps for MIPv6 and 0.7442 Mbps for FMIPv6, while at 36 Mbps of bitrates, the throughput was of about 0.732 Mbps for MIPv6 and 0.743 Mbps for FMIPv6. With 48 Mbps, we obtained a performance of 0.732 Mbps for MIPv6 and 0.7439 Mbps for FMIPv6 in throughput. Finally, the implementation kept the output throughput at almost the same level as we obtained a throughput of 0.73 Mbps for MIPv6 and 0.74 Mbps for FMIPv6. UDP Packet Loss Rate for Video Stream Services Video Stream services simulation introduced Packet Loss Rate majorly caused by handover latency. It is as well a critical network performance and QoS metric that helped us evaluate the numbers in percentage of packets that the client was not able to receive from the server after computing the difference between the total number of packets sent by the server and the total number of packet received by the wireless client form two different networks that were served by the server during the client mobility. Figure 26 illustrates differences established between the total of packets sent by the server and those received by the client, and then calculated the percentage of the number based on the total of packets sent by the server. Figure 26: UDP Video Stream Packet Loss Rate Implementation Results Therefore, PLR measurements were as well established with different bitrates parameters between MIPv6 and FMIPv6. At 6 Mbps the simulated PLR was of about 5.734% for MIPv6 while it was down to 4.418% in extended FMIPv6. With 9Mbps, the PLR was of about 6.19% for MIPv6, and 4.18% for FMIPv6 as shown by the respective graph lines. With a slightly different variation, at 12 Mbps of bitrates, we observed 5.23% in PLR for MIPv6 and 4.87% for Fast MIPv6. At 18 Mbps bitrates, the PLR went a little higher for MIPv6 with 6.5 % and kept the pace for FMIPv6 with 4.86%. At 24 Mbps, the PLR were of 5.27% for MIPv6 and 4.58% Mbps for FMIPv6, while at 36 Mbps of bitrates, the PEL were of about 6.09% for MIPv6 and 4.64% for FMIPv6. Furthermore, with 48 Mbps-bitrates, we obtained a PLR performance of 6.034% for MIPv6 and 4.61% for FMIPv6. Lastly, the output throughput obtained at 54 Mbps produced a PLR of 6.3% for MIPv6 and 5.03 % for FMIPv6. TCP Network Throughput for File Transfer Protocol Services Using TCP protocol, we measured the overall performance and calculated the network throughput on the server to establish the performance differences between MIPv6 and the implemented FMIPv6 networks solutions. Based on the output in the created datasets, TCP throughput was measured in Kbps for both of the networks. The impression is that MIPv6 still has poor performance than the implemented FMIP6. TCP service performance was recorded on the server in this instance as opposed to UDP protocol. Figure 27 presents the overall throughput performance in terms of network QoS for both MIPv6 and FMIPv6 establishing differences based on configured bitrates values. Throughput for this service is measured in Kbps. Figure 27: TCP FTP Throughput Implementation Results For TCP protocol, throughput measurement was important since it paved the way to figuring out how fast the file was transmitting and could help calculate the time it took to transmit the total file size. Therefore, here we found out that with different simulations and configuration parameters in terms of bitrates between MIPv6 and FMIPv6 technologies gave as well different values in the output that at 6 Mbps the simulation output a throughput value of about 64.62 Kbps for MIPv6 and 68.77 Kbps in FMIPv6 implementation. At 9 Mbps, we got a throughput of 58.01 Kbps for MIPv6 and 59.74 Kbps for FMIPv6, while for 12 Mbps of bitrates, the simulation produced a throughput of 52.53 Kbps and 75.25 Kbps for MIPv6 and Fast MIPv6, respectively. At 18 Mbps, the throughput for MIPv6 was of about 72.33 Kbps and 75.38 Kbps for FMIPv6. At 24 Mbps, the throughput was of 71.64 Kbps for MIPv6 while for FMIPv6 we four a throughput of 102.97 Kbps. While at 36 Mbps of bitrates, we found a throughput of 66.41 Kbps for MIPv6 and 73.305 Kbps for FMIPv6. At 48 Mbps, we obtained a throughput performance of 81.47 Kbps for MIPv6 and 84.05 Kbps for FMIPv6. Finally, at 54 Mbps, the implementation opened a wider gap between throughput performance instances with 73.2 Kbps for MIPv6 and 111.29 Kbps for FMIPv6. TCP Packet Error Rate for File Transfer Protocol Services As mentioned in Chapter 2, PER is very critical for TCP communication (Abukharis et al., 2014) . The simulation results demonstrated a PER of 0 % and in one separate case (54 Mbps) a value vas produced but as well insignificant and close to 0 %, we were able to conclude on the importance of the sensitivity of TCP Packet Errors that the simulation research output produced the expected network performances. Figure 28 illustrates the fundamental issue of packet error rate (PER) in File Transfer Protocol service implementation with significantly negligible values which appeared appropriately respondent to the exigence of TCP protocol. Figure 28: TCP File Transfer Protocol Packet Error Rate Implementation Results File Transfer services were part of the implementation bull’s eyes of both MIPv6 and FMIPv6 technologies in client-server architecture using the eight different bitrates values to measure the overall performance of the services contributing to the network QoS. The differences as displayed in Figure 28 still gave a slight favor to the FMIPv6 implementation since the PER for MIPv6 showed a slight pick than FMIPv6 only at the last bitrate measure (54 Mbps). From 6 Mbps up to 48 Mbps bitrates values, all the PER values were promisingly recorded at 0% error. Only 54 Mbps bitrates did not exactly equate 0% but rather took extremely low values for both MIPv6 and FMIPv6 technologies with 0.00000000728% and 0.00000000707%, respectively. This performance confirms the reliability of the implemented TCP protocol. TCP Throughput for HTTP Services For HTTP service implementation with TCP protocol, we considered only one bitrate value as defined in the general parametrization of the network solutions in Table 7. The value was set to 54 Mbps based on IEEE 802.11b standard. With a single wireless bitrates value, we as well-found differences in simulation results as the throughput performance for MIPv6 client-server is considerably lower than the one obtained by FMIPv6 implementation for HTTP service. Figure 29 shows the throughput implementation results for HTTP services over TCP protocol as we opted to consider only one instance of bitrate for both MIPv6 and FMIPv6 in client-server network. Figure 29: TCP HTTP Throughput Implementation Results Based on the Web services configuration, QoS performance shows that with 54 Mbps of bitrate, the throughput is 418.05 Kbps for MIPv6, but higher with the implementation of FMIPv6 which gave us a throughput of 520.67 Kbps. The difference in performance is of over 100 Kbps between the two network implementation instances. For TCP protocol implementation, QoS throughput performance values were measured on the server side of the network as shown in Figure 29. TCP Packet Error Rate for HTTP Services As implemented with HTTP service, the networks’ overall Packet Error Rate was extremely low for MIPv6 tending to 0%, while for FMIPv6 the value was exactly 0% considering the general bitrate implementation parameter that was set to 54 Mbps based on IEEE 802.11b. With a single wireless bitrates value and as displayed in Figure 30, TCP protocol with HTTP service demonstrated once again how important it is to have a minimum or 0% error in the transmitted packets with respect to the TCP’s zero tolerance posture in packet errors. The results here demonstrated the level of reliability of TCP protocol in Web services due to the insignificant values of the PER for both MIPv6 and the FMIPv6 in Figure 30. Figure 30 illustrates the Packet Error Tate implementation results for HTTP services over TCP transport protocol considering one bitrate instance (54 Mbps) as configured in the general simulation configuration that we set up for both MIPv6 and FMIPv6 network simulation instances in Table 7. Figure 30: TCP HTTP Packet Error Rate Implementation Results Network performance and Quality of Services metric (here PER) showed that with the identical 54 Mbps bitrates for both MIPv6 and FMIPv6, the Packet Error Rates was as insignificant as 0.000000022 % for MIPv6, and for FMIPv6 implementation, the value went down to 0% in Packer Error Rate which qualified the reliability of the implemented TCP protocol. Chapter Summary This chapter gave insights on the findings from the defined objectives that this study was pursuing. It presented results based on the accumulated literature as instructed in the first specific objective. The overall findings brought to light different literature pieces that were important to the study, and that led to the actual implementation of the proposed solution. Furthermore, suggested in the literature, IPv6 protocol was the focal subject and was used based on its favorability not only in its capacity to accommodate a larger number of IP addresses than the legacy Internet (IPv4), but also since it provides ability to users, as clients (in this study), to manage mobility tasks instead of the network side taking over the role. So MIPv6 not only solved the problem of seamless connection to the servers, but also allowed any client that is close to any IPv6-supported Access Router (Router and AP) to immediately start exchanging or receiving packet and information from the server as soon as it is granted access to internet, i.e. as soon as it finishes registering with the Home Network through HA. Internet accessibility is then conditioned by the handover process that could still be damaging to real-time applications such as Video Stream services. Therefore, to