Considering the evolution of embedded system technology and use, we can envisage the imminent availability of multiple, highly integrated and networked devices blended into human and machine environments. Such communication and computing resources may offer new opportunities for implementing ambient intelligence, or the adaptation of environment services to human and machine needs.

New services could be expected to assist people during professional activities and leisure time. They could, for instance, be related to identification (face recognition, language analysis and translation, gesture and motion classification), automatic delivery of relevant information (database, statistics, information filtering, classification, data-mining, intelligent display), home automation, multimedia control, automatic translations, and so on.

These new capabilities are based on the following distributed framework:

sensor networks, including cameras, microphones, RFID ...
computing elements for information processing dedicated to different kinds of data: image, audio, motion, ...
decision elements for proposing specialized and personalized services based on AI, evolutionary algorithms, ...
configuration controller for safe transitions between various hardware and software configurations - interface devices for feedbacks and service proposals: displays, speakers, ...
This emerging application domain requires powerful computing capabilities and short response times, but it must also be designed according to the constraints of embedded systems, including resource limitations and energy efficiency.

The use of reconfigurable architectures in this context provides the opportunity to design self-adaptive systems-on-chip that combine GPP facilities with parallel, specialized and configurable processing elements which can implement the data-processing, analysis and decision algorithms. In other words, reconfigurable architectures present adaptation capabilities that fit ambient intelligence requirements.

The objective of this special session is to cover different research directions combining centralized or distributed decision mechanisms coupled with reconfigurable architectures as a first step towards ambient intelligence based on reconfigurable architectures.

Thus, as part of FPT'09, we solicit original research papers related to on-chip configuration decision-making with special interest in the following topics:

General concepts of self-adaptive reconfigurable architectures
Decision issues for reconfigurable architectures (Energy, QoS, Security, Reliability, ...)
Networked/Distributed reconfigurable architectures
Context-aware reconfigurable architectures
Applications of self-reconfigurable systems
Papers should be submitted to the Special Session on Self-Adaptive Reconfigurable Architectures for Ambient Intelligence using the FPT'09 submission system. Submissions close on 8 June 2009.

IMPORTANT DATES
Submission of papers: 8 June 2009
Notification of acceptance: 10 August 2009
Camera-ready papers and author registration due: 7 September 2009

FURTHER ENQUIRIES
Further enquiries should be directed to the Special Session Chair:

Jean-Philippe Diguet
Lab-STICC
CNRS / Université Européenne de Bretagne
jean-philippe.diguet@univ-ubs.fr