Recent progress in hollow-core fiber (HCF) technology is raising the prospects for the deployment of this type of fiber in future infrastructure upgrades. Although earlier deployments of HCF will take advantage of its low latency property to realize point-to-point links (e.g., in the metro and aggregation network segments), HCF generations with better specifications (improving performance and deployment robustness) may become suitable for more demanding regional and long-haul network applications. In these scenarios, they may co-exist with traditional fiber types in order to ensure, at the network-level, an appropriate balance between latency reduction and the increase of capital expenses associated to both the higher cost of HCF spans and the potential reductions in reach. Therefore, fiber infrastructure owners will face a new design problem, which consists of optimizing the spans where HCF is to be used to minimize service latency while meeting optical performance degradation and cost increase bounds. This paper describes an optimal method to solve this problem, by leveraging the approximation of incoherent accumulation of noise to derive an integer linear programming (ILP) model, and two variants of a heuristic algorithm. A set of detailed simulations over a reference optical transport network highlight the optimality and scalability of the proposed method.

While there has been significant progress in evaluating and comparing different representations for learning on protein data, the role of surface-based learning approaches remains not well-understood. In particular, there is a lack of direct and fair benchmark comparison between the best available surface-based learning methods against alternative representations such as graphs. Moreover, the few existing surface-based approaches either use surface information in isolation or, at best, perform global pooling between surface and graph-based architectures. In this work, we fill this gap by first adapting a state-of-the-art surface encoder for protein learning tasks. We then perform a direct and fair comparison of the resulting method against alternative approaches within the Atom3D benchmark, highlighting the limitations of pure surface-based learning. Finally, we propose an integrated approach, which allows learned feature sharing between graphs and surface representations on the level of nodes and vertices \textit{across all layers}. We demonstrate that the resulting architecture achieves state-of-the-art results on all tasks in the Atom3D benchmark, while adhering to the strict benchmark protocol, as well as more broadly on binding site identification and binding pocket classification. Furthermore, we use coarsened surfaces and optimize our approach for efficiency, making our tool competitive in training and inference time with existing techniques. Code can be found online: https://github.com/Vincentx15/atomsurf

We introduce DrugFlow, a generative model for structure-based drug design that integrates continuous flow matching with discrete Markov bridges, demonstrating state-of-the-art performance in learning chemical, geometric, and physical aspects of three-dimensional protein-ligand data. We endow DrugFlow with an uncertainty estimate that is able to detect out-of-distribution samples. To further enhance the sampling process towards distribution regions with desirable metric values, we propose a joint preference alignment scheme applicable to both flow matching and Markov bridge frameworks. Furthermore, we extend our model to also explore the conformational landscape of the protein by jointly sampling side chain angles and molecules.

The increasing importance of satellite data downlink in remote sensing missions has underscored the role of antenna performance in determining data transfer speed and volume. The authors propose a novel approach to enhance data downlink performance in remote sensing applications for Low-Earth Orbit (LEO) satellites. The proposed method reduces the complexity of the telecommunications system by employing an X-Band digital beam steerable reflectarray for Earth Observation (EO) Synthetic Aperture Radar (SAR) and ground communications, thereby streamlining satellite operations. This method was evaluated over a simulated two-day period in a 550 km sun-synchronous orbit to estimate access time and data transfer efficiency. Our results demonstrate that beam steerable reflectarray antennas offer considerable advantages over traditional antenna solutions, with improved access time and increased data transfer efficiency. These findings hold significant implications for the design and deployment of satellite constellation systems, potentially enhancing the performance of data downlinks in remote sensing missions.

Nowadays, optical transport and data center networks utilize wavelength-division multiplexing (WDM) and band division multiplexing (BDM) as possible solutions to increase the capacity to support the ever-increasing traffic demand. It is an alternative solution to simple spatial division multiplexing (SDM) intended to utilize more than one fiber per link. In this perspective, we propose a new modular photonic integrated multi-band wavelength selective switch (WSS) to implement a novel reconfigurable optical add-drop multiplexer (ROADM) architecture. The proposed WSS can operate on a broad spectrum range, covering the three S+C+L bands, and is potentially scalable to large numbers of output fibers and routed channels while maintaining a small footprint. We investigated the network performance assessment of the ROADM architecture based on the proposed multi-band WSS on a German network, and a detailed comparison is given for both SDM and BDM solutions. Results show that the increase in network capacity by exploiting more bands in the cost-efficient BDM scenario is quite close to the SDM scenario using existing infrastructure replicated over multiple fibers.

Low latency is becoming progressively more relevant in Earth observation applications, being constellations of mi-crosatellite a forthcoming approach. Active sensors are fundamental for high-availability (24/7) with low latency, making SAR a technology of choice. The related technological challenges include satellite miniaturisation and format standardisation for versatility regarding launcher adaptation. This work proposes a constellation of 16U compatible satellites exclusively for SAR remote sensing. A dual-polarization digitally beam steerable reflectarray is proposed for SAR antenna and ground-link communications. It will then be possible to exploit different SAR imaging modes and angles of regard without changing the satellite attitude, simplifying and reducing the footprint of ADCS. By having smaller satellites with reduced deployment costs, it is possible to launch more satellites, addressing latency reduction. The dual purpose beam steerable reflectarray antenna enables high downlink data throughput with a network of simpler ground stations.

Both spatial-division multiplexing (SDM) and band-division multiplexing (BDM) emerge as possible solutions to increase the optical network capacity to support the traffic demand which has been rising over time. In this work, two different ROADM (Re-configurable Optical Add Drop Multiplexer) switching techniques, namely SDM-InS (Independent switching) and SDM-CCC (Core Continue Constant) are investigated and the resulting network capacity is compared with the BDM approach. In the BDM case, both L- and S-bands have been used in addition to C-band to increase the network capacity. The launch power is optimized to control the QoT (Quality of Transmission) summarized by the generalized SNR (GSNR) per channel. Due to: stimulated Raman scattering, frequency variation of loss, frequency variation of dispersion coefficient and noise figures, an optimum power tilt and offset are calculated for each band. We show that the total network capacity increased by $\sim 2 \times$ and $\sim 3 \times$, when using the L-band and L+S-bands in addition to the C-band, respectively, in both a reference German and a reference US network. Additionally, it was also shown that using additional bands, the increase in network capacity is close to the result of using additional optical fibers in the SDM case.