Current students

Nowadays, the interest over industrial control and automation has skyrocketed due to the emergence of the Industry 4.0 and Intelligent Transportation Systems (ITS). In these scenarios, environmental awareness and sensing have become a key factor for improving efficiency and safety in critical applications, such as collaborative robots and automated driving. Wireless cloud networks have the potential to overcome current limitations based on single device/vehicle sensing, exploiting the networked infrastructure for exchanging massive amounts of data through dedicated device-to-device (D2D) and vehicle-to-vehicle (V2V) connections. Cooperative techniques can work synergically within these networks to enable the fusion of information coming from interconnected devices/vehicles and augment the perception of the cooperating nodes.

The goal of this Ph.D. research is to develop machine learning-based cooperative sensing techniques for Beyond 5G (B5G) industrial wireless networks, with main applications to smart factory and smart mobility. The research is focused on federated learning setups for wireless edge cloud networks, where on-device intelligence is exploited for distributed model training and testing. Experimental activities will target Industrial Internet of Things (IoT) and Internet of Vehicles (IoV) use-cases, where cooperating agents (e.g., automated vehicles) use distributed cooperative learning approaches for sensing and mapping the surrounding environment.