Titre de la thèse: Robust information gathering and dissemination in wireless sensor networks
Résumé de la thèse: Wireless sensor networks have witnessed a growing interest in the last decade. These low-cost devices allow to efficiently collect and process measurements of spatially spread physical phenomena. They enable many applications, ranging from environmental monitoring (pollutants level, air and water quality), or traffic monitoring in the context of smart cities, to equipment monitoring and data logging in industrial applications. Many research and industrial problems still remain open in the context of wireless sensor networks. This thesis will consider distributed and robust data compression to facilitate the collection of correlated data. It will also consider the development of distributed estimation algorithms robust to sensor failures or to the presence of misbehaving nodes in the wireless network. The aim of the thesis is twofold. The first objective is to identify, for well-defined scenarios of interest, the fundamental limits of communication and distributed signal processing capabilities of the network. The second objective consists in the individuation, on the basis of the obtained theoretical results, of a set of design guidelines (i.e. which architecture/signal processing technique is better suited for a task under given energetic/complexity constraints), to orient the development of effective solutions. In order to attain this goal, prototypical implementations will be developed.