Research

I am currently pursuing a Ph.D. with the MARACAS team at Inria Lyon, France, since November 2024.

• Thesis title: Deep Learning for Radio Signal Classification
• Doctoral School: ED160 EEA
• Funded by Inria and the Direction générale de l’armement (DGA)

Thesis Overview

My research investigates how deep learning can be used to automatically recognize and characterize radio signals. The goal is to predict key transmission parameters—such as carrier frequency, bandwidth, and modulation scheme—directly from raw waveform data.

By combining advanced neural architectures with modern signal processing, this work aims to enable adaptive communication receivers capable of operating in complex and dynamic radio environments.

To this end, I design realistic datasets using the CorteXlab testbed. The overarching goal is to contribute to next-generation intelligent spectrum monitoring and sensing systems.

Publications

• CorteXforge: Scenario-Based RF Dataset Generation on Real Testbeds
  Andréa Joly, Leonardo S. Cardoso, Jean-Marie Gorce
  A framework for generating realistic, reproducible, and annotated RF datasets on programmable SDR testbeds.
  Preprint, 2026 · HAL · arXiv

Research Tools

I develop open-source tools to support different aspects of my research, from experimental RF dataset generation to automated scientific literature monitoring.

Academic Service

• Journal Reviewer: IEEE Wireless Communications Letters (WCL), 2026
• Conference Reviewer: IEEE Wireless Communications and Networking Conference (WCNC), 2026

Recognition

• Finalist at the Three-Minute Thesis Competition (3MT) IEEE ICMCLCN 2025 (May 2025)