Self-Organizing Multimedia Architecture

Akronym:   SOMA
Start:     12.2008
Ende:     11.2011

 

Abstract
The Lakeside Labs SOMA project, under the leadership of Prof. Dr. Laszlo
Böszörmenyi
is preformed in cooperation with the scientific partner Alpen-Adria Universität Klagenfurt and the industrial partner ASFiNAG Mautservice GmbH.
SOMA will contribute significantly to the research in the area of multimedia systems and plans to develop a prototype "LOOK2" showing the practicability of SOMA in a real world scenario and which can be used as a multimedia platform for traffic information.

Scenarios in layers
The principle of SOMA is to gather the full life-cycle of multimedia data from data units and to integrate them in a larger multimedia information system.
SOMA's central concept is to focus on scenarios, which are defined over specific time periods and recorded as events with a semantic meaning. More precisely, the system architecture is composed by three self-organizing layers:

  • The sensor network layer is a network of intelligent sensors, which can collect data. This layer is self-configurating and adjusts its behaviour to the environment and the needs of the layers on-top.
  • The distribution network layer optimizes resource management in terms of storage, processing, adaptation, and delivery. The relevance of information is computed based on the observed use of the data and the specific intentions of users.
  • The user layer, or user interface, of SOMA is a platform, which gets implicitly or explicitly feedback from the layers below, i.e. the data is either  automatically provided by the other layers or actively requested from the other layers.

Soma is based on the interaction of the different layers; the resulting effects build the core of the SOMA research project.


Exchange by users
The three layers continuously exchange information on events and the associated data and metadata to ensure a multi-media functionality and permanent interaction on all three layers. The transparency of the interactive layers and the functionality of the distribution layer enable sensors to react on requests defined by users. The distribution layer observes sensor capabilities and usage patterns in order to harmonize and organize input both from users and sensors. The users are able to follow reports almost in real time without however relying on automatic routines.


Prototype and test scenario
In principle, SOMA allows to analyze, besides the investigation of the self-organized layers, a multiplicity of other research problems too; in both fundamental and applied research.
At the end of each year the three layers of SOMA are subject to a major test scenario and a substantial evaluation. This will ensure that further research challenges within layers are identified and can easily be organized in terms of research questions on the working prototype "LOOK2".


For further and more detailed information please visit the SOMA Homepage.

 

Lakeside Labs
Copyright 2010 Lakeside Labs