provide solution to that issue we proposed a solution that included a faster handover process with the goal of reducing the handover delay, therefore improving on the overall network performance in term of Quality of Service. This chapter exclusively presented the results recorded and computed for the various network services (Video Stream, FTP and HTTP) as they were differentiated based on the transport protocol (TCP or UDP) that were used to perform the simulation instances. As the recorded results show, regardless of the transport protocol chosen, every MIPv6 metric or performance measurement ended up short as opposed to FMIPv6 service implementation. The displayed results gave us the idea of how important FMIPv6 can be in preserving the overall network Quality of Services. CHAPTER 6 DISCUSSION, CONCLUSIONS AND RECOMMANDATIONS Introduction Following the implementation of MIPv6 technology in client-server architecture with a reduced and fast handover technique, all technical implementations and data collection, this chapter discusses the results and findings of the research and draw subsequent conclusions from the data analysis and representation. It illustratively expands on the delivered values by both MIPv6 and the Fast Handover MIPv6 client-server architectures in data analysis, helping address the problem of client (Mobile Node) mobility and handover latency for better or acceptable network QoS based on either TCP, or UDP protocol with the implemented Video Stream, HTTP, and FTP application services. This chapter therefore, is broken down into a brief introduction, summary of the important elements of the project, discussion of the major findings of the study, and finally conclusions from the research findings. Summary Purpose of the Study The purpose of this study was to enhance capacity and performance of network services in client-server architectures using MIPv6 host-based technology and to introduce technological measures in extension that helped reduce the handover latency, hence enhance the overall performance of network services in terms of QoS. The research purpose also implied that the network should also be able to ensure security of network traffic since the client is operating in mobile wireless environment and is highly dependent on the insecure Internet connectivity. As developed, designed, implemented and later analyzed, the proposed solution to the research problem satisfactorily met the basic and important requirements of the overall project purpose. Specific Objectives In a hierarchical order, the research aimed at meeting its specific objectives with the intent of archiving the overall expectations. The specific objectives were established as follows: Specific objective 1 was to evaluate MIPv6 technology and client-server network mobility problems though literature review and propose a solution framework. Specific objective 2 was to design and implement client-server architecture using an optimized and secure MIPv6 solution in a simulated environment. Specific objective 3 was to evaluate network Quality of Service of the implemented MIPv6 solution for FTP, HTTP and Video Stream services. Specific objective 4 of the project was to implement and evaluate client-server Fast Handover MIPv6 solution for better Quality of Qervice. Research Methodology Chapter 3 introduced the research methodology that led to the implementation of the proposed solution providing a path involving steps and approaches taken toward the fulfilment of research objectives. The research methodology was organized in such a way that the proposed research design that introduced all the technical and logical needs to design, setup and configure in simulation of the platform needed to implement the proposed solution as elaborated in section 2.9. The methods used to implement the proposed research solution involved a comparative approach in performance of two different technologies (MIPv6 and Fast Handover MIPv6) in client-server networks. To collect data, we used dynamic rate selection method that consisted of eight different bitrate values computed and recorded through different simulation instances to produce data that lead to a comparative analysis of the data based on implemented services (HTTP, FTP and Video Stream). The implementation of application services involved the use of both connectionless and connection-oriented transport protocols (UDP/TCP) that led to recording and extracting simulation results that we finally analyzed using statistical quantitative methods in comparing results outputs in scalar, vector, as well as histogram file types. To perform the implementation of the proposed solution, we simulated the designed network in OMNET++ simulation tool with INET Framework extension. The research procedures proposed in this study focused on the objective of simulating MIPv6 and Fast Handover MIPv6 in implementation process that is depicted in a diagram detailing different steps taken from the network design to implementation and results extraction and analysis as we used OMNET++ simulation tool and INET Framework in extension. Finally, all simulation results recordings were configured in mean values. This helped us record more implicit values and ensured efficiency in memory management. To measure packets numbers simulation results, some statistics such as Received Packet, Packet Sent were singled out in order to be recorded considering their importance in network performance and overall QoS measurements. All statistics were then kept in the “results” file in OMNET++ that we created in the INET’s xMIPv6 model directory and that were gathered in a constructed dataset setting to collectively be analyzed. In the simulation results files, data was recorded and generated in three different types of files including scalars (.sca), vectors (.vec) and event log files (.elog), all saved in data analysis (.anf) file in OMET++. The obtained data was then exported to spreadsheet files to be analyzed as a combination that established the final statistical and quantitative differences in network performance and overall Quality of Service preferences between the proposed MIPv6 and FMIPv6 technologies established in client-server architectures. Major Findings The implementation of the proposed solution as presented in the proposed design confirms the profound dependence of Mobile IPv6 on the handover delay for the network to improve on its performance in term of QoS. Handover delay was of a larger proportion for MIPv6 than for the proposed fast handover MIPv6 solution in results findings in Table 11, which drove almost all the performance metrics to higher standards for the proposed fast handover MIPv6 than for the implemented standard MIPv6. So, based on the collected literature, MIPv6 on its own, as implemented in this project, bridged the gap of mobility capability and service performance for a mobile client in constant need of services and communication with the server. Therefore, after improving on the handover delay time, which we found contributed considerably on the performance of the implemented MIPv6 network, the newly proposed fast handover solution improved all network performance metrics regardless of the transport protocol (TCP or UDP) in use (Al-Kasasbeh et al., 2008). Network security for the volatile wireless network and the unpredictability of attacks on internet drove to the implementation of the required IPSec that protects the client’s Binding Update message to the HA by creating a IPv6 tunnel until the client is ready for route optimization which is also protected through Return Routability procedure based on the increased vulnerability in the RO process (Xianhua & Sui, 2011). As configured, the network provided results statistically recorded in the nature of scalars, vectors, or histograms and drove to the constitution of datasets that helps plot the results for a clearer interpretation through graphs and statistically represented data. The results analysis was presented based on two major transport protocols split in service support as UDP bore Video Stream services and TCP took care of FTP and HTTP services. Therefore, based on this transport protocols, services were used individually for simulation. As the recorded results show, regardless of the transport protocol chosen and services implemented, every MIPv6 performance measurement or metric ended up short as opposed to FMIPv6 implementation. The displayed results in section 5.4 gave the idea of how important FMIPv6 can be in preserving the overall network QoS. The statistical computation of the simulation results allowed us to record and compute results, which showed significant performance differences between the implemented MIPv6 network and the proposed Fast Handover MIPv6. As a result, both technologies implemented a Video Stream application using UDP, which demonstrated a difference in performance on Throughput, Packet Loss Rate, Handover delay and End-to-End delay. Using TCP protocol, FTP and HTTP application demonstrated a better performance measuring the networks’ Throughputs and Packet Error Rates. Finally, it is safe to conclude that the overall network QoS improved with less handover delay. Discussions MIPv6 and Client-Server Mobility Problems Specific objective 1 in Section 1.3 aimed at establishing the research problem through literature and identifying problems surrounding the inability for traditional client-server architectures to provide mobility services in terms of required coverage and resources, which we discussed at lengths in chapter 2. Indeed, in Chapter 2 we revealed the historical perspective of mobility in Internet Protocol with more approaches driven by the need for Mobile Nodes to communicate seamlessly, as it has long been the case for telecommunication networks with the