General Information

It is widely predicted that the next generation of the Internet will be comprised of trillions of IP connected nodes. Furthermore, that these nodes will be specialized devices that produce and consume content in a different way than today's predominant web server/browser combination. One analogy could be that this new Internet will resemble a very large sensor and actuator network (VLSANET), collecting, storing and organizing the data for retrieval by man and machine. Given the scale of the task, existing storage, retrieval and computational models fall short.

It is the position of this workshop to show that a new class of data base system is required to meet the emerging application needs of the next generation Internet. The characteristics of the new class demonstrate the write speed requirements of a large transaction processing databases, the relaxing of persistence guarantees, the calculation of time series statistics within ad hoc queries and read speeds guaranteed by a quality of service.

Information in an Internet device environment is both supply and demand flexible, specifically devices may go on and off line without warning. Common schema for supply and demand will not exist, some standards may emerge, but the more flexible the logical data model, the more useful the system will be. Questions are not pre-determined, therefore an optimization is difficult to achieve with a priori information, such as pre-aggregation, indexing, optimal storage architectures. By its very nature, data is not in the same location, in some cases the data may be mobile. Furthermore there is no central concept of state in an internet of devices. How can complex multi-level transactions be handled ? Standard transaction mechanisms are meaningless. New methods need to be introduced if state is required.

A useful subset of this problem is framed by the use of the data in a control and signaling application. Conventional record-based meta data are not part of the system. Four dimensional attributes become interesting, location (in 3D) and time -- and the physical reading, say temperature, at that point. This can be simplified to a 2D storage problem if the devices are assumed to be stationary. This is a sampled reality and therefore can be applied to models as discrete mathematics and follows well known principles of control, signal processing and statistical analysis. Problems are then bounded by sampling rates, precision of the reading and any other systems anomalies introduced by the information system. The information system will introduce lags.

We seek papers describing ideas and current work in the general area of new generation database systems for a sensor and object rich internet. Interesting areas include (but are not limited to):

• Applications
• Artificial Intelligence
• Handling four dimensionality of data objects
• Information retrieval and decision making
• Mathematical support models
• Mobile data
• Optimisation
• Locating and interconnecting
• Transaction management
• Querying data streams
• Sampling
• Security and Reliability
• Semantics
• Statistical methods
• Test beds

We would like to receive contributions from those who have an involvement in researching, developing or using device rich internet systems who have considered the above issues and have some questions, thoughts or solutions to offer. Research papers, application area reviews, demonstrations, panel and discussion topics are welcome.

Invited Speaker

Prof. Keith Jeffery from STFC will give a guest lecture at the workshop. Keith is currently Director IT and International Strategy of the Science and Technology Facilities Council, based at Rutherford Appleton Laboratory in the UK. Keith has been a major contributor for many years to the subject area of internet and data base systems serving on many esteemed professional and research committees such as Next Generation GRIDs expert group of DG INFSO F2 and Endowment Board of the VLDB (Very Large Database) Conference. He has numerous publications in refereed journals, books and conference proceedings and holds three honorary / visiting professorships. Keith has been considering the idea of the internet of billions of devices for some years and will share his findings in this area with us.

Formation of Consortium Opportunity

It is hoped that an outcome of the workshop will be the formation of a consortium through which we can make funding bids to advance this area. Participants from European countries are most welcome as are participants from the rest of the world.