Thesis: Mohamed Amine BOUZAIDI TIALI

Rapporteurs:

• Tan-Phu VUONG - Grenoble Polytechnic Institute.
• Charles TATKEU - Université Gustave Eiffel.

Examiners:

• Fadi KURDAHI - University of California.
• Assia NAJA - International University of Rabat

Invited guests:

• Marwane AYAIDA - Université Polytechnique Hauts-de-France.
• Ahmed DIDOUH - King Abdullah University Of Science And Technology.

Encadrants:

• Yassin EL HILLALI - Université Polytechnique Hauts-de-France.
• Mourad OUADOU - Faculty of Sciences of Rabat.

Directors:

• Atika RIVENQ - Université Polytechnique Hauts-de-France.
• Khalid MINAOUI - Faculty of Sciences of Rabat.

With the rise of autonomous vehicles in the automotive industry, requirements for on-board perception have evolved considerably. GPS-based localization systems and on-board multimodal sensors (cameras, radar, LIDARs) are no longer sufficient to guarantee reliable, wide-ranging perception of the environment, particularly in complex, dynamic or partially obstructed urban contexts. Vehicles now need to be able to "see beyond
" their own sensors, by integrating them into their systems. "of their own sensors, by integrating exogenous data from their environment. In this context, Vehicle-to-Everything (V2X) communications are becoming a fundamental pillar of augmented perception. However, these communications face major challenges linked to high mobility, radio interference and wireless channel variability. Channel estimation thus becomes a critical task, conditioning the reliability and responsiveness of information exchanges. In addition, identification of the environmental context (urban, motorway, rural) is essential for dynamic adaptation of communication and perception strategies. This thesis proposes the study and development of a hybrid vehicular network combining the advantages of ITS-G5 technologies (IEEE 802.11p standard) and next-generation cellular networks (LTE-V2X / 5G). The aim is to design an adaptive communication architecture capable of ensuring enhanced cooperative perception and continuity of service in various mobility scenarios. Theproposed approach is part of the Cooperative Intelligent Transport Systems (C-ITS) framework and aims toimprove the safety, reliability and overall performance of autonomous vehicles.