Réseaux de communication et récolte d'énergie : optimisation et démonstration – E-CROPS

“Anytime, anywhere, anything” has been the recent catch-phrase used by technology evangelists promising untethered wireless data flow not only among people, but also among devices of any imaginable sort. Trillions of autonomous devices are foreseen in this promise. While wireless sensor networks solve the wiring problem, we would still need to recharge or replace hundreds of batteries every day or supply the network infrastructure (e.g. base stations in the cellular world) with an ever increasing amount of power. Harvesting available ambient energy, such as solar, thermal or electromagnetic, to power wireless devices is the only viable solution to realize this promise in a sustainable manner. Building upon this requirement, in E-CROPS we study the design, optimization and implementation of a wireless network in which, nodes can harvest renewable energy and store the extra energy in their batteries, be it at the terminal or the network infrastructure side.
In our theoretical study, we aim to formulate a mathematical theory of communication for energy harvesting networks, considering the communication network jointly with the energy network consisting of the harvesters and the storage units. We propose to design the protocols enabling the adaption of the physical and network layer design to the temporal changes in the available energy as well as to the distribution of the energy within the network.
In parallel, we will study novel vibration and thermal energy harvesters as well as the appropriate storage units, and integrate these devices into a wireless sensor network application as a proof of concept for our scientific results. Bringing together researchers from theoretical and experimental backgrounds, we expect that both sides will benefit significantly from this interaction: 1) The design of the harvesting and storage units will be adapted to the needs of the specific communication scenarios; 2) Practical energy and data profiles will be used to design and compare our algorithms, and finally 3) A proof-of-concept sensor network application will be implemented to test the practicality of our ideas in real world.