Federated learning (FL) can be used to distribute machine learning (ML) tasks across edge and Internet of Things (IoT) devices with limited resources. FL provides an alternative and much more practical solution to classical artificial intelligence (AI), which requires moving large data volumes to energy-hungry data centers. On the other hand, sustainability of FL processes should be accurately quantified as limiting energy consumption might require sacrificing accuracy. This article proposes a framework for real-time monitoring of energy and green house gas (GHG) emissions (carbon footprints) of FL systems. The framework is developed for both classical FL policies relying on the parameter server and emerging fully decentralized ones. The proposed approach considers, for the first time, the impact of ML model quantization and sparsification on the energy/carbon budget while also discussing novel gradient tracking (GT) FL strategies that are robust to data heterogeneity but require higher communication bandwidth. General guidelines for energy-efficient designs are discussed based on several case studies on real datasets. This article quantifies the energy footprint of continual FL processes that implement periodic adaptation on new data as foreseen by emerging IoT industry verticals. Results show that centralized FL is advantageous when strict carbon budgets are imposed or energy-inefficient (80%), provided the ML model compression is properly tuned.

Bayesian Federated Learning (FL) policies enable multiple nodes to collaboratively train a shared Machine Learning (ML) model while accounting for the uncertainty of its predictions. This is accomplished by estimating the global posterior distribution in the model parameter space. Currently, Bayesian FL strategies are impaired by large communication costs that need to be reduced to provide more sustainable training platforms. This letter investigates the impact of compression strategies in centralized Bayesian FL setups, where a Parameter Sever (PS) is tasked to supervise the learning process. The goal is to study how compression affects the ability of Bayesian FL systems to provide high-quality, yet well-calibrated ML models. The analysis is carried out in the healthcare domain, where the prediction reliability is particularly critical, focusing on a medical imaging task. Numerical results show that applying aggressive compression policies highly reduces the ability of Bayesian FL systems to provide accurate and reliable ML models. On the contrary, light compression stages maximize accuracy and calibration at the cost of larger communication overheads.

This paper explores the potentialities of deploying vehicular Cooperative Positioning (CP) systems in urban scenarios utilizing real-world data collected via experimental campaigns. We examine the case of two prototype vehicles equipped with LiDAR sensors for perceiving their surrounding environment and with Global Navigation Satellite System (GNSS) receivers for positioning. The considered use case focuses on the cooperative detection of static landmarks, to be used for improving the vehicles’ GNSS positioning. The experimental campaign points out a severe degradation in ego vehicle localization performances due to complex multipath propagation experienced in the urban scenario. To cope with such a problem, we integrate into the CP system a compensation method able to mitigate the position bias originating from the adverse propagating conditions. Experimental results show that integrating the developed compensation into the CP solution enables an accurate detection of the landmark positions, leading to an enhancement of the vehicle localization accuracy.

The rapid evolution of artificial intelligence (AI) technologies holds transformative potential for the healthcare sector. In critical situations requiring immediate decision-making, healthcare professionals can leverage machine learning (ML) algorithms to prioritize and optimize treatment options, thereby reducing costs and improving patient outcomes. However, the sensitive nature of healthcare data presents significant challenges in terms of privacy and data ownership, hindering data availability and the development of robust algorithms. Federated Learning (FL) addresses these challenges by enabling collaborative training of ML models without the exchange of local data. This paper introduces a novel FL platform designed to support the configuration, monitoring, and management of FL processes. This platform operates on Platform-as-a-Service (PaaS) principles and utilizes the Message Queuing Telemetry Transport (MQTT) publish-subscribe protocol. Considering the production readiness and data sensitivity inherent in clinical environments, we emphasize the security of the proposed FL architecture, addressing potential threats and proposing mitigation strategies to enhance the platform's trustworthiness. The platform has been successfully tested in various operational environments using a publicly available dataset, highlighting its benefits and confirming its efficacy.

This paper focuses on highly precise localization and tracking of vehicles in race circuits, where centimeter-level accuracy is required for safety and for enabling complex maneuvering. Recently, the Annealed Stein Particle Filter (ASPF) has been proposed as a promising Bayesian tracking tool for tracking, showing its superior performances against conventional Bayesian filtering methods, such as the Extended Kalman Filter (EKF) and the Particle Filter (PF). Despite its excellent performances, the ASPF entails large computational complexity, making it unsuitable for highly dynamic vehicular scenarios. To address this shortcoming, we propose a Deep Unfolded ASPF (DU-ASPF), a novel Bayesian tracking algorithm integrating the deep unfolding paradigm where the ASPF operations are rearranged into a sequential structure with learnable weights. Experimental results using raw Ultra-Wide Band (UWB) measurements show that the DU-ASPF is able to substantially speed up the tracking process while maintaining the ASPF accuracy.

Federated Learning (FL) has emerged as a promising approach for privacy-preserving machine learning, particu-larly in sensitive domains such as healthcare. In this context, the TRUSTroke project aims to leverage FL to assist clinicians in ischemic stroke prediction. This paper provides an overview of the TRUSTroke FL network infrastructure. The proposed archi-tecture adopts a client-server model with a central Parameter Server (PS). We introduce a Docker-based design for the client nodes, offering a flexible solution for implementing FL processes in clinical settings. The impact of different communication pro-tocols (HTTP or MQTT) on FL network operation is analyzed, with MQTT selected for its suitability in FL scenarios. A control plane to support the main operations required by FL processes is also proposed. The paper concludes with an analysis of security aspects of the FL architecture, addressing potential threats and to increase trustworthiness.

