The goal of FIA (as stated in their site) is to permit open interactions and cross-fertilization across technical domains, reaching out to whoever has talent. FIA works towards:

• Common deliverables creating value for the projects concerned
• Joint strategic research agenda regarding common actions and requirements
• Developing a consolidated calendar of events aiming at avoiding fragmentation of efforts

The Madrid event consisted of roughly four session. The first session contained discussed the national research in various countries. Then came two parallel track sessions. The first one with with three tracks, the second one with four. Report of the activity in the seven parallel tracks made up the last session. The session agendas and the position papers are available online.

All parallel track were also to contribute suggestions for experiments to an experimental testbed called FIRE. The session about the national research initiatives contained talks from Germany (by Volkmar Dietz, BMBF, “G-Lab”), France (Francois Jutland, Scientific Director Telecom Institute), Finland (Reijo Paajanen, CEO of TIVIT), and Spain (Fernando Fournon, Executive President of Telefonica I+D).

The German and Finish presentation were mostly focussed on the various technologies that are needed in a future internet (100 gigabit ethernet, security, quantum communication, various internet technologies related to routing, transport, and information processing)

The French and Spanish presentation were more high-level. The French presentation focused on the creation and cooperation between labs. Interesting to me were the explicit negative remarks about internet user’s preference for free content.

The Spanish presentation was the most abstract about what the future internet should look like:

• Web 1.0: text, some music, and small images
• Web 2.0: Web 1.0 + user generated content, video, and graphics
• Web 3.0: Web 2.0 + HD video, virtual 3d worlds, 3d navigation

Some keywords: 3D interfaces, 360 degrees screens, hybrid world (real and virtual seamless and blurred), holography. A need for open interfaces, open applications, interoperability to create an ubiquitous ambient intelligence.

The topics in the first parallel track session were: Future Content Network(s), Management & Service Aware Networking Architecture, and Trust and Identity in the Future Internet. I went to the Future Content Network(s) track.

A large part of this session was about various forms of video: 3D, multi-view (where the viewer can select a viewing angle), and beyond HD. The obvious effect on the network is that more bandwidth is required. However, with layered encoding (low bandwidth low resolution stream is part of a high bandwidth HD stream), it is suggested that networking devices that route video should be able to extract those low bandwidth streams and pass them to low bandwidth devices. For video, when the network becomes overloaded, it is important to maintain the quality of the audio channels and reduce the quality of the video. This typically needs to be done ‘in’ the network.

More general, the future internet should be less about sending packets to hosts and naming hosts, and more about naming data, and finding the nearest copy of a piece of data in the network.

The idea is to move to a content aware network, where messages are much more self-describing, the network understands the semantics of content, and can route based on content.

The second parallel tracks session contained four tracks: Future Internet Service Offer, Real World Internet, Socio-Economics, and Usage of Experimental Facilities based on Use Cases. I went to the first half of Real World Internet and to the second half of Future Internet Service Offer.

The Real World Internet session was mostly about sensors. One interesting project is GSN. However, there did not seem to be any new requirements for a future internet. One presentation expressed the need for support for events and ‘Sense-decide-act’-loops.

I went to the second half of Future Internet Service Offer. I was unable to find an coherent story of what they were trying to do, let alone requirements for a future internet.

The fourth part contained reports reports of the parallel track sessions. I’ll report on the ones I didn’t go to. From Management and Service Aware Networking Architecture: unclear to me. From Trust and Identity in the Future Internet: need for ‘real-time’ trust, based on historical data. Support for compartmentalization. ‘Trust can’t be outsourced’, but the actual implementation of trust mechanisms can. Need for a proper level of regulations (comparable to consumer protection laws). Privacy: people leave life-long traces on the internet. Privacy versus accountability, anonymity, and traceability. Using the right technology, for example, minimal disclosure protocols. Privacy is not a priority for most people (low user awareness). with respect to regulations: the legal landscape is cluttered and lagging. There is a need for reference scenarios.

Apparently, the main feedback on the FIRE (experimental testbed) was that the offer was not clear enough.