The current Internet architecture focuses on communicating entities, leaving aside the exchanged information. However, trends show that WHAT is exchanged is becoming more important than WHO are exchanging it. As a result, the Internet is effectively moving from interconnecting machines to interconnecting information, questioning the necessity of endpoint identifiers as a foundation of internetworking.
There have been efforts to introduce identifiers for information rather than endpoints on internetworking level, e.g., the FP7 PSIRP and the PARC CCN project, the former based on an information-centric publish/subscribe model and the latter based on a content-centric overlay. These efforts promise, among other things, more flexibility in adapting to new services, efficiency improvements on lower layers, and native multicast support.
However, crucial questions remain, like (1) the role of caching, potentially leading to new concepts for congestion and error control, (2) the role of transport functions, (3) service-specific delivery solutions, and (4) support for disconnected operations. There are also open questions regarding the influence on lower layer operations, including resource allocation and reservation, resulting from the increased visibility of the exchanged information. Also, efficient re-allocation of optical resources and bundling of wireless resources need to be studied to gain insight into how such novel architecture can efficiently work with lower layers. Last, but not least, the socio-economic environment needs to be understood. The markets being created as well as the impact of socio-economics on design choices are as important as outlining the opportunities and threats for existing and new players to come. Demonstrating and validating practical solutions in experimental testbeds is needed for complementing such understanding.
PURSUIT will pursue these endeavours in a collaborative project setting, building on recognized expertise in relevant areas.