adoption 1G, 2G, 3G, 4G and today 5G (Banerji & Chowdhury, 2013 ; Céspedes, Shen & Sherman, 2015). As IETF established partnership with other group such as Mobile IP working group to continue upgrade features in response to demands from individual to enterprise levels, a new charter was rolled out outlining various interactions between MIP and AAA (Lin et al., 2009). AAA servers were then used for technical and security services enhancement in MIP technologies. On the other hand, client-server architectures were found to be restricted to an established LAN or WLAN with inability to provide mobility services for clients in need of moving to further distances. However, with the integration of MIPv6 in its host-based configuration, it was found to be possible to provide mobility services to a moving client (Xiaorong, et al., 2013). Therefore, Mobile Internet Protocol was introduced in IPv4 protocol architecture to ensure mobility and usability of mobile services by a mobile user. But due to different deficiencies of IPv4 protocol such as a limited address range, lack of mobility, security and so on, IPv6 protocol was introduced and so was MIPv6 protocol to expand the network and take care of network mobility and security functions. The ability for IPv6 to coexist with IPv4 made it possible for researchers such as Punithavathani at el. (2009) and Sha et al. (2017) to establish some methods of communication between IPv4 and IPv6 networks that included three categories of communication which are : dual-stack, tunneling, as well as translation. IPv6 was initially based on IPv4 architecture but was adopted for better and more secure services. Mobile IPv6 technology was standardized to provide solution to MIPv4 user mobility and different other challenges under the leadership of IETF working group. Network mobility was introduced on most of the OSI layers. MN was then able to manage and decide on its mobility and association procedure with any other network, hence host-based mobile technology. The network included a HA, CN and MN as nodes that could communicate and exchange information via other network resources and gateways using a process called “triangle routing” (Sanguankotchakorn & Jaiton, 2008). Furthermore, other technologies appeared useful in the implementation of MIPv6. The research process adopted IPsec techniques by establishing a bidirectional tunnel between the MN and its HA to secure communications that are transmitted between the CN and the MN and the MN and its HA. In IP addressing, a decision was split into two different addressing approaches that included stateful and stateless address configuration modes, where stateful mode appeared more useful and adapted to the proposed host-based MIPv6 technology in proposed solution. Additionally, through the same IETF technology proposals, the notion of Route Optimization was introduced. In fact, Route Optimization (RO) as we implemented it in the MIPv6 and FMIPv6 network, was used in order to decrease overhead at the IPv6 Router that connects the HA to the FA. It was also of capital utility since it offered a way for both the client (MN) and Server (CN) to forward packets to each other directly without the intermediary of the HA (Al-Kasasbeh et al., 2008). Optimizing packet route also helped increase the number of packet to possibly be sent in the network per second, which resulted in a positive impact on the overall network performance. But RO, as well as MIPv6 technology on its own introduced security concerns that should be taken care of by using special security measures such as IPsec and Return Routability Procedure. With the implemented MIPv6, if there is no security mechanism such as IPsec, the CN does not know which MN sent the BU before even deciding of the RO process. Security concerns introduced by RO is of a broad scope since communications between the CN and the MN are no longer controlled by the HA through IPv6 tunnel but transmitted directly over the Internet to the remote CN. To protect that information, IETF introduced Return Routability procedure, which in general allows CN to generate BU messages exchange and overall communication security keys that MNs will have to use in the process to identify and authenticate themselves in the network so they can keep communicating and exchanging information with the CN. IPsec on the other hand, was used in advance for the first BU between the MN and HA, opening a secure and bidirectional IPv6 tunnel between the two nodes to communicate safely, with the MN using its CoA and HoA to contact the HA. However, BU is not secret, but it always needs to be sent from a legitimate MN. So, knowing the way the MN was attributed its IPv6 address and its CoA is very important. In fact, based on the consulted literature, MIPv6 network is usually configured in such a way that IPv6 addresses of the MN is assigned using the stateless address configuration, considering the prefix of the MN’s (host) IPv6 address. So based on these abilities and the client’s sense of independence in mobility decisions, and considering the wider IPv6 address range, multiple technology domains have been adapting the technology including the revolutionary Internet of Things, Wireless Sensor Network, and so on (Jara et al., 2013). To address handover latency performance deficiencies, Fast Handover MIPv6 solution was introduced and adopted in the initial MIPv6 architecture. In fact, as Vassiliou and Zinonos (2010), and also Kwon, Kim, Bae, and Suh (2010) studied MIPv6 and found a handover latency period of 3.6 and 5 seconds, respectively. These delays are all acceptable but rather damaging in real-time applications where the MN needs a constant and seamless connectivity to access services. FMIPv6 was introduced to roughly narrow the latency gape of both L2 and L3 handover. Therefore, in this project’s proposed solution, we systemically introduced factors that commonly had the potential to influence network performance and QoS in both wireless client-server and MIPv6 networks and introduce enhanced or adapted version of the factors to produce a set of factors characteristics respondent to Fast Handover MIPv6 handover mobility requirements. The aim of this research however, was to create an environment where client-server architectures could be technically adapted in MIPv6 technology. So, we had to make sure clients (or mobile users) remained connected and that they could seamlessly communicate with the server or (CN) even after the clients change network attachments. Furthermore, Network performance in both MIPv6 and FMIPv6 was measured on virtually all the TCP/IP model layers with emphasis on some important metrics such as handover latency, end-to-end delay, throughput, Packet Error Rate, Packet Loss Rate. All these metrics measured the overall network performance based on different applications services such as FTP, HTTP and Video Stream that would later be implemented and tested through a combination of both UDP and TCP protocols in both MIPv6 and Fast Handover MIPv6 technologies. Design and Implementation of Client-Server MIPv6 As stated by specific objective 2, the research focused on network design and implementation of two different solutions adapted in client-server architectures. At first, we implemented Mobile IPv6 technology in client-server architecture, and then we improved upon it to guarantee a well- performing solution that reduced the overall handover latency. Hence, the proposed solutions focused on implementing MIPv6 technology, and then improve on the handover latency by implementing Fast Handover Mobile IPv6 technology in client-server architecture. After designing the network architecture using Microsoft Visio network diagram module as illustrated in Figure 20, we transformed the architectural design into implementable network using Discrete Event Simulation approach. The choice of a simulator was driven by the characteristics of a DES where simulated entities and network units are moving between different states based on time (Maidstone, 2012). As considered in this project, we used test statistics event-driven method, which reflects that the approaches taken can be compared to each other (Fajar et al., 2018). Therefore, after examining different discrete event simulators, we found some similarities and abilities to respond to the project purpose with just OMNET++ and NS2. However, based on some characteristics such as hierarchical models support, GUI support, time to learn, and others, NS2 fell short on giving satisfaction to the demanded abilities as projected by the research objectives. Hence, we fairly chose OMNET++ to implement both MIPv6 and its extension, being Fast Handover MIPv6. Using OMNET++, the simulation took into account the modularity of the simulator, which based on INET Framework and its extensions such as xMIPv6, two modular element embodying two different programming languages (Network Definition and Network Initialization) were chosen as they helped develop the required elements for the implementation of the proposed solution (Mayer & Gamer, 2008). The design model was entirely based on the architectural planning of INET Framework’s mobile IPv6 modules from which we constructed and developed working client-server architecture based on IETF specifications and standards as illustrated in Figure 21 in section 4.3. The network was built as a standard MIPv6 from which we latter proposed an enhancement in capacity in implementing a reduced handover latency technique known as Fast Handover Mobile IPv6. This project however, focused on two different network mobility aspects that we explained in Chapter 2, and that included handover latency and route optimization. In including these aspects in the implementation of the project referring to the configuration parameters in