The cost of learning dilemma:

In the last years, Artificial Intelligence (AI) has become an increasingly central actor in our lives, enabling technological solutions to adapt to the immediate needs of the end-users. The shift from a static to a continuously evolving technology does not come for free but presents a critical cost in terms of communication bandwidth, computational power, and other network resources. Hence, telecommunication networks are facing competition between traditional data flows and those supporting the training of AI algorithms. Despite this, most works still assume that learning and user data are exchanged on separate channels, ignoring the dependency between agent training and network conditions. This project aims to define new strategies for combining the optimization of network resources and learning algorithms, analyzing how the improvement of agent training implies a resource reduction for target applications.

Lightweight EEG acquisition:

For remote EEG monitoring, it is necessary to implement ad hoc architectures combining wearable devices and Body Sensor Networks (BSNs) for processing the signals. The main challenge is tuning the number of sensors used for signal acquisition, which, in the case of EEG, can go from 15 to 64 electrodes or more. This project aims to design new algorithms for processing EEG signals, obtaining an acquisition setup with a limited number of electrodes that still enables the reconstruction of a full EEG at the receiver. The project involves the design of new algorithms for reconstructing a full EEG and the analysis of the trade-off between the reconstruction accuracy and the amount of input information. To evaluate the final performance, it will consider both the mathematical difference between the original and the reconstructed signals and the signal clinical evaluation, assessed in terms of the probability of detecting relevant clinical features.

UAV networks for remote health monitoring:

Body Sensor Networks (BSNs) offer a non-intrusive means of monitoring health, as they enable the continuous tracking of vital signs without causing discomfort or disruption to daily lives. A reliable telecommunication system is pivotal for ensuring data transmission between BSN and clinical facilities, where health parameters are analyzed. However, there are still vast regions, especially in rural and remote areas, where establishing a stable internet connection is a significant challenge. This project aims to design a dynamic telecommunication system, integrating local sensors, that record health data, and Unmanned Aerial Vehicles (UAVs), liable for data collection and processing. The challenge is to adapt the different system components, from wireless communication technology to routing protocols, towards the maximization of the overall performance, measured in terms of clinical assessment of the users' health. More practically, the project involves the modeling of the described scenario, the design of a suitable strategy for system optimization, and the evaluation of the proposed architecture via simulative tools.