Beatriz Soret

In this work, we optimize unicast and multicast traffic delivery over non‐terrestrial networks (NTNs) to users on the ground through direct satellite downlink (DL). We consider a gateway‐free mesh architecture with low Earth orbit (LEO) satellites interconnected via inter‐satellite links (ISLs), where unicast and multicast services compete for both ISL and DL resources. The (E2E) problem is decomposed into: (i) a space‐segment multicast distribution problem over capacity‐constrained ISLs, modeled via Steiner‐tree approximations; and (ii) a DL resource allocation problem aligned with a 3GPP cell‐based NTN RAN model, where we formulate a demand‐aware allocation scheme for mixed multicast/unicast traffic. Results show that Steiner‐based multicast reduces ISL utilization and risk of bottlenecks, achieving up to lower load at the cost of a moderate delay increase below compared to shortest‐path unicast routing. These gains are further amplified in scenarios with spatially correlated satellites. In the DL segment, demand‐aware allocation maximizes service availability and improves throughput by up to 700% compared to a proportional fair benchmark through flexible resource allocation.

Modern Earth Observation (EO) missions generate massive volumes of imagery that challenge existing downlink and ground‐processing capabilities, particularly for time‐critical applications. This work investigates how a low Earth orbit (LEO) satellite constellation equipped with heterogeneous edge computing resources can enable real‐time semantic processing of data acquired by EO satellites. We introduce an energy‐aware framework that optimizes the use of resources accounting for data acquisition, computing, and communication constraints. Although we focus on maritime surveillance, the formulation is task‐agnostic and accommodates a broad class of semantic and goal‐oriented inference problems. Specifically, we formulate two coupled optimization problems: (i) observation scheduling, which selects image acquisition opportunities while accounting for turbulence‐induced image degradation and energy budget, and (ii) processing scheduling, which allocates semantic workloads across onboard and ground processors. We evaluate these mechanisms for the task of detection and localization of vessels, for which we quantify the benefits of turbulence‐aware observation scheduling for preserving image quality and experimentally characterize the execution‐time distribution of YOLOv8 on different computing platforms. Results demonstrate that task‐ and turbulence‐ aware observation scheduling can significantly improve the quality and quantity of observed targets. Furthermore, cooperative edge processing within the constellation substantially reduces power consumption compared with traditional downlink‐centric architectures. These findings highlight the potential of distributed edge intelligence to enhance the responsiveness and autonomy of future satellite‐based EO systems.

In this work, we explore the use of intelligent reflecting surfaces (IRSs) as transparent relays in free-space optical (FSO) links for low earth orbit (LEO) satellites constellations. We introduce a novel IRS design featuring a slightly convex surface geometry to broaden the reflected beam, thereby mitigating the beam alignment challenges typically encountered in satellite-based FSO systems. We investigate the impact of IRS size, curvature, and transmitter beam width on end-to-end system performance, taking into account attenuation, atmospheric turbulence, and pointing errors. To accommodate the effects of surface curvature, we enhance and refine the channel pointing model and derive novel closed-form expressions for key performance metrics, including bit error rate, outage probability, and ergodic capacity. Numerical results based on realistic system parameters confirm the effectiveness of the proposed approach, albeit at the cost of increased geometric channel losses. Finally, we carry out numerical simulations to validate the new channel model and derived closed-form expressions, providing valuable insights for the design of future non-terrestrial networks architectures based on FSO systems.

We introduce the concept of 1Q, the first wireless generation of integrated classical and quantum communication. 1Q features quantum base stations (QBSs) that support entanglement distribution via free-space optical links alongside traditional radio communications. Key new components include quantum cells, quantum user equipment (QUEs), and hybrid resource allocation spanning classical time-frequency and quantum entanglement domains. Several application scenarios are discussed and illustrated through system design requirements for quantum key distribution, blind quantum computing, and distributed quantum sensing. A range of unique quantum constraints are identified, including decoherence timing, fidelity requirements, and the interplay between quantum and classical error probabilities. Protocol adaptations extend cellular connection management to incorporate entanglement generation, distribution, and handover procedures, expanding the Quantum Internet to the cellular wireless.

