| The evolution of wireless devices, such as laptops, personal digital assistants (PDAs), and cell phones, has changed the way networking and communication are perceived. The growing dependence on these devices, along with the high mobility of users, has increased the need to be connected in all places at all times. This increase in user demand, along with the evolution of new applications such as satellite networks, sparse mobile networks, and remote disconnected communities, has ultimately led to the rise of the delay and disruption tolerant networks (DTNs) research area.; The future of computer networks includes an Internet that encompasses numerous heterogeneous networks, reaches out to the most remote areas, and provides communication in the most extreme and unstable conditions. With such a vision, new problems with large complexities arise. Communication devices in the future will be required to remain connected despite the rise of new challenges such as network partitioning, intermittent connectivity, large delays, the high cost of infrastructure deployment, and the absence of an end-to-end path.; This dissertation addresses some of the major problems that arise, as a result of these challenges, in delay and disruption tolerant networks (DTNs). We study several protocol and architectural challenges in DTNs. We specifically introduce and study a new class of DTNs, known as Delay Tolerant Mobile Networks (DTMNs). We propose various controlled flooding schemes in DTMNs, and investigate the performance of these schemes over different DTMN architectures. Additionally, we study the dedicated messenger ownership schemes and scheduling strategies in clustered DTMNs. Furthermore, we address transport layer issues, particularly reliability, in DTN environments. We present a solution for mobile devices experiencing opportunistic intermittent connectivity. Finally, we propose a novel architecture, ParaNets, that encompasses our vision of how future DTN solutions will be built. |
