Type 1 diabetes is a growing global health challenge. Standard clinical practice often relies on manual insulin injections, which can lead to suboptimal glucose regulation. Recent advancements have shifted focus towards Artificial Pancreas systems, integrating continuous glucose monitoring with automated insulin delivery. This work presents a quantitative comparison of four Deep Reinforcement Learning algorithms for autonomous glycemic regulation via insulin injection: DDPG, PPO, SAC, and TD3. The validation is conducted using the Hovorka model, in presence of uncertainties on number, time and amount of meals. Results show that all four controllers are able to maintain blood glucose levels within the target range. The TD3 algorithm outperforms the others in terms of several key performance indicators such as time in range, time in hypo/hyperglycemia and total insulin usage, while also exhibiting fewer hyperglycemic episodes compared to prior works in academic literature.

Artificial Intelligence and Machine Learning have brought transformative changes to clinical diagnostics, especially in image classification via deep convolutional neural networks. The latter are crucial in analyzing and identifying diseases from medical visuals, ensuring accurate diagnosis and prevention. However, the limited availability of medical imaging data poses a serious challenge and leads to poor performance of the classifier. To address this issue, both geometric/color augmentation and synthetic data generation (through Generative Adversarial Networks and Variational Autoencoders) methods are employed, both operating on the entire dataset without an intrinsic optimization of the classification process. This study introduces an automated data augmentation strategy in which the editing operation is customized with respect to the individual images contained in the training set. This is done by using the Deep Reinforcement Learning framework provided by the Proximal Policy Optimization algorithm, with the reward being the test accuracy of the image classifier. The application of the proposed procedure on a meager dataset related to gastrointestinal diseases demonstrates an improvement in image classification by over 3%.

In digital transformation, Industrial Data Space (IDS) is a key enabler for industry 4.0 to improve the industrial process, efficiency, and energy consumption by exploiting learning-based techniques. The present paper focuses on improving the decision-making process in complex industrial environments by developing a Deep Reinforcement Learning (DRL) based real-time assistant. Mainly, we address a use case from the space industry to improve the launcher throughput and efficiency and reduce cost by optimally managing the industrial resources. A mathematical formulation of the Industrial Production System (IPS) and a simulated environment are developed to train the DRL-based Proximal Policy Optimization (PPO) agent. The proposed method is scalable and in line with the dynamic nature of the industrial production systems to overcome the domain-dependent heuristics extensively used in the manufacturing industry. Furthermore, simulation results show that the proposed method can provide industrial operators and managers with a real-time decision support system to increase the Return on Assets.

This paper deals with the interworking of a satellite system for mobile multimedia applications with the terrestrial environment. The concept of an "overlying network" is introduced with the aim of giving a common approach for the interworking of the considered satellite system with the various terrestrial networks (B-ISDN, N-ISDN, TCP/IP). The paper particularizes the introduced concept to the case in which the considered satellite system is interfaced with the broadband ISDN (B-ISDN).

This paper deals with the medium access control (MAC) subsystem which has been implemented in the MEDIAN project (wireless broadband LANs for professional and residential multimedia applications) belonging to the European Union's (EU's) advanced communications technologies and services (ACTS) program. The proposed MAC aims at providing a low complexity and efficient transport for asynchronous transfer mode (ATM) cells from a base station (BS), interfaced with an ATM local exchange (LEX), to ATM portable stations (PSs) and vice versa. A time division duplex (TDD) scheme is adopted in which uplink and downlink frame durations can be varied frame-by-frame, in order to fit the present uplink/downlink traffic mix. Moreover, in the uplink, a contention-free based technique is considered in which the slots are assigned to the calls depending on the present load of the PS uplink buffers.

We refer to the possible performance improvement consequent to the introduction of dynamic carrier allocation (DCA) strategies in the pan-European Global System for Mobile communications (GSM) cellular network. The impact of the introduction in a real network of two of the more promising DCA strategies is assessed, also in terms of the entailed signalling exchanges between base stations. A performance evaluation of a realistic cellular network exploiting the above-mentioned DCA strategies is carried out, taking into account all the main GSM transmission and network aspects. The simulations have been performed by considering a realistic cell layout corresponding to the urban area of Rome, its realistic propagation environment, and an user mobility model based on the building percentages in such area. The results of the performance evaluation show that, by carefully choosing the values of the parameters characterizing each of the considered DCA strategies, a conspicuous increase in the number of users supported by the cellular network can be achieved.

