Current students

Precise localization, with cm-level accuracy, is vital for vertical application scenarios envisioned by the fifth generation (5G) of mobile communications (and beyond). By integrating collaborative communication and localization into a same infrastructure, millimeter-wave (mmWave) transmission technologies have enabled a paradigm change in positioning services. However, conventional localization algorithms, mainly based on geometry techniques, suffer from non-line-of-sight (NLOS), where multipath comes at different angles and with delay dispersion, resulting in positioning inaccuracy. Unlike geometry-based algorithms, Machine Learning (ML) tools are able to deliver user equipment (UE) localization by training ad-hoc models on space-time fingerprints extracted from the environment and based on the whole raw Channel State Information (CSI).

In my Ph.D. research, I will take advantage of my telecommunication and ML background to investigate the open problems of 6D localization (3D position + 3D orientation) in 6G ultra-wide bandwidth (UWB) systems based on high-dimensional CSI.
In particular, I will extend state-of-the-art works by studing cooperative Deep Neural Network (DNN)-based localization algorithms based on CSI, which, in UWB systems, is like a thumbprint of the user location.