Scalable approaches for DiffServ multicasting

Over the last several years, there has been an explosion in the introduction of new Internet technologies. Whereas the Internet in its original form was a medium primarily for academia and research interests, the Internet has been redefined as business and consumer interests have dominated the focal points of Internet technology. The dominant question facing the Internet today is, how can the network meet the needs of the users and their applications while trying to keep such implementations scalable to the billions of users present on the Internet? Two of the emerging technologies for answering the question are Differentiated Services (DiffServ) and multicasting. Although the two technologies share complementary goals, the integration of the two technologies is a non-trivial issue due to three fundamental conflicts. The three fundamental conflicts are the scalability of per-group state information, senderversus receiverdriven QoS, and resource management. The issues surrounding how to solve these conflicts provide the basis for this dissertation. In this dissertation, two architectures (DiffServ Multicasting (DSMCast) and Edge-Based Multicasting (EBM)) are proposed to satisfy the requirements for scalable DiffServ multicasting architectures. In addition to the two architectures, this dissertation also presents the first study regarding single tree support for heterogeneous QoS multicasting. Furthermore, the dissertation proposes a novel application of DSMCast for fault tolerance and management of the DiffServ network. Finally, the dissertation comments on future applications of the architectures and proposes several areas for future research.