Multi - level classification and formulation of an integration framework for estimation/ communication/ computation (EC2) co-design

This report is an overview of the research activities regarding the WP06 (C4E co-design) of the FeedNetBack European project for the first six months of the year 2010 at INRIA. The research team consists of Post Doctoral Fellow Alireza Farhadi and Director Carlos Canudas de Wit. In preparing this report we had: a) A meeting with our industrial partner Ifremer. b) A three days visit at the University of Padova (UNIPD), Italy. c) A short discussion with our industrial partner Videotech. d) Long discussions with Sandro Zampieri and Luca Schenato from UNIPD. We also received some results from our colleagues working in ETH (Swiss).The objective of the FeedNetBack project is to propose a co-design framework, which allows the integration of control-estimation, communication, computation, complexity, and energy in networked control systems. This co-design framework is developed for the following case studies: a) A fleet of Autonomous Underwater Vehicles (AUVs) b) Intelligent camera networks for motion capture c) Surveillance systems using a network of smart cameras. The three case studies have been selected to demonstrate the wide spectrum of possible applications of the FeedNetBack project: From systems with relatively few, highly mobile nodes, communicating over a network subject to communication imperfections; to systems with a very high number of immobile nodes, with high available bandwidth but also high computation requirements (smart camera network for surveillance applications and motion capture). To create such a co-design framework we first need to fully understand the constraints imposed by control, communication, computation, complexity, and energy on the above case studies. This is the first objective of this report. The second objective is to formulate an estimation/ communication/ computation co-design framework which is applicable to the above case studies. To achieve these goals, in Section 1, we study fleet of AUVs; and following our discussions with Ifremer, we identify the interactions between control, communication, computation, etc. in this case study. In Section 2, we study smart networks of cameras for motion capture; and following our discussions with Sandro Zampieri and Luca Schenato, we identify the interactions between different components (control, communication, etc.). In Section 3, we study smart networks of cameras for surveillance applications; and following our discussions with Videotech, Sandro Zampieri and Luca Schenato, we identify the interactions between different components. Then, in Section 4, based on this studies, we formulate an integration framework for estimation/ communication/ computation co-design which is applicable to fleet of AUVs and smart camera network for surveillance applications.