Browsing by Author "Ahmed, Muhammed Zaharadeen"

Nowadays the need for a reliable Internet access across the globe is very crucial because of the huge size of useful data it stores. This stimulates most ISPs to choose a reliable routing protocol such as Border Gateway Protocol in order to ensure guarantee of customer's data on the Internet. By multi-homing a network using BGP, a more control of logical division of the traffic load and achieving failover becomes possible. This enhances network performance, throughput and provide better routing policies on the Internet. Single-homed Internet communication link for an ISP networks does not inspire with confidence because of excessive traffic and device/human errors that may arise. BGP is used because of its ability to choose the best path to a destination particularly in a Multihoming environment. Link fault tolerance is highlighted on its challenge to Multihoming network. Simulation of BGP network was carried out on single homing where the result presents limited reliability in guarantee of customer's data. Alternatively, two independent simulations were carried out to run BGP Multihoming to the same Internet service provider and to different Internet service provider in other to provide an improvement. In BGP Multihoming to the same Internet Service Provider, a failover technique was introduced to providing realizing reliability on the network while the technique of traffic division was introduced in realizing scalability. Also, in simulating BGP Multihoming to different ISP, the technique of providing failover was ensured by realizing reliability while network analysis was carried out to ensure the throughput in the network. In ensuring the throughput, 6 routers were benchmarked in AS 100 to perform the analysis. At the end, simulation results were achieved using Graphical Network Simulator 3 emulator and Wireshark network analyzer. The simulation was carried out to show an improvement in reliability and scalability of the BGP network.

The problem of managing network mobility has been an open issue in network mobility management. This is because of the huge volume of multimedia contents across IP networks. The current IP architecture seems inadequate to continuously manage the ever-increasing number of mobile and network devices, that consumes and generate multimedia contents. This initiated and stimulated the concept of CCN/NDN to solve challenges faced in IP networks. Named Data Networking is a new content centric network architecture that can possibly tackle most issues of IP mobility and security. NDN approach is commonly identified under Information Concentric Network or Content Concentric Network and is centered on addressing contents by themselves using names, rather than assigning IP addresses to packets on hosts where information is located on the global internet. This thesis proposed a mobility management scheme with improved network availability and scalability for the mobile producer during and after inter-technology handoff. This is achieved by using rendezvous-based mobility management approach. For network availability, anchorless (locator-free) approach is used for producer’s connectivity during handoff for intra and inter access point change over. This ensures mobile producers’ registration and location update as it successfully achieves mobile handoff within an outside the network core. Forwarding paths and labels are also enhanced with minimal signaling cost and using best route approach to ensure easy transmission and reception of Interest and Data contents between mobile producers and rendezvous servers. For network scalability, simulation is used. The number of rendezvous server is increased for each simulation scenarios of pull, push, upload and shear to ensure traffic redistribution and by synchronizing the proposed scheme with scenario aware protocol. Each of the rendezvous server has an amount of data streams it can respond to in a network session when a mobile producer or consumer transmits its interest request. In the first scenario, four rendezvous nodes were deployed to implement the proposed number of data content streams in the NDN network. These data streams range between 100 to 1500 in each rendezvous node and is accessed multiple times to by producers and consumer for specified time interval. For the second scenario the number of rendezvous servers were increased to seven where number of data streams were maintained to 1500 in each rendezvous with maximum of 300 connection request per consumers and producers. For the third scenario, the number of consumers 10 and mobile producers are 15 to access 1500 maximum number of data streams in each rendezvous. In the final simulation link loss rate is at maximum of 10% and all other parameters were maintained constant based on first, second with 15% and third scenario with 23%. This can enable mobile producers to be able to retransmit their interest request in situation of total packet losses for at least three times. Based on the implementation, Algorithms are designed for network availability during and after successful mobile producer handoff, and for unique prefix name assignment. These are supported by an effective flowchart for clear description. For the results, ndnSIM2.1 is used in coding the scenarios based on random way point movement model and the analysis is simulated in Linux environment. Wireshark is interfaced to observe some basic parameters such as average throughput, round trip time,