Mohamad Atwi

Research Engineer @L2S - Laboratoire des signaux et systèmes

Paris, FR

I am currently undertaking a research internship at the L2S Laboratory, affiliated with CentraleSupélec, where I am engaged in advanced research at the intersection of machine learning and next-generation wireless networks. My primary focus lies in the application of Federated Learning to wireless systems and the investigation of semantic communication frameworks for 6G networks.As part of this work, I have contributed to the study and development of novel methodologies that tackle critical challenges in semantic communications, with an emphasis on robustness, scalability, and edge intelligence. My involvement spans the following areas:Robust Semantic Transmission: Analyzing the performance limits of DeepSC under multi-interference scenarios and proposing interference-resilient designs.Multi-User Semantic Access: Designing scalable sequence-spreading solutions and signal refinement networks for communication under harsh RF interference.Zero-Shot Wireless Sensing: Leveraging physical-layer semantic features to enable model generalization across environments without labeled data.Structural Semantics & Topology: Utilizing persistence diagrams for efficient and inference-oriented semantic compression.Knowledge Graph-Based Communication: Employing reinforcement learning and attention mechanisms to optimize semantic data selection and resource allocation.Edge-Aware Pilot Optimization: Integrating semantic encoders with OFDM systems and proposing hierarchical learning-based pilot placement for improved channel estimation.This internship is deepening my expertise in intelligent and distributed communication systems, and reinforcing my contribution to the evolution of wireless networks from traditional bit-oriented paradigms to meaning-driven, AI-native architectures for 6G and beyond.

CentraleSupélec

Master's degree

Lebanese University - Faculty of Engineering III

Diploma of Education