Scientific direction Development of key enabling technologies
Transfer of knowledge to industry

PhD : selection by topics

Engineering science >> Electronics and microelectronics - Optoelectronics
2 proposition(s).

Participatory localization of connected objects through deep learning techniques

Département Systèmes

Laboratoire Communication des Objets Intelligents

01-10-2018

SL-DRT-18-0806

benoit.denis@cea.fr

Conventional localization methods based on low-cost and low-complexity wireless communication standards (e.g., Long Range - IoT Lora or Sigfox systems, WiFi/BT-LE) can provide a positioning accuracy level (respectively, 500m to 1km in outdoor, 5m to 10m in indoor) that is not suitable to the needs of emerging applications, such as geo-referencing of the data produced by mobile IoT sensor nodes, or smartphone-based pedestrian navigation in public areas or office buildings. In the frame of these PhD investigations, we thus propose to rely on Artificial Intelligence on the one hand (typically, on deep learning techniques), as well as on advanced management/representation methods adapted to spatialized data on the other hand, so as to improve localization performance, while relaxing underlying technological specifications (more particularly, in terms of wireless transmission capabilities). The idea is to learn (over both space and time) and to fuse multi-parameter maps, benefitting not only from radio metrics notoriously ill-adapted to localization (e.g., received power, packet error rate...) and contributed in a participatory way by the different mobile agents on the field, but also from other available modalities depending on the application context (e.g., prior road network map/building layout, physical environmental measurements such as ambient sound, embedded inertial units, etc...).

Télécharger l'offre (.zip)

micro-concentrator for spatial applications

Département des Technologies Solaires (LITEN)

Laboratoire Photovoltaïque à Concentration

01-09-2018

SL-DRT-18-0861

philippe.voarino@cea.fr

CPV (Concentrator PhotoVoltaics) technologies take more and more an important play in the field of existing PV solutions. Record modules reach efficiency of 36.7% with cells of 5x5mm². By miniaturizing concentrating systems, it becomes possible to reduce costs and also to increase the efficiency by using microelectronics processes developed in LEDs manufacturing. These based micro-concentrator systems can also be used for spatial missions for which the surface reduction of III-V solar cells has a direct impact on the cost and the robustness of the solar generator. LITEN has lied on micro-concentration for many years, and continue to develop this activity. Within the framework of the thesis, we propose to study different solutions of lenses on wafer which can be applied for spatial applications, and the associated physical phenomenon. Material behaviors should be taken into account for extreme environmental conditions. A micro-concentrator system will be realized and tested under solar simulation.

Télécharger l'offre (.zip)

Voir toutes nos offres