Bayesian Federated Learning (FL) offers a principled framework to account for the uncertainty caused by limitations in the data available at the nodes implementing collaborative training. In Bayesian FL, nodes exchange information about local posterior distributions over the model parameters space. This paper focuses on Bayesian FL implemented in a Device-to-Device (D2D) network via Decentralized Stochastic Gradient Langevin Dynamics (DSGLD), a recently introduced gradient-based Markov Chain Monte Carlo (MCMC) method. Based on the observation that DSGLD applies random Gaussian perturbations to the model parameters, we propose to leverage channel noise on the D2D links as a mechanism for MCMC sampling. The proposed approach is compared against a conventional implementation of frequentist FL based on compression and digital transmission, highlighting advantages and limitations.

This paper considers the problem of cooperative localization of passive objects in a vehicular environment through the fusion of lidar point clouds collected at different moving vehicles and sent to the road infrastructure. Object localization is then used to improve the position estimate of vehicles according to the implicit cooperative positioning paradigm. At first, each vehicle uses a deep neural network (a 3D object detector) to process its lidar point cloud and localize static objects. Then, the set of estimated bounding boxes is sent to the road infrastructure, which performs data association through a message passing neural network to identify the set of measurements originating from the same detected object. Lastly, cooperative localization of objects is backward used to improve vehicle positioning. Simulations of a realistic cooperative lidar sensing scenario with CARLA software highlight improved positioning compared to non-cooperative tracking.

The exchange of sensing information through Vehicle-to-Everything (V2X) communications enables the development of cooperative systems for localization augmentation in connected automated vehicles. In V2X scenarios, the integration of measurements from multiple vehicles enhances the environmental perception which is of the utmost importance for enhanced safety services. In this paper, we propose a Deep Neural Network (DNN)-assisted cooperative localization method that relies on a centralized road infrastructure and a network of lidar sensors at vehicles. The proposed algorithm is referred to as DNN Implicit Cooperative Positioning (DNN-ICP) and performs two tasks. At first, each vehicle processes its lidar point cloud by a 3D object detector to identify static objects in the surrounding. Then, the estimated objects are collected at the road infrastructure which uses the aggregated information to improve the localization. Numerical results in a realistic vehicular scenario are presented to quantify the improvement provided by DNN-ICP with respect to a non-cooperative vehicle positioning scheme, showing the reduction of uncertainty on vehicle positioning.

This paper addresses the problem of indoor positioning, where complex propagation characteristics call for advanced Bayesian filters for accurate position tracking. We propose to employ the Stein Particle Filter (SPF) to approximate the posterior distribution with a set of particles, using the Stein Variational Gradient Descent (SVGD) method. A novel SPF tracking method, referred to as Annealed Stein Particle Filter (A-SPF), is designed by exploiting the annealed scheduling of SVGD. Compared to SPF, the A-SPF captures multi-modal distributions that easily arise in indoor localization problems from multipath without requiring higher number of particles. Experimentation activities are carried out in two indoor scenarios, an office and a machinery area, where Ultra Wide-Band (UWB) technology is used to collect raw data. Results show the improved positioning performance of the proposed A-SPF compared to conventional solutions based on extended Kalman filter and particle filter, as well as with standard SPF.

In recent years, automotive systems have been integrating Federated Learning (FL) tools to provide enhanced driving functionalities, exploiting sensor data at connected vehicles to cooperatively learn assistance information for safety and maneuvering systems. Conventional FL policies require a central coordinator, namely a Parameter Server (PS), to orchestrate the learning process which limits the scalability and robustness of the training platform. Consensus-driven FL methods, on the other hand, enable fully decentralized learning implementations where vehicles mutually share the Machine Learning (ML) model parameters, possibly via Vehicle-to-Everything (V2X) networking, at the expense of larger communication resource consumption compared to vanilla FL approaches. This paper proposes a communication-efficient consensus-driven FL design tailored for the training of Deep Neural Networks (DNN) in vehicular networks. The vehicles taking part in the FL process independently select a pre-determined percentage of model parameters to be quantized and exchanged on each training round. The proposed technique is validated on a cooperative sensing use case where vehicles rely on Lidar point clouds to detect possible road objects/users in their surroundings via DNN. The validation considers latency, accuracy and communication efficiency trade-offs. Experimental results highlight the impact of parameter selection and quantization on the communication overhead in varying settings.

In Industry 4.0, real-time location systems are emerging as a key technology to improve the efficiency of industrial processes, as they allow to track any assets or material movement and collect data on their usage. Ultra Wideband (UWB) systems offer unrivaled localization accuracy, but they call for augmentation strategies in environments with complex propagation conditions such as plants or factories with high density of scattering objects and obstructions. In this paper, we focus on Bayesian filtering techniques to counterbalance the detrimental effects induced by non line of sight and dense multipath in a smart factory scenario. We first conduct a set of experimental tests with commercial devices in an industrial facility of Pirelli Tyre S.p.A. located in Milan, Italy. We then use the collected data to design and test augmentation algorithms based on Extended Kalman Filter (EKF) and Particle Filter (PF), fusing Time Difference of Arrival (TDoA) and Angle of Arrival (AoA) signals. Experimental results show that, despite the harsh environment, accurate localization is possible by fusion of hybrid measurements and integration of prior information on the target dynamics and the industrial propagation environment.