We address the challenge of preserving the simultaneity and chronology of sensing events in multisensor systems with wireless links. The network uses temporal windows of integration (TWIs), borrowed from human multisensory perception, to preserve the temporal structure of the sensing data at the application side. We introduce a composite latency model for propagation, sensing, and communication that leads to the derivation of the probability of simultaneity violation. This is used to select the TWI duration aiming to achieve the desired degrees of chronological preservation, while maintaining the throughput of events. The letter provides important insights and analytical tools about the TWI impact on the event registration.

Open Radio Access Network (ORAN) provides a flexible framework for network slicing in 4G/5G systems, with xApps enabling efficient near-real-time resource management. The purpose of slicing is to allocate resources in a way that meets the heterogeneous demands posed by diverse services, while minimizing the amount of unused resources. We demonstrate experimentally how xApps can support slicing through near real-time control of Radio Access Network (RAN) resources and at the same time avoid resource waste due to overprovisioning. We develop an open-loop implementation for Radio Resource Management (RRM) to allocate resources efficiently across slices to serve multiple users of Virtual Reality (VR) traffic with latency guarantees. Prior research concluded that VR traffic is highly variable in its demand. We evaluate the methods’ ability to meet latency targets and minimize overprovisioning under variable traffic using specialized xApps with open-loop prediction. Our findings reveal distinct trade-offs between the methods: our proposed solution, Latency Estimation Correction (LEC), offers superior latency control but at the cost of resource efficiency; it emphasizes estimating end-to-end latency and guides the allocation accordingly. Conversely, a simpler Packet Size Prediction (PSP) approach demonstrates improved resource efficiency, though it sacrifices control over latency. The method leverages traffic forecasting to predict the required allocation. This comparison underscores the balance between latency management and resource utilization in the context of network slicing and ORAN. Aggregating users reduces resource overprovisioning by 30%, though it introduces a slight increase in end-to-end latency offset. Across target latencies ranging from 10 to 16.67 ms, our evaluations show a stable average offset of 0.6 ms.

This paper introduces a full solution for decentralized routing in Low Earth Orbit Satellite Constellations (LSatCs) based on continual Deep Reinforcement Learning (DRL), specifically designed for on-board implementation in satellites with limited computational and communication resources. This requires addressing multiple challenges, including the partial knowledge at the satellites and their continuous movement, and the time-varying sources of uncertainty in the system, such as traffic, communication links, or communication buffers. We follow a multi-agent approach, where each satellite acts as an independent decision-making agent, while acquiring a limited knowledge of the environment based on the feedback received from the nearby agents. The solution is divided into two phases. First, an offline learning phase relies on decentralized decisions and a global Deep Neural Network (DNN) trained with global experiences to learn the optimal paths. Then, the online phase with local, on-board, and pre-trained DNNs requires continual learning to evolve with the environment, which can be done in two different ways: (1) Model anticipation, where the predictable conditions of the constellation, resulting from its orbital dynamics, are exploited by each satellite sharing local model with the next satellite; and (2) Federated Learning (FL), where each agent’s model is merged first at the cluster level and then aggregated in a global Parameter Server (PS) at ground or at a geostationary orbit (GEO) satellite. Results from simulations with State-of-the-Art (SoA) constellations such show that the proposed approach converges in less than a second to similar end-to-end (E2E) latency than the shortest-path centralized approach with full knowledge of the network. Moreover, the Centered Kernel Alignment (CKA) metric quantifies the necessary alignment of the models when the dynamics of the environment change.

This article investigates a multi-tier 3D network that models a hybrid space-air-ground communication system and two main relay scenarios. In the first scenario, the aerial tier, consisting of entities such as Unmanned Aerial Vehicles (UAVs) or aircrafts, uses the space layer, represented by satellites, to connect to ground User Equipment (UE) devices or ground stations. In the second scenario, the space layer cooperates with satellites in a lower orbit or with aerial devices to communicate with ground-based devices. In both scenarios, it is assumed that the device acting as a relay is distributed throughout its tier according to a Poisson Point Process (PPP), and randomly selected from the coverage area that depends on the position of the source and destination devices. For both scenarios, the paper calculates the distribution function for the link distance, which permits to obtain the average delay of signal propagation. Further, in the paper the outage capacity and ergodic rate expressions for the amplify and forward (AF) and decode and forward (DF) relaying strategies under the shadowed-Rician fading channel model are derived. Also, the asymptotic behavior in the high-SNR regime is analyzed. These novel analytical expressions provide insights of the fundamental performance limits of integrated terrestrial and non-terrestrial wireless networks.