Table 7, we practically ensured an improvement of the overall QoS in terms of performance of MIPv6 network achieving a faster and reliable mobility process (FMIPv6) the services (FTP, HTTP and Video Streaming) as configured in Tables 8, 9, and 10, and used with both TCP and UDP protocols between the server and its clients. Furthermore, through both aspects, we implemented required security measures such as IPv6 Packet Tunneling and Return Routability Procedure, that protected communications between the client and the HA, and the client and Server (the CN). The simulation included one client, one server, which gave room to network expansion in terms of number of users and servers based on needs, while the network part is managed by three routers (HA, IPv6Router, and FA) with HA and FA being connected to the client through their respective APs. To verify the accuracy of the simulation, we attested a rapprochement in simulation behavior and results with the measured produced in (Yousaf et al., 2008). All configurations that made the difference between the two implemented technologies (MIPv6 and FMIPv6) were set up in the general configurations in Table 7, potentially influencing all modules and applications that we implemented to test and improve the network QoS. Evaluation of Network QoS of Client-Server MIPv6 and FMIPv6 The evaluation of the implemented technology was practically based on definition and testing of simulation characteristics that we deem deliverables of the proposed solution. Therefore, the network setup is performed based on different network modules, protocols and parameters working together to collectively define and configure the functionalities needed for the whole network configuration. Three languages were used implement the solution. Hence, as C++ language was used to defined network modules and components behavior with a set of libraries and messages, NED language was used to define network entities (nodes), connections and different interfaces on nodes, and INI programming language to configures the behavior of network nodes in the simulation configuration file for OMNET++ environment (Varga & Hornig, 2008). Therefore, serving as network configuration file, INI module was used to configure network parameters. The simulation was configured with a set of general parameters where simulation time limit was set to 120 seconds to all the implemented services (FTP, HTTP, and Video Stream) based on both TCP and UDP transport protocols, and the system set to record the “event logs” that highlighted the timing and names of events as nodes communicates or send messages to each other. To test the network, we also considered different simulation assumptions pertaining to the mobile client behavior, network and radio transmission characteristics. The MN speed was set to 10 meters per second (mps) for both network implementation instances (MIPv6 and FMIPv6) with mobility constraints and client positions limited at (195m, 25m) and (945m, 180m). All the existing and needed implementation characteristics were developed and configured to allow a swift realization of the project’s objectives. The findings presented in Chapter 5 extrapolates statistical representation of the results based on different network performance metrics including throughput, handover delay, end-to-end delay, packet loss rate, as well as packet error rate that we tested and applied to services such as FTP, HTTP, and Video Streaming. UDP Handover Delay for Video Stream Services Koodli (2009b) suggested that there is a period during which the MN (Client) is unable to communicate with other nodes in the network due to link switching delay and the delay for configuration of IPv6 address within the FA. Handover latency was then measure using UDP protocol since it helped implement Video Stream services, the most important application to be preserved in terms of packet loss as it requires a real-time format for the client to what the stream at its best performance. Based on the 8 bitrates values that we considered (6, 9, 12, 18, 24, 36, 48, 54 Mbps), we were able to extract different handover latency values from the difference in time that the last packet was received by the client with the time the very next packet is received by the client, all recorded in the PCAP file. Each resulted value each was plotted on a histogram chart to explicitly demonstrate the differences between handover performances of MIPv6 and FMIPv6. According to the displayed results, the highest handover latency was found where bitrates is equal to 18 Mbps with a latency of 5 seconds, whilst the lowest level of handover latency could be seen at 12 Mbps with 3.57 seconds. However, after implementing the proposed FMIPv6 in the client-server environment, we used a comparative evaluation with the implemented standard MIPv6 and noted a decline in handover latency where the maximum value is 3.26 seconds at 54 Mbps of bitrates, and the minimum value being 2.56 seconds at 12 Mbps. Differences established between handover latencies on both MIPv6 and Fast Handover MIPv6 instances highlighted the imperative outperformance of the MIPv6 handover latency by Fast Handover implementation in MIPv6 technology. So, with just shorter handover time, there is a potential of achieving much greater performances in terms of network QoS. Much of these values obtained in Handover Latency could be compare with values such as 7 seconds in MIPv6 and 3 seconds in FMIPv6 in a test bed evaluation as studied by Pieterse et al. (2013b). But also in variation of handover delay, values such as those obtained by Yousaf et al., 2008 could also be compared to the values obtained in this research. UDP Packet End-To-End Delay for Video Stream Services The end-to-end delay was recorded at the server level since unlike the client, the server is connected to the network using Ethernet interface. Based on the Video Stream service configuration, packets transmitted between the client and the server introduced a very low end-to-end delay implementation instances (MIPv6 and Fast Handover MIPv6) with a proportion of 0.6 ms as the lowest value and 1.985 ms as the highest value for MIPv6, whilst Fast Handover MIPv6 process introduced a lower level of delay in end-to-end communications between the client and the server with the lowest and the highest delays being of 0.5 ms and 1.903 ms, respectively. The End-to-End delay values obtained out of both simulated technologies (MIPv6 and FMIPv6) in client-server architecture presented steadily decreasing values with 6 Mbps corresponding to the highest end-to-end delay value and 54 Mbps corresponding to the lowest value. Hence, end-to-end delay decreased as bitrates values increased, despite the low delay registry that is impressively good for both of MIPv6 and FMIPv6 implementation instances in the proposed client-server architecture. UDP Throughput for Video Stream Services Throughput analysis for the implemented solution in client-server architecture weighed a very high in seeking the improved handover latency, hence improved QoS for a better network performance. Throughput was also measured considering different bitrates values that were adapted to both client-server implementation instances. The values obtained with throughput measurements were recorded through the WLAN interface at the client node. Based on the volatility and unpredictability, and the mobility characteristics of wireless medium as used in this project, the throughput values were rather unstably established along with the underlying bitrates values. Therefore, we found out that the highest throughput performance in MIPv6 implementation was when the client was transmitting at 12 Mbps of bitrates with a relative value of 0.739 Mbps of throughput performance, while the lowest values was recorded at 18 Mbps with 0.729 Mbps of throughput performance. On the other hand, the introduction of Fast Handover MIPv6 technology bolstered the throughput performance with at least 0.7406 Mbps as the lowest throughput value at bitrates of 54 Mbps, and 0.747 Mbps as the highest throughput value at 9 Mbps. UDP Packet Loss Rate for Video Stream Services As mention in Chapter 2 of this research project, Packet Error Rate is one of the most important network performance metrics in handover-related performance measurements especially in measuring QoS of real-time applications such as Video Stream. This metric helped realize how many packets the client did not get to receive while watching the streaming video. Calculating the rate fell into seeking the percentage level of packet loss that the client encountered during the streaming of the video stream as opposed to what the server was transmitting in real-time (total number of packets sent). Therefore, the total number of packets transmitted by the server minus the number of packet that the mobile client was able to receive while he was doomed to momentarily encounter disconnect from the entire network produced a difference from which we evaluated the percentage that was also based on the total number of packets sent. Therefore, the overall view shows a better performance for FMIPv6 with less packet loss than MIPv6 which had a higher loss rate recorded. Thus, MIPv6 present its highest packet loss rate at 18 Mbps with 6.5 % of sent packets lost, and the lowest at 12 Mbps with 5.2 % of sent packets lost. On the other hand, the implementation of FMIPv6 expectedly decreased the loss rate value for all the tested bitrates values with the highest packet loss rate having been recorded at 54 Mbps with 5 % of sent packets lost and the lowest packet loss rate recorded at 9 Mbps with 4.1 % of sent packets lost. TCP Network Throughput for File Transfer Protocol Services As for UDP, we used TCP protocol to measure network throughput of the implemented