Internet content distribution has, over the recent years, undergone profound and radical changes. Aside the traditional client-server mode of diffusion, lacking a wide support for IP multicast that is made complex by group management issues, new paradigms such as Content diffusion networks (CDN) or Peer-to-peer (P2P) have been proposed over the last decade.
While CDN and P2P are very succesful technologies, they are both end-to-end solutions layered over-the-top of the TCP/IP infrastructure, which implies the content has to be injected over the network possibly several times. Additionally, CDN and P2P are application-specific solutions, and are thus generally considered as patches over the current infrastructure, rather than long-term solutions.
Under this light, new paradigms such as Information Centric Networks (ICN) are appearing that advocate a redesign of the architecture around the concept of content — rather than around the concept of address as in classical TCP/IP. Among the architectures under the ICN umbrella, Content Centric Network (CCN) seem to be the most promising.
The activities carried out at LINCS encompass all these diffusion methods, historically including P2P and CDN, and moving naturally toward ICN and CCN. Activities in these areas span a rather large spectrum of methodologies, including fluid and stochastic modeling, simulation, prototyping and experiments, and Internet traffic measurement.