We refer to the implementation and the performance of a specific class of dynamic carrier allocation (DCA) strategies which is suitable for being introduced in the pan-European Global System for Mobile communications (GSM) cellular network. One of the main concerns of this work is to verify if and to what extent strategies belonging to this class can offer better performance than fixed channel allocation (FCA) in realistic traffic and propagation conditions. The impact of the introduction in a real network of two of the more promising DCA strategies in the considered class is assessed. In particular, solutions are presented concerning the signaling exchanges entailed by the implementation of each of the considered DCA strategies. A performance evaluation of a realistic cellular network exploiting the above-mentioned DCA strategies is carried out, taking into account all of the main GSM transmission and network aspects. The simulations have been performed by considering a realistic cell layout corresponding to the urban area of Rome, its realistic propagation environment, and a user mobility model based on the building percentages in such an area. The results of the performance evaluation show that, by carefully choosing the values of the parameters characterizing each of the considered DCA strategies, a conspicuous increase in the number of users supported by the cellular network can be achieved.

Code-division multiple-access (CDMA) is one of the major candidate access techniques for third generation systems. A great deal of effort has been devoted to the study of the capacity it can support. This paper presents analytical derivations which allow the determination of the link availability in the presence of user mobility and power control imperfections in a CDMA network; moreover, it provides the guidelines which permit the implementation of a simple and flexible simulation tool which is independent of the specific CDMA implementations. As a matter of fact, the reported concepts can be applied to any asynchronous CDMA system, i.e., they hold both for the American Standard IS-95 and for the European Community Standard developed in the framework of the RACE CODIT Project.

TEST 02 - Elsevier's Scopus, the largest abstract and citation database of peer-reviewed literature. Search and access research from the science, technology, medicine, social sciences and arts and humanities fields.

In third-generation mobile systems cell dimensions will become lower and lower. So, on the one hand, an higher variance of the offered traffic per cell is expected and, on the other hand, the issue of reducing the signaling exchanges among base stations (BS's) becomes very important. A solution to the above-mentioned problems is the implementation of a dynamic channel allocation (DCA) strategy with distributed control. The DCA strategy foresees that the assignment of the carriers to the various cells is dynamically rearranged on a real time basis, in order to track the traffic evolution in the various cells. The distributed control entails that carrier acquisitions (and releases) are made by the BS's with minimal signaling exchanges. A fundamental problem of DCA strategies with distributed control concerns the coordination among BS's in order to avoid conflicting acquisitions of a same carrier from two BS's located at a distance lower than the reuse distance. This paper considers two techniques for performing the above-mentioned coordination: the synchronous technique and the asynchronous technique. Both techniques are suitable for being applied in a possible evolution of the pan-European Global System for Mobile communications (GSM). These techniques base on the presence of an acquisition delay between the acquisition stimulus time and the actual acquisition time; such acquisition delay permits coordination among BS's, but worsens performance in terms of blocking and dropping probabilities. It is shown that, in most cases, the asynchronous technique entails shorter acquisition delays than the synchronous one and thus attains better performance.>

In this paper an expression of the link availability in CDMA cellular networks which takes into account user mobility is determined. The performed simulations are based on this expression and on a simple but meaningful mobility which also considers hot cells where mobile users tend to accumulate. Two mechanisms for limiting the negative effects yielded by the bursty user mobility are presented. Such mechanisms aim at preventing the accumulation of a large amount of mobile users in a given cell, thus limiting the occurrence of situations in which the self-noise is so high that the target quality requirements are not met (i.e. the link is unavailable). The consequent increase in link availability takes place at the expense of the rising of non-null blocking and dropping probabilities. In this view, an optimization process is presented which achieves the maximum capacity while satisfying the given performance targets.

A study has been carried out to ascertain the potential benefits of asynchronous transfer mode (ATM) in the context of satellite communications and to determine the most suitable system architectures. System design guidelines are examined, and the network architecture, transmission arrangements, multiplexing and switching schemes, medium access control, and protocol layering are discussed. Particular consideration is given to the definition of the onboard switch structure.>