client-server architecture based on different bitrates values for both handover latency technologies. Based on the display result, Fast Handover MIPv6 is still performing better than the standard MIPv6. Remarkably, throughput in both MIPv6 and FMIPv6 is performing poorly for TCP implementation than for UDP implementation. Therefore, only two simulation instances reached a throughput of 0.1 Mbps while others were stagnating in the proportions of dozens of Kbps. However, the performance differences were any different from the gaps displayed in UDP analysis. Therefore, the lowest value of the overall throughput implementation for MIPv6 was 53.5 Kbps recorded at 12 Mbps, and the highest valued being of 81.47 Kbps was recorded at 48 Mbps. On the other hand, FMIPv6 displayed a starling increase in some instances while in others the gap was of a narrow proportion. Thus, for Fast Handover MIPv6 the highest displayed throughput was of 111.29 Kbps recorded at 54 Mbps, and the lowest value being of 59.74 Kbps was recorded at 9 Mbps. TCP Packet Error Rate for File Transfer Protocol Services As suggested by Abukharis et al. (2014), Packet Error Rate in a metric is very critical to connection-oriented communications. Therefore, the implementation of MIPv6 as well as the enhanced FMIPv6 demonstrated the sensitivity of TCP (with FTP service here) protocol in terms of PER since both technologies recorded 0 % of loss in packets for almost all the bitrates values. However, even though some values were recorded for both 48 Mbps with and 54 Mbps in packet error rate estimates, they were of a very insignificant (very close to 0%) proportion, responding to the sensitivity of TCP communication to errors in packets. So, both MIPv6 Fast Handover MIPv6 again displayed a difference in their performances in terms of packet error rate, which is one of the very important metrics of the network overall QoS. TCP Throughput for HTTP services For HTTP implementation, we considered the general bitrates implementation parameter that was set to 54 Mbps based on IEEE 802.11b. With just a single wireless bitrates value for HTTP services, we found differences as well in simulation results as throughput performance for MIPv6 client server is considerably (over a 100 Kbps margin) lower than the one obtained by FMIPv6 implementation for HTTP services. Therefore, with MIPv6 implementation, the network throughput value was of 418.05 Kbps, whilst it was up to 520.678 Kbps for Fast Handover MIPv6, highlighting the importance of handover latency improvement driven techniques in MIPv6 network experiences. TCP Packet Error Rate for HTTP Services The network overall quality of service for TCP-related services relied very much on the ability of the simulated network to produce acceptable values for PER, since connection-oriented transport protocols are rather sensitive to errors in transmitted packets. Thus, based on a single bitrates value (54 Mbps) for both MIPv6 and Fast Handover MIPv6 client-server implementation, and with an insignificant difference, Fast Handover MIPv6 outperformed the standard MIPv6 with recorded PER of 0 %, while PER for MIPv6 in HTTP services implementation was very low (close to 0 %), reiterating the consistency of a low or inexistent PER for TCP-related services. Conclusions All the project’s specific objectives have been met. Indeed, for specific objective 1, literature was obtained, analyzed and applied to different study areas. It gave giving insights on MIPv6 and related areas and on the prospect their interaction with client-server technology to attain the purpose of the project. Specific objective 2, which aimed to and produced a technical or architecture commodity that implemented (through simulation) MIPv6 technology in client-server architecture with respect to important technical requirements such as security and Route Optimization. Since MIPv6 was poised to introduce rather longer handover delay, which is was deemed unsatisfactory to the needed better network performance, specific objective 3 has been met, since the project implementation was able to plainly evaluate network performance of MIPv6 technology in client-server architecture and ensured a reduction in handover latency by introducing another technology approach, FMIPv6. As intended, this approach shaped the output and produced a better network performance. Based on the comparative approach taken in the research methodology, results from objective 3 were straightforwardly presented and the comparison was directly established responding to the specific objective 4. MIPv6 and Client-Server Mobility Problems Specific objective 1 in this research was to evaluate research problem through literature review and in response, propose the research solution framework. This specific objective has been met. The research study reached the milestone of making possible the concept of integration of Mobile IPv6 technology with client-server architectures in parallelism with other mobile technologies and telecommunication networks such as 3G/UMTS and 4G/LTE, and 5G/IoT for their abilities to seamlessly provide internet connection to mobile users. As stated in Section 1.2, the problem the research sought to solve had its catalyst in network coverage limitations of WLANs, security threats due to wireless network volatility in Mobile Networks, and as well the willingness to solve the shortcomings of IPv4 technology as it relates to the extinction of its IP addresses range, its inability to provide smooth mobile capabilities to Mobile users, and its rather unreliable security capabilities and lack of autonomous route optimization process. This objective allowed us to determine more optimized techniques that could conceivably and properly be applied in MIPv6 to ensure security and better network performance. Therefore, these techniques included bidirectional tunnel of information between the MN and the HA, which not only allowed a secure line of communication between the nodes, but also provided the CN ability to contact the MN through its HoA since it had no knowledge of the new MN’s CoA. Another introduced and applied technique was Route Optimization. RO process was later decided between the MN (client) and CN (server), allowing them to directly share information without the intermediary of the HA. This process as well introduced security various risks, but another security procedure, Return Routability Procedure, was put in place to take care of some potential attacks to the network nodes (client/server) by encrypting nodes’ IDs, signaling and binding update messages. Finally, this research showed that due to the potential projected and technological promises of MIPv6, some important and future technology innovations such as IoT, and as to this project, client-server, MIPv6 could usefully help and expand capabilities in nodes and their interactions. Design and Implementation of Client-Server MIPv6 The second specific objective touched on the design and implementation of the actual MIPv6 technology as proposed in a client-server architecture. This objective has been met. To perform the implementation of the proposed solution, the research methodology suggested a simulation environment, OMNET++ which was the most well-adapted simulation environment and could as well accommodate INET framework in extension, which contained all needed libraries and modules vital to the implementation of the technology in a host-based mobility setting. Based on the proposed architecture design which presented a network with one client, one server and the HA and FA, the project implementation proceeded through simulation of the topology that was adopted to both MIPv6 and client-server architectural provisions. The research proposed MIPv6 solution introduced a critical handover latency that sparked an analysis of the handover delay and introduction of measures that could influence the delay and consequently improve on the network performance. This process triggered the implementation of a faster handover process, FMIPv6. Implementation includes services such as FTP, HTTP and Video Stream that became helpful along with TCP and UDP protocols in the implementation process. Finally, to establish performance differences according to the proposed methodology, metrics such as throughput, Packet Error Rate (PER), Packet Loss Rate (PLR), Handover and Packets end-to-end delays were measured. Evaluation of Network QoS of Client-Server MIPv6 The third objective was met, and it advanced the necessity to evaluate the implemented MIPv6 technology measuring network performance of the solution using various service models such as FTP, HTTP, and Video Stream, while ensuring network security and more optimized routing process. This solution provided technology abilities to a client-server technology to seamlessly manage and connect with the clients as they move and attach to other IPv6-based networks. However, the handover delay was still critical that it could potentially be damaging for real-time and availability demanding application services such as Video Stream. Therefore, If we implemented a technology that could reduce the delay time, it would ensure a better network performance. Evaluation of Client-Server FMIPv6 for Better QoS To respond the handover delay concern, and meet the specific objective 4 of the research, Fast Handover MIPv6 technology was explore and implemented using the same client-server architecture since it is an extension of MIPv6 proposed by IETF research group. The fourth objective was to apply technical measures to improve the shortcomings of the implemented MIPv6 by introducing fast handover techniques, then evaluate the solution for better QoS as opposed to the performance of the proposed MIPv6 solution. Finally, all simulation outputs were measured comparatively