Development and Deployment of Delay Tolerant Networks: An Arctic Village Case

Sammanfattning: In the late 1990s, NASA conducted a study of the Interplanetary Internet (IPN) architecture. In order to build and deploy IPN infrastructure, the network technology had to be able to cope with long radio signal propagation delays and frequent radio link disruptions. The concept of a Delay Tolerant Networking (DTN) emerged after recognizing that such a networking paradigm can also be applicable for terrestrial use. DTN technology can be applied, for instance, in disaster situations, military battlefields, economically developing areas, and remote regions. This thesis follows the process of applying DTN technology to a remote, communication-challenged area in the Arctic part of Sweden. The aim of the DTN deployments in the remote villages of Sarek and Padjelanta National Parks, between 2008 and 2011, was to provide a basic set of ICT services to the nomadic Sami population. Therefore, the research presented here acknowledges and considers the specific geographical, technical, and cultural conditions of theseareas, and how these conditions profoundly shaped the development of the deployed technology as well as the research methodology. As a result, this thesis makes scientific contributions to several research topics, spanning the fields of DTN routing, DTN service development, DTN evaluation methodologies, and ICT deployments. The first contribution in this thesis is the proposal of a new and improved version of the PRoPHETv2 routing protocol. The development of this routing protocol was driven by actual protocol use and the results of experiments conducted during the course of the DTN deployments. Secondly, this thesis proposes an alternative DTN routing objective for a typical remote village DTN scenario. Weaknesses of a conventional DTN routing research objective are exposed by outlining concrete geographical, social, and technical conditions discovered in DTN deployments on the field. When these conditions are overlooked, they can profoundly affect DTN deployments. Thirdly, this thesis discusses the development and deployment of the Not-So-Instant-Messaging (NSIM) DTN service. The NSIM service was designed to leverage from the decentralized DTN infrastructure. Its success in the field demonstrates the importance of localized DTN services. Fourthly, using qualitative reading of DTN routing related papers, this thesis describes shortcomings of established DTN routing evaluation methodologies. Extensive use of simulated environments and scarce real-world experiments in the DTN research field often leads to usage of specific hypothetical scenarios. These scenarios are difficult to compare or relate to each other. Additionally, DTN research that does contextualize itself in remote, extreme, and challenging scenarios performs evaluations of proposed routing schemes in urban or academic environments. The DTN evaluation model that is proposed here tries to improve the readability, comparability, and validity of DTN routing evaluations. This thesis also pays attention to the issue of how to evaluate thecomplex interplay that occurs between researchers, users, technology and environment throughout the deployment process. The suggested method highlights the dynamics of resistance, as conceptualized within Actor Network Theory (ANT). It illustrates how employment of the concept of resistance facilitates the recognition of different driving forces in the design process that emerge from the events in the deployment. Ultimately, the thesis contributes with the PRoPHET routing protocol specification in the "Request for Comments" (RFC) document series that is the official publication channel for the Internet Research Task Force (IRTF) and other Internet communities. The protocol specification published as the RFC6693 document allows for actual protocol implementation and assures interoperability. The discussion that follows the RFC document in this thesis focuses on the process of transferring scientific findings gained from the experiments on the deployment field into the Internet draft document that was finally recognized as an experimental RFC within the IRTF.