Thèse Architecture Assistée par Ris pour des Systèmes de Communication sans Fil Fiables H/F

Doctorat.Gouv.Fr

Cette recherche de doctorat vise à concevoir et à valider une nouvelle architecture assistée par RIS pour des systèmes de communication sans fil fiables. Les objectifs principaux sont d'améliorer la fiabilité des communications sans fil en configurant dynamiquement les éléments RIS pour atténuer l'évanouissement et les interférences, d'intégrer des mécanismes d'authentification cross-couches qui exploitent les signatures de la couche physique telles que les informations d'état du canal pour compléter les méthodes cryptographiques, et d'utiliser des algorithmes d'intelligence artificielle, y compris l'apprentissage par renforcement et les réseaux de neurones, pour optimiser la configuration RIS et les paramètres d'authentification en temps réel.
Next-generation wireless networks, including 6G, the Internet of Things (IoT), and Vehicular Ad Hoc
Networks (VANETs), are expected to support massive connectivity, ultra-low latency, high reliability, and
secure data exchange. However, these networks face several fundamental challenges:
- Channel impairments: Multipath fading, interference, and blockage significantly degrade
communication reliability in dynamic environments such as urban areas and highways [1].
- Security vulnerabilities: Conventional security mechanisms operate at higher layers, introducing
latency and computational overhead while failing to address physical-layer threats such as
eavesdropping, jamming, and spoofing [2].
- Resource constraints: IoT and vehicular devices often have limited computational power and energy,
making complex cryptographic techniques inefficient [3].
Reconfigurable Intelligent Surfaces (RIS) have emerged as a promising technology capable of reconfiguring
the wireless propagation environment by dynamically controlling electromagnetic wave characteristics such
as phase and amplitude. RIS can significantly enhance signal quality, coverage, and energy efficiency while
enabling new physical-layer security mechanisms [4].
Despite these promises, the integration of RIS into reliable and secure communication
architectures remains largely unexplored. Most existing work treats RIS as a passive reflector without
considering cross-layer interactions with higher-layer protocols, adaptive authentication, or AI-driven
configuration [5]. The main objective is to design, implement, and validate an RIS-assisted communication framework that significantly improves link reliability and authentication efficiency for vehicular and IoT networks under realistic channel conditions and attack scenarios.
The proposed research will follow a multi-stage methodology:
- System Modeling: Develop analytical models for RIS-assisted wireless communication systems,including channel modeling and mobility scenarios.
- Algorithm Design: Design adaptive RIS configuration strategies using optimization and machine learning techniques (e.g., Deep Reinforcement Learning).
- Security Framework: Develop cross-layer authentication schemes integrating physical-layer fingerprinting and lightweight cryptographic methods.
- Simulation and Validation: Implement and evaluate the proposed system using MATLAB, Python,NS-3, and LabVIEW environments.
- Performance Evaluation: Assess key metrics such as SINR, Bit Error Rate (BER), outage
probability, secrecy rate, and latency.