between MIPv6 and Fast MIPv6 solutions and allowed us to draw the ultimate conclusion of a better network performance for FMIPv6 approach than MIPv6-based network implementations. Differences could be seen as the performances were displayed based on respective metrics including Throughput, PER, PLR, Handover and end-to-end delays, which assumed a conclusive better network Quality of Services. Recommendations for the Project Since the solution is implemented in a client-server architecture, it is important to implement some technologies such as Route Optimization and IPsec along with MIPv6 for not only optimizing the packet routes, but also to insure secure communications as the client is moving toward unknown networks. On the other hand, based on the project conclusion that established a real difference in performance with Fast Handover MIPv6 producing a better level than the standard MIPv6, it is then encouraged to take the Fast MIPv6 approach, especially in client-server architecture. This is ultimately based on the criticality of delays in services and utility of availed services even when the clients are moving across other networks. Recommendations and Future Work This research contains a broad spectrum of networking technologies by integrating well-known technology approaches, MIPv6 and Client-Server. The implementation and integration of both technologies introduced a systematic excavation of contributing technologies in the explicit measure of their complexities and depths. While conducting the research, numerous avenues for further research became evident. Therefore, an elaborated description of options associated to future work is outlined below: The proposed MIPv6 architecture does not consider use of more than one HA. This may increase security and service availability issues in case of disaster occurrence since it represents a single point of failures. So, more HA can possibly be securely added and synchronized with the MN to increase availability posture and prevent fatal security breaches. Simulation was carried out in a homogeneous environment where only Wi-Fi IEEE 802.11b standard was used. However, other access technologies like WiMAX, LTE, and son on, could be used to test network handover latency influence on the overall performance of both MIPv6 and FMIPv6 technology schemes. Given the mobility decision being taken by the client or MN in accordance to the network environment provisions with Wi-Fi standard, it would be possible to do the same with WiMAX or LTE standards. In INET Framework, it would be more effect to introduce and develop a whole source code for the entire and sole implementation of FMIPv6 in order to clearly produce and respond to technology requirements as to those suggested by IETF. Thus, then manage to substantially improve MIPv6 network performance. In MIPv6 handover, RO and binding update processes, it is extremely critical to protect the Server or CN based on is status in the entire network as it may contain important and classified data, personal and public information on the server and the network itself. Through these procedures is exposed to a lot of potential security attacks as its IP address is unique and cannot be changed. Developing an environment where even the server of CN is mobile and can change location and IP attachments, it may furtherly protect information on the network and avoid disastrous events. References Abukharis, S., Alzubi, J. A., Alzubi, O. A., & Alamri, S. (2014). Packet error rate performance of IEEE802.11g under bluetooth interface. Research Journal of Applied Sciences, Engineering and Technology, 8(12), 1419–1423. Al-adhal, Khaled Mahmood, S. S. T. (2012). Mobile IP?: A Study Of Issus , Challenges , And Comparison Of IPv4 & IPv6. International Journal of Engineering Research and Applications (IJERA), 2(6), 616–621. Al-Kasasbeh, B. M., Al-Qutaish, R. E., & Al-Sarayreh, K. T. (2008). Indirect Routing of Mobile IP?: A Non-Encapsulation Approach. International Journal of Computer Science and Network Security, 8(7), 124–131. Al-Rubaye, A., Aguirre, A., & Seitz, J. (2016). Enabling Soft Vertical Handover for MIPv6 in OMNeT++. Retrieved from Albrecht, M. (2010). Introduction to Discrete Event Simulation. Proceedings of the 2^{nd} DYCOMANS Workshop on “Management and Control: Tools in Action”, (January), 367–376. Retrieved from to DES.pdf Ali, A. N. A. (2012). Comparison study between IPV4 & IPV6. International Journal of Computer Science Issues, 9(3), 1–4. Retrieved from Ali, Z., & Bhaskar, S. B. (2016). Basic statistical tools in research and data analysis. Indian Journal of Anaesthesia, 60(9), 662–669. Alomari, S. A., Sumari, P., & Taghizadeh, A. (2011). A comprehensive study of wireless communication technology for the future mobile devices. European Journal of Scientific Research, 60(4), 583–591. Amin, M. A., Bakar, K. B. A., Abdullah, A. H., & Khokhar, R. H. (2011). Reducing handover latency in mobile IPv6-based WLAN by parallel signal execution at layer 2 and layer 3. Communications in Computer and Information Science, 154 CCIS, 201–211. Aristomenopoulos, G., Bouras, C., & Stamos, K. (2008). Performance study of the Mobile IPv6 protocol and its variations. 2008 6th International Symposium on Communication Systems, Networks and Digital Signal Processing, 438–442. Arvind, T. (2016). A Comparative Study of Various Network Simulation Tools. International Journal of Computer Science and Engineering Technology (IJCSET), 7(08), 374–378. Babatunde, O., & Al-debagy, O. (2014). A Comparative Review of Internet Protocol Version 4 (IPv4) and Internet Protocol Version 6 (IPv6). International Journal of Computer Trends and Tecnology, 13(1), 10–13. Bai, G., & Williamson, C. (2004). The effects of mobility on wireless media streaming performance. Proceedings of Wireless Networks and Emerging Technologies (WNET), 596–601. Retrieved from Baig, Z. A., & Adeniye, S. C. (2011). A trust-based mechanism for protecting IPv6 networks against stateless address auto-configuration attacks. ICON 2011 – 17th IEEE International Conference on Networks, 171–176. Banerji, S., & Chowdhury, R. S. (2013). On IEEE 802.11: Wireless Lan Technology. International Journal of Mobile Network Communications & Telematics, 3(4), 45–64. Bilal, S. M., Khan, A. R., & Othman, M. (2012). A performance comparison of open source network simulators for wireless networks. In 2012 IEEE International Conference on Control System, Computing and Engineering (pp. 34–38). Branco, K., Pinto, A., & Pigatto, D. (2017). Communication in Critical Embedded Systems. Communications in Computer and Information Science, 702, 1–131. Brotherton, B. (2008). Researching Hospitality and Tourism: A Student Guide. Caicedo, C. E., Joshi, J. B. D., & Tuladhar, S. R. (2009). IPv6 security challenges. IEEE Computer Society, 42(2), 36–42. Céspedes, S., & Sherman, S. (2015). On Achieving Seamless IP Communications in Heterogeneous Vehicular Networks, 16, 1–15. Çetin, G., & Çetin, A. (2015). Mugla Journal of Science and Technology THE ANALYSIS OF LAYER-2 HANDOVER PERFORMANCE FOR MOBILE IPV6 USING OMNeT ++ SIMULATION TOOL OMNeT ++ BENZET?M ARACI KULLANILARAK MOB?L IPV6 KATMAN-2 HÜCRE, 1(1), 34–38. Chen, J.-L., Lee, Y.-F., & Chang, Y.-C. (2014). Mobile IPv6 network: implementation and application. Internation Journal of Network Management, (October 2005), 17–31. Choi, S. Il, & Koh, S. J. (2016). Use of Proxy Mobile IPv6 for Mobility Management in CoAP-Based Internet-of-Things Networks. IEEE Communications Letters, 20(11), 2284–2287. Choudhary, A. R., & Sekelsky, A. (2010). Securing IPv6 network infrastructure: A new security model. 2010 IEEE International Conference on Technologies for Homeland Security, HST 2010, 500–506. Chun, S.-M., Kim, H.-S., & Park, J.-T. (2015). CoAP-Based Mobility Management for the Internet of Things. Sensors (Basel, Switzerland), 15(7), 16060–16082. Cisco systems, I. (2012). Implementing Mobile IPv6. Cooper, D. R., & SChindler, P. S. (2011). Business research methods/Donald R. Cooper, Pamela S. SChindler (11th ed.). McGraw-Hill. Retrieved from;jsessionid=16FD0E739867E3D2AF152639FB6E5F9E?cc=at&lang=en& D. Rudolf. (2009). Next Generation Internet?: IPv4 Address Exhaustion, Mitigation Strategies and Implications for the U.S. Ieee Usa, 1–26. Retrieved from Devarapalli, V., Wakikawa, R., Petrescu, A., & Thubert, P. (2005). Network Mobility (NEMO) Basic Support Protocol. Ebalard, A. (2010). Mobile IPv6 IPsec Route Optimization (IRO) draft-ebalard-mext-ipsec-ro-02. Working Draft,IETF Secretariat, Internet-Draft Draft-Ebalardmext-Ipsec-Ro-02, (July). Retrieved from Fajar, A., Sarno, R., & Fauzan, A. C. (2018). Comparison of Discrete Event Simulation and Agent Based Simulation for Evaluating the Performance of Port Container Terminal. In 2018 International Conference on Information and Communications Technology (ICOIACT) (pp. 259–265). Retrieved from Fei, Z. S., Xing, C. W., & Li, N. (2014). QoE-driven resource allocation for mobile IP services in wireless network. Science China Information Sciences, 58(1), 1–10. Fu, S., & Atiquzzaman, M. (2005). Handover latency comparison of SIGMA, FMIPv6, HMIPv6, and FHMIPv6. In GLOBECOM – IEEE Global Telecommunications Conference (Vol. 6, pp. 3809–3813). Ganz, F., Li, R., Barnaghi, P., & Harai, H. (2012). A resource mobility scheme for service-continuity in the Internet of Things. Proceedings – 2012 IEEE Int. Conf. on Green Computing and Communications, GreenCom 2012, Conf. on Internet of Things, IThings 2012 and Conf. on Cyber, Physical and Social Computing, CPSCom 2012, 261–264. Goralski, W. (2014). Learn About Differences in Addressing Between IPv4 and IPv6. Juniper Networks, 1–11. Retrieved from Goudru, N. G., & Vijayakumar, B. P. (2016). Evaluation of TCP Performance Using NDG Loss Predictor in Wireless Networks. Proceedings – 2015 21st Annual International Conference on Advanced Computing and Communications, ADCOM 2015, 16–21. Hanumanthappa, J., & Annaiah, H. (2006). DW&C:Dollops Wise Curtail IPv4/IPv6 Transition Mechanism using NS2 . 