Overview

The Internet is currently experiencing a huge change in the way that users are communicating over it and using provided services. Unstructured peer-to-peer protocols and overlay network architectures have been proposed, and to a lesser extent deployed, to address limitations of today's Internet. The dominant client-server architecture suffers from centralisation and all the problems deriving from it. As more users request services from a single server, the server's performance and observed data rates degrade. Centralised servers, or centralised clusters of servers, suffer from the single-point-of-failure problem, that can lead to service interruption due to random failures or failures of malicious intent. Peer-to-peer protocols are able to solve such problems by using the resources of all participating peers and replicating the available data without the need of central indexing. Furthermore, the vision of utilising the Internet for providing Voice-over-IP (VoIP) services and converging mobile networks with fixed telecommunications and data networks is quickly becoming a reality. Even today users are able to access Internet services via their mobile phones, and communicate in the reverse direction as well. Since it is unlikely that a single network architecture will become prevalent, the research community has developed a number of proposals that aim to allow the interoperation between them and converge them to a unified system. Naming is a central issue in such an environment.

This research project investigates the design principles and operational properties behind next generation naming systems. Next generation naming will enable interoperability between the various existing naming systems (such as DNS(sec), fixed and mobile telecommunication networks, ENUM, as well as proprietary naming systems like Skype).