ED Mathématiques et Informatique
Study of the Performance of ITS Communication Technologies for Micromobility
by Rima BOUGHARIOU (LaBRI - Laboratoire Bordelais de Recherche en Informatique)
The defense will take place at 10h00 - Salle des thèses Ecole Nationale d'Electronique et des Télécommunications de Sfax (ENET'Com), Route de Tunis km 10, Cité el Ons, Technopôle de Sfax, Sakiet Ezzit, Tunisie
in front of the jury composed of
- Mohamed MOSBAH - Professeur - Bordeaux INP - Directeur de these
- Hichem BESBES - Professeur - École Supérieure des communications de Tunis (Sup'Com) - Rapporteur
- Hassene MNIF - Professeur - Ecole Nationale d'Electronique et des Télécommunications de Sfax (ENET'Com) - CoDirecteur de these
- Lamia CHAARI - Professeure - Institut supérieur d'informatique et de multimédia de Sfax (ISIMSF) - Examinateur
- Layth SLIMAN - Professeur - EFREI - Examinateur
- Riadh DHAOU - Professeur - Toulouse INP-ENSEEIHT - Rapporteur
The rapid growth of micromobility is progressively transforming urban transportation systems. Electric bicycles and e-scooters have emerged as efficient alternatives to conventional transportation modes, contributing to reduced traffic congestion and lower environmental impact. However, integrating these vulnerable road users into Cooperative Intelligent Transportation Systems (C-ITS) introduces new challenges related to connectivity, service continuity, mobility management, and communication reliability. Vehicle-to-Everything (V2X) communications constitute a fundamental component of modern C-ITS architectures. They primarily rely on ITS-G5 technology, based on the IEEE 802.11p standard, and next-generation cellular networks such as 5G NR-V2X. Although these technologies offer complementary capabilities, their efficient utilization in micromobility environments remains challenging due to dynamic user mobility, limited onboard resources, and highly variable wireless conditions. This thesis investigates the performance of ITS communication technologies for micromobility and proposes several mechanisms to improve their efficiency. First, multiple representative use cases are analyzed within an ITS-G5-based Vehicular Ad Hoc Network (VANET) environment, including adaptive speed control, dynamic geofencing, and energy-efficient communication infrastructure management. Intelligent Road Side Unit (RSU) management mechanisms, including the deployment of mobile RSUs, are proposed to improve network coverage while reducing energy consumption. The obtained results demonstrate energy consumption reductions of up to 44.5% for RSU infrastructures while maintaining V2X communication continuity. Second, a hybrid ITS-G5/5G communication architecture is developed to leverage the complementary strengths of low-latency local communications and wide-area cellular connectivity. An intelligent radio interface selection strategy based on Deep Reinforcement Learning and the Double Deep Q-Network (DDQN) algorithm is proposed to dynamically optimize network access decisions. The proposed approach improves communication reliability, packet delivery ratio, throughput, and energy efficiency under dynamic urban mobility conditions. Performance evaluation shows a Packet Delivery Ratio (PDR) of up to 99% and an energy consumption reduction of 15.6% compared with conventional communication selection approaches. Finally, this thesis addresses the challenge of vertical handover management in hybrid ITS-G5/5G environments. An innovative handover optimization framework based on Offline Reinforcement Learning and the Conservative Q-Learning (CQL) algorithm is introduced. The proposed Radio Interface Management Approach for Handover Optimization (RIMA-HO) exploits historical communication data to reduce unnecessary handovers while maintaining service continuity and communication quality. The RIMA-HO solution reduces unnecessary handovers and improves service continuity, achieving an average energy reduction of 16.3%, with savings reaching 41.8% under the most favorable scenarios. The proposed contributions are validated through simulations conducted using a SUMO and OMNeT++ co-simulation environment, as well as real mobility and communication traces. Experimental results demonstrate significant improvements in connectivity, energy efficiency, communication stability, and mobility management. This work contributes to the development of intelligent communication solutions that support the safe and efficient integration of micromobility into future Cooperative Intelligent Transportation Systems.