1 1. Haraty, M., Mohammadi, A., & Mehbodnia, M. (2014). Estimation of SINR Distribution Function in Clustered Wireless Sensor Networks Using Statistical Distribution Modeling. Telecommunications (IST), 2014 7th International Symposium On, 740–745. Hasan, A., & Al, R. (2014). A Survey?: MOBILE IPV4 / IPV6 Fundamentals , Advantages and Disadvantages. International Journal of Advanced Research in Computer Science and Software Engineering, 4(4), 874–877. Retrieved from Hasan, S. S., & Hassan, R. (2013). Enhancement of return routability mechanism for optimized-nemo using correspondent firewall. ETRI Journal, 35(1), 41–50. Heydari, V., Kim, S., & Yoo, S.-M. (2016). Secure VPN using Mobile IPv6 based Moving Target. EEE, 7:1-7:8. Hung, T. C., & Duong, V. T. T. (2011). Mobile IPv6 fast handover techniques. 13th International Conference on Advanced Communication Technology (ICACT2011), 2–4. Hussien, L. F., Aisha-Hassan, A. H., Anwar, F., Mahmoud, O., Zakaria, O., & Saeed, R. A. (2010). Design of robust protocol to enhance QoS in mobile IPV6 environment. In International Conference on Computer and Communication Engineering, ICCCE’10 (pp. 11–13). Hyun, J., Li, J., Kim, H., Yoo, J. H., & Hong, J. W. K. (2015). IPv4 and IPv6 performance comparison in IPv6 LTE network. 17th Asia-Pacific Network Operations and Management Symposium: Managing a Very Connected World, APNOMS 2015, 1–6. Islam, S., Abdalla, A. H., Khalifa, M. K. H. O. O., Mahmoud, O., & Saeed, R. A. (2012). Macro mobility scheme in NEMO to support seamless handoff. 2012 International Conference on Computer and Communication Engineering, ICCCE 2012, (July), 234–238. Issariyakul, T., & Hossain, E. (2012). Introduction to network simulator NS2. Introduction to Network Simulator NS2 (Vol. 9781461414). Iványi, A. (2005). ALGORITHMS OF INFORMATICS Vol. 2. (Vol. 2). Kempelen Farkas Hallgatói Információs Központ. Jakubiak, J., & Koucheryavy, Y. (2007). Precise Delay Analysis for IEEE 802.11 Legacy Ad-Hoc Networks. In 2007 IEEE 65th Vehicular Technology Conference – VTC2007-Spring (pp. 2956–2960). Jara, A. J., Fernandez, D., Lopez, P., Zamora, M. A., & Skarmeta, A. F. (2013). Lightweight mobile IPv6: A mobility protocol for enabling transparent IPv6 mobility in the internet of things. GLOBECOM – IEEE Global Telecommunications Conference, 2791–2797. Jara, A. J., Fernandez, D., Lopez, P., Zamora, M. A., & Skarmeta, A. F. (2014). Lightweight MIPv6 with IPSec support. Mobile Information Systems, 10(1), 37–77. Jara, A. J., Zamora, M. a, & Skarmeta, A. F. G. (2010). An Initial Approach to Support Mobility in Hospital Wireless Sensor Networks based on 6LoWPAN ( HWSN6 ). Journal of the Wireless Mobile Networks, Ubiquitous Computing, and Dependable Applications, 1, 107–122. Jauhari, A. S., & Kistijantoro, A. I. (2015). INET Framework modifications in OMNeT++ simulator for MPLS traffic engineering. In Proceedings – 2014 International Conference on Advanced Informatics: Concept, Theory and Application, ICAICTA 2014 (pp. 87–92). Johnson, D., & Arkko, J. (2011). Mobility Support in IPv6. Internet Engineering Task Force (IETF). Jun, L., Liyan, Z., & Pierre, S. (2009). Intelligent Fast Handover Scheme for Mobile IPv6-based Wireless Local Area Networks. Ijcsns, 9(8), 60–64. Retrieved from Jung, Y., & Peradilla, M. (2013). Host mobility management using combined MIPv6 and DNS for MANETs. 2013 International Conference on Selected Topics in Mobile and Wireless Networking, MoWNeT 2013, 100–105. Kavitha, D., Sreenivasa Murthy, K. E., & Zahoor Ul Huq, S. (2010). Security analysis of binding update protocols in route optimization of MIPv6. In ITC 2010 – 2010 International Conference on Recent Trends in Information, Telecommunication, and Computing (pp. 44–49). Khan, S. (2010). Towards a reliable seamless mobility support in heterogeneous IP networks, (January). Retrieved from Kim, S., & Kim, D. Y. (2006). Home Address allocation to mobile terminal over mobile IP environments, (0). Kim, Y., Kwon, D., Bae, K., & Suh, Y. (2005). Performance Comparison of Mobile IPv6 and Fast Handovers for Mobile IPv6 over Wireless LANs, 807–811. Koodli, R. (2009a). Mobile IPv6 Fast Handover. Koodli, R. (2009b). Mobile IPv6 Fast Handovers. Koodli, R. (2010). Mobile IPv6 Fast Handovers. Kukhmay, Y., Glasman, K., Peregudov, a., & Logunov, a. (2006). Video Over IP Networks: Subjective Assessment of Packet Loss. 2006 IEEE International Symposium on Consumer Electronics. Kwon, D. H., Kim, Y. S., Bae, K. J., & Suh, Y. J. (2010). Access router information protocol with FMIPv6 for efficient handovers and their implementations. GLOBECOM – IEEE Global Telecommunications Conference (Vol. 6). Lai, J., Sekercioglu, Y. A., Jordan, N., & Pitsillides, A. (2006). Performance Evaluation of Mobile IPv6 Handover Extensions in an IEEE 802 . 11b, 6, 1–8. Lai, J., Wu, E., Varga, A., Sekercioglu, Y. A., & Egan, K. G. (2002). A Simulation Suite for Accurate Modeling of {IPv6} Protocols. Proceedings of the 2nd International OMNeT++ Workshop, (January), 2–22. Le, D., & Chang, J. (2010). Tunnelling-Based Route Optimization for Mobile, 509–513. León, O., Hernández-Serrano, J., & Soriano, M. (2010). Securing cognitive radio networks. International Journal of Communication Systems, 23(5), 633–652. Lin, P., Cheng, S. M., & Liao, W. (2009). Modeling key caching for mobile IP authentication, authorization, and accounting (AAA) services. IEEE Transactions on Vehicular Technology, 58(7), 1–3. Liu, Y., & Yang, C. (2011). OMNeT ++ based modeling and simulation of the IEEE 1588 PTP clock. 2011 International Conference on Electrical and Control Engineering, ICECE 2011 – Proceedings, 4602–4605. Loehnert, S. (2010). About statistical analysis of qualitative survey data. International Journal of Quality, Statistics, and Reliability, 2010. Lytchkina, N. (2009). Simulation modeling of regions’ social and economic development in decision support systems . Proceedings of the 27th International Conference of the System Dynamics Society. Ma, D., & Ma, M. (2014). Network selection and resource allocation for multicast in HetNets. Journal of Network and Computer Applications, 43, 17–26. Madhuri, D., & Reddy, D. C. (2016). Performance Comparison of TCP, UDP and SCTP in a Wired Network. 2016 International Conference on Communication and Electronics Systems (ICCES), (October). Maidstone, R. (2012). Discrete Event Simulation, System Dynamics and Agent Based Simulation: Discussion and Comparison. System, 1–6. Matzer, H., & Levy, S. (2004). Antenna Gain, 1–6. Retrieved from Mayer, C., & Gamer, T. (2008). Integrating real world applications into OMNeT++. Institute of Telematics, University of Karlsruhe, …. Retrieved from McCarthy, B., Jakeman, M., & Edwards, C. (2009). Supporting nested NEMO networks with the unified MANEMO architecture. Proceedings – Conference on Local Computer Networks, LCN, (October), 609–616. MobiWorld 2010. (2010). The 2nd International Workshop on Mobile IPv6 and Network-based Localized Mobility Management. Las Vegas. Retrieved from Modares, H., Moravejosharieh, A., Salleh, R. Bin, & Lloret, J. (2014). Enhancing Security in Mobile IPv6. ETRI Journal, 36(1), 51–61. Moravejosharieh, A., Modares, H., & Salleh, R. (2012). Overview of Mobile IPv6 security. Proceedings – 3rd International Conference on Intelligent Systems Modelling and Simulation, ISMS 2012, 1–3. Mrugalski, T., Wozniak, J., & Nowicki, K. (2010). Remote stateful autoconfiguration for mobile IPv6 nodes with server side duplicate address detection. 2010 Australasian Telecommunication Networks and Applications Conference, ATNAC 2010, 141–146. Mühleisen, M., Jennen, R., & Kirsche, M. (2010). Wireless Networking Use Cases. In Modeling and Tools for Network Simulation (pp. 305–325). Springer, Berlin, Heidelberg. Networks, F. (2006). OPENT Technologies, Inc. Managing network and application performance. Nikander, P., Arkko, J., Aura, T., Montenegro, G., & Nordmark, E. (2005). Mobile IP Version 6 Route Optimization Security Design Background, 6, 2004–2008. OMNeT++. (2012). INET Framework for OMNeT++, 157. Opoku, S. K. (2012). A Simultaneous-Movement Mobile Multiplayer Game Design Based on Adaptive Background Partitioning Technique, 2–4. Orooji, M., Soltanmohammadi, E., & Naraghi-Pour, M. (2013). EVALUATING THE EFFECTS OF CO-CHANNEL INTERFERENCE IN WIRELESS NETWORKS. Division of Electrical and Computer Engineering School of Electrical Engineering and Computer Science Louisiana State University , Baton Rouge , LA 70803, 4693–4697. Retrieved from Park, M., Lee, J., Kim, D., Baek, H., Lim, J., & Choi, H. (2013). A distributed dynamic address assignment scheme for tactical mobile ad hoc networks. Proceedings – IEEE Military Communications Conference MILCOM, 2–5. Perkins, C. E. (2002). Mobile IP and the IETF Mobile Networking at IETF 45, 3(3), 4–7. Perkins, C., Johnson, D., & Arkko, J. (2011). Mobility Support in IPv6. Internet Engineering Task Force, (July), 1–169. Retrieved from Phoomikiattisak, D., & Bhatti, S. N. (2013). Network Layer Soft Handoff for IP Mobility, 6, 3–5. Phoomikiattisak, D., & Bhatti, S. N. (2015). Mobility as a first class function. 2015 IEEE 11th International Conference on Wireless and Mobile Computing, Networking and Communications, WiMob 2015, 850–859. Pieterse, J., Wolhuter, R., & Mitton, N. (2013). Implementation and analysis of FMIPv6, an enhancement of MIPv6. Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST, 111, 351–364. Punithavathani, D. Shalini & Sankaranarayanan, K. (2009). IPv4 / IPv6 Transition Mechanisms. EuroJournals Publishing, Inc., 34(1), 110–124. Retrieved from Radhakrishnan, R., Jamil, M., & Mehfuz, S. (2008). A Robust Return Routability Procedure for Mobile IPv6, 8(5), 234–240. Ratola, M. (2004). Which Layer for Mobility?? – Comparing Mobile IPv6 , HIP and SCTP. Seminar on Internetworking, 1–9. Ravi, N., A, M. S., & Periyasamy, M. (2014). Implementation of IPv6 / IPv4 Dual-Stack Transition Mechanism, 5(5), 6326–6332. Sailan, M. K., Hassan, R., & Patel, A. (2009). A Comparative Review of IPv4 and IPv6 for Research Test Bed. 2009 International Conference on Electrical Engineering and Informatics, 2(August), 1–6. Sanguankotchakorn, T., & Jaiton, P. (2008). Effect of triangular routing in mixed IPv4/IPv6 networks. Proceedings – 7th International Conference on Networking, ICN 2008, (May 2008), 357–362. Sarkar, N. I., & Membarth, R. (2015). Modeling and Simulation of IEEE 802 . 11g using OMNeT ++. International Journal, (July), 379–381. Sha, F., Chen, X., Ye, R., Wu, M., Zhang, Z., & Cai, W. (2017). A communication method between IPv4 server and IPv6 network in virtual machine environment. Proceedings – 16th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2017, 2, 885–888. Sheikh, M. A., Singh, N., & Afroz, S. (2016). A study of Quality of Service in Network Mobility (NEMO) using Qualnet 7.0. In 2016 International Conference on Computing for Sustainable Global Development (INDIACom) (Vol. 1, pp. 1–5). Sheikhahmadi, A., Nematbakhsh, M. A., & Shokrollahi, A. (2015). Improving detection of influential nodes in complex networks. Physica A: Statistical Mechanics and Its Applications, 436, 833–845. Shih, C.-H., & Chen, Y.-C. (2010). A FMIPv6 Based Handover Scheme for Real-Time Applications in Mobile WiMAX. Journal of Networks, 5(8), 929–936. Siebers, P. O., MacAl, C. M., Garnett, J., Buxton, D., & Pidd, M. (2010). Discrete-event simulation is dead, long live agent-based simulation! Journal of Simulation, 4(3), 204–210. Sookun, Y., & Bassoo, V. (2016). Performance analysis of IPv4/IPv6 transition techniques. 2016 IEEE International Conference on Emerging Technologies and Innovative Business Practices for the Transformation of Societies, EmergiTech 2016, 188–193. Suárez, F. J., Nuño, P., Granda, J. C., & García, D. F. (2015). Computer networks performance modeling and simulation. Modeling and Simulation of Computer Networks and Systems: Methodologies and Applications. Tramarin, F., Vitturi, S., & Luvisotto, M. (2017). A dynamic rate selection algorithm for IEEE 802.11 industrial wireless LAN. IEEE Transactions on Industrial Informatics, 13(2), 846–855. Tripathi, A. K., Radhakrishnan, R., & Lather, J. S. (2012). Factors Affecting the Handover Latency in MIPv6. Special Issue of International Journal of Computer Applications (0975 – 8887), (November), 32–36. Varga, A., & Hornig, R. (2008). AN OVERVIEW OF THE OMNeT++ SIMULATION ENVIRONMENT. Proceedings of the First International ICST Conference on Simulation Tools and Techniques for Communications Networks and Systems. Varga, A., & OpenSim. (2016). OMNeT++ Installation Guide. Retrieved from Vassiliou, V., & Zinonos, Z. (2010). An Analysis of the Handover Latency Components in Mobile IPv6. Journal of Internet Engineering, 230–240. Retrieved from Waehlisch, M., Koodli, R., Fairhurst, G., & Liu, D. (2014). Multicast Listener Extensions for Mobile IPv6 (MIPv6) and Proxy Mobile IPv6 (PMIPv6) Fast Handovers. World Telecommunication/ICT Policy Forum. (2013). IPv4 and IPv6 issues. International Telecommunication Union, (May), 14–16. Wozniak, J. (2016). Mobility management solutions for current IP and future networks. Telecommunication Systems, 61(2), 257–275. Xian, X., Shi, W., & Huang, H. (2008). Comparison of OMNET++ and other simulator for WSN simulation. 2008 3rd IEEE Conference on Industrial Electronics and Applications, ICIEA 2008, 1439–1443. Xianhua, J., & Sui, D. (2011). Study of IPv6 Security Technology. The Proceeding of 2011 International Conference on Network and Electronic Engineering. Xiaorong, F., Jun, L., & Shizhun, J. (2013). The research on mobile Ipv6 security features. IEEE Symposium on Wireless Technology and Applications, ISWTA, 125–128. Xie, J., & Narayanan, U. (2010). Performance Analysis of Mobility Support in IPv4/IPv6 Mixed Wireless Networks. IEEE Transactions on Vehicular Technology, 59(2), 962–973. Retrieved from Yang, G., Lei, Z., Wang, H., Dou, Y., & Xie, Y. (2014). Analysis of the RADIUS signalling based on CDMA mobile network. Proceedings – 2014 6th International Conference on Intelligent Human-Machine Systems and Cybernetics, IHMSC 2014, 2, 270–274. Yang, S.-J., & Chen, S.-U. (2010). QoS-Based Fast Handover Scheme for Improving Service Continuity in MIPv6. 2010 Ieee International Conference on Wireless Communications, Networking and Information Security (Wcnis), Vol 2, 403–408. Yousaf, F. Z., Bauer, C., & Wietfeld, C. (2008). An Accurate and Extensible Mobile IPv6 (xMIPV6) Simulation Model for OMNeT++. OMNeT++, 6(October). Yousaf, F. Z., Müller, C., & Wietfeld, C. (2010). A comprehensive MIPv6 based mobility management simulation engine for the next generation network. Proceedings of the 3rd International ICST Conference on Simulation Tools and Techniques, 16. Zaidi, M., Bhar, J., Ouni, R., & Tourki, R. (2011). Reducing Wi-Fi handover delay using a new positioning process. In 2011 International Conference on Communications, Computing and Control Applications, CCCA 2011. Zhang, D.-Y., Hu, M.-Z., Zhang, H.-L., & Kang, T.-B. (2005). The research on metrics for network performance evaluation. 2005 International Conference on Machine Learning and Cybernetics, 2(August), 1127–1131 Vol. 2. Zhang, L. J., & Pierre, S. (2012). Mobility Support for IPv6-based Next Generation Wireless Networks: A Survey of Related Protocols. Int. J. Wirel. Netw. Broadband Technol., 2(3), 18–41. Zhang, N., & Bao, H. (2008). Study on Mobile IP Technology in Wireless Communication Systems. In ICWMMN200B Proceedings (pp. 1–4). Beijing, China. Zheng, X., & Sarikaya, B. (2009). Handover keying and its uses. IEEE Network, 23(2), 1–8. Zhu, Z., Wakikawa, R., & Zhang, L. (2011a). A Survey of Mobility Support in the Internet. Zhu, Z., Wakikawa, R., & Zhang, L. (2011b). A Survey of Mobility Support in the Internet. APPENDIX A: MOBILE IPV6 .NED SOURCE CODE package inet.examples.mipv6ClientServer; import inet.applications.ethernet.EtherAppCli; import inet.applications.httptools.configurator.HttpController; import inet.applications.httptools.server.HttpServer; import inet.common.geometry.common.OsgGeographicCoordinateSystem; import inet.environment.common.PhysicalEnvironment; import inet.examples.adhoc.hostautoconf.Host; import inet.examples.mobility.MobileHost; import inet.examples.wireless.ratecontrol.Client; import inet.linklayer.ethernet.EtherHub; import inet.linklayer.ieee80211.Ieee80211Nic; import inet.mobility.static.StationaryMobility; import inet.networklayer.configurator.ipv4.IPv4NetworkConfigurator; import inet.networklayer.configurator.ipv6.FlatNetworkConfigurator6; import inet.networklayer.icmpv6.IPv6NeighbourDiscovery; import inet.networklayer.ipv6.IPv6; import inet.networklayer.ipv6.IPv6RoutingTable; import inet.networklayer.multi.MultiRoutingTable; import inet.node.ethernet.Eth1G; import inet.node.httptools.DirectHost; import inet.node.inet.AdhocHost; import inet.node.inet.Router; import inet.node.inet.WirelessHost; import inet.node.internetcloud.InternetCloud; import inet.node.ipv6.Router6; import inet.node.ipv6.StandardHost6; import inet.node.wireless.AccessPoint; import inet.node.xmipv6.CorrespondentNode6; import inet.node.xmipv6.HomeAgent6; import inet.node.xmipv6.MobileHost6; import inet.node.xmipv6.WirelessHost6; import inet.physicallayer.antenna.DipoleAntenna; import inet.physicallayer.base.packetlevel.AntennaBase; import inet.physicallayer.contract.packetlevel.IRadioMedium; import inet.physicallayer.contract.packetlevel.IReceiver; import inet.physicallayer.ieee80211.packetlevel.Ieee80211ScalarRadioMedium; import inet.physicallayer.pathloss.BreakpointPathLoss; import inet.transportlayer.sctp.SCTPNatRouter; import inet.visualizer.common.InfoOsgVisualizer; import inet.visualizer.common.InfoVisualizer; import inet.visualizer.contract.IIntegratedVisualizer; import inet.visualizer.contract.IInterfaceTableVisualizer; import inet.visualizer.integrated.IntegratedVisualizer; import inet.visualizer.power.EnergyStorageOsgVisualizer; import ned.DatarateChannel; channel fiberline extends ned.DatarateChannel { parameters: delay = 1us; datarate = 1Gbps; } channel ethernetline extends ned.DatarateChannel { parameters: delay = 0.1us; datarate = 100Mbps; } network mipv6ClientServer { parameters: double total_mn; double total_cn; // **.mgmt.numChannels = 5; @display(“bgb=1111.8151,891.405”); types: submodules: visualizer: like IIntegratedVisualizer if hasVisualizer() { parameters: @display(“p=55.49,53.7”); } radioMedium: Ieee80211ScalarRadioMedium { parameters: @display(“p=171.84001,53.7;is=s”); } Home_Agent: HomeAgent6 { parameters: @display(“p=334,383”); } servertotal_cn: CorrespondentNode6 { parameters: @display(“p=560.27,805.50006;i=device/server;is=n”); } clienttotal_mn: WirelessHost6 { parameters: @display(“p=85.920006,232.70001”); } AP_Home: AccessPoint { parameters: @display(“p=334.73,288.19”); } FA: Router6 { parameters: @display(“p=741.06006,358”); } AP_FA: AccessPoint { parameters: @display(“p=741,263”); } IPv6Router: Router6 { parameters: @display(“p=560.27,562.06”); } hub: EtherHub { parameters: @display(“p=560.27,683.78”); } flatNetworkConfigurator6: FlatNetworkConfigurator6 { @display(“p=404.54,53.7”); } connections allowunconnected: FA.ethg++ ethernetline IPv6Router.ethg++; Home_Agent.ethg$o++ –> ethernetline –> IPv6Router.ethg$i++; Home_Agent.ethg$i++ hub.ethg$i++; serveri.ethg$i++ Home_Agent.ethg$i++; AP_Home.ethg$i++


Remember. This is just a sample

You can get your custom paper from our expert writers

Get custom paper