Search

Tesi amb menció de Doctorat Internacional, (English) In the wake of a digital revolution, contemporary society finds itself entrenched in an era where network applications' demands surpass the capabilities of conventional network management solutions. This dissertation navigates through the intricacies of modern networked environments, where traditional management approaches are falling short due to emerging applications like augmented and virtual reality, holographic telepresence, and vehicular networks, demanding ultra-low latency and robust adaptability. These evolving networks form the backbone of modern society, sustaining numerous vital services but posing elevated complexity and operational hurdles for Internet Service Providers (ISPs) and network operators. Amidst this complexity, the need for innovative solutions to optimize and manage today's networks is more pronounced than ever. A central proposition of this dissertation is the MAGNNETO framework, a groundbreaking Machine Learning (ML) based initiative that stands for Multi-Agent Graph Neural Network Optimization. This framework is at the heart of the endeavour to facilitate distributed optimization in networked scenarios. By integrating a Graph Neural Network (GNN) architecture into a Multi-Agent Reinforcement Learning (MARL) setting, it instigates a fully distributed optimization process and capitalizes on the inherent distributed nature of networked environments, hence potentially addressing scalability issues and facilitating real-time applications. This initiative is adaptable, offering versatility in addressing various use cases and showcasing robustness to meet the challenging requisites of real-world applications. A substantial contribution of this work is the successful implementation of MAGNNETO across different relevant networked cases, prominently focusing on two highly impactful scenarios within the computer network field. Initially, it re-examines the pivotal issue of Traffic Engineering (TE) optimization in ISP networks. With the g, (Español) En el contexto de una revolución digital, la sociedad contemporánea se encuentra inmersa en una era donde las demandas de las aplicaciones en red superan las capacidades de las soluciones de gestión convencionales. Precisamente, esta tesis navega a través de las complejidades de los entornos de redes modernas, donde los enfoques de gestión tradicionales se están quedando cortos debido a aplicaciones emergentes como la realidad aumentada, la realidad virtual o la telepresencia holográfica, las cuales exigen ultra baja latencia y adaptabilidad dinámica. Estas redes en evolución conforman la columna vertebral de la sociedad moderna, manteniendo numerosos servicios vitales pero planteando una complejidad elevada para los Proveedores de Servicios de Internet (ISP) y los operadores de red. En medio de esta complejidad, la necesidad de soluciones innovadoras para optimizar y gestionar las redes actuales es más evidente que nunca. Una proposición central de esta tesis es el marco MAGNNETO (Optimización Multiagente con Redes Neuronales Gráficas, en sus siglas en inglés), una iniciativa revolucionaria basada en Aprendizaje Automático (ML). Su objetivo principal es facilitar la optimización distribuida en escenarios de redes. Al integrar una arquitectura de Redes Neuronales Gráficas (GNN) en un entorno de Aprendizaje por Refuerzo Multiagente (MARL), da pie a un proceso de optimización completamente distribuido y aprovecha la naturaleza distribuida inherente de los entornos de red, abordando problemas de escalabilidad y facilitando aplicaciones en tiempo real. Esta iniciativa es adaptable, ofreciendo versatilidad para abordar varios casos de uso y mostrando robustez para cumplir con los desafiantes requisitos de aplicaciones reales. Una contribución sustancial de este trabajo es la implementación exitosa de MAGNNETO en diferentes casos relevantes de redes, centrándose prominentemente en dos escenarios altamente impactantes en el campo de las redes de computadores. En, Postprint (published version)

This draft introduces the concept of a Network Digital Twin (NDT), including the architecture as well as the interfaces. Then two specific instances of the NDT are introduced, the first one for packet networks. This produces performance estimates (delay, jitter, loss) for a packet network with a specified topology, traffic demand, and routing and scheduling configuration. Second, a NDT for optical networks, this produces transmission performance estimates of an optical network with specified optical service topologies and network equipment types, topology and status., Preprint

GNNITIOn is a powerful library that allows the user to generate quick GNNs without needing to learn to use the deep learning libraries. STGNNs (Spati-Temporal GNNs) are a variation of the GNN that takes into account the variability of the data through the time. The goal of this master thesis is to learn how the STGNNs work by developing one from literature and give IGNNITION the tools to be able to develop STGNNs.

El TFG en educación infantil que se presenta a continuación tiene como objetivo explorar y analizar el papel del trabajo con los animales de la granja en el desarrollo educativo de los niños en el ámbito de la educación infantil. Se examina la importancia de esta experiencia educativa en relación con el aprendizaje de los niños, su conexión con la naturaleza y el fomento de habilidades socioemocionales. Este TFG también aborda la importancia de la planificación y la integración de estas actividades en el currículo de educación infantil, así como la necesidad de proporcionar un entorno seguro y adecuado para los niños y los animales. Se lleva a cabo a su vez un análisis de las implicaciones prácticas y las recomendaciones para los educadores que deseen implementar actividades con animales de granja en sus aulas. Todo lo comentado anteriormente se encuentra encuadrado en un marco teórico contrastado y detallado que dota de sentido y rigor a la propuesta de intervención construida., Grado en Educación Infantil

Graph neural networks (GNN) have shown outstanding applications in fields where data is essentially represented as graphs (e.g., chemistry, biology, recommendation systems). In this vein, communication networks comprise many fundamental components that are naturally represented in a graph-structured manner (e.g., topology, routing, signal interference). This position article presents GNNs as a fundamental tool for modeling, control and management of communication networks. GNNs represent a new generation of data-driven models that can accurately learn and reproduce the complex behaviors behind real-world networks. As a result, these models can be applied to a wide variety of networking use cases, such as planning, online optimization, or troubleshooting. The main advantage of GNNs over traditional neural networks lies in their unprecedented generalization capabilities when applied to other networks and configurations unseen during training. This is a critical feature for achieving practical data-driven solutions for networking. This article starts with a brief tutorial on GNNs and some potential applications to communication networks. Then, it presents two state-of-the-art GNN models respectively applied to wired and wireless networks. Lastly, it delves into the key open challenges and opportunities yet to be explored in this novel research area., This publication is part of the Spanish I+D+i project TRAINER-A (ref. PID2020-118011GB-C21), funded by MCIN/AEI/10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA) and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund., Peer Reviewed, Postprint (author's final draft)

Network models are an essential block of modern networks. For example, they are widely used in network planning and optimization. However, as networks increase in scale and complexity, some models present limitations, such as the assumption of Markovian traffic in queuing theory models, or the high computational cost of network simulators. Recent advances in machine learning, such as Graph Neural Networks (GNN), are enabling a new generation of network models that are data-driven and can learn complex non-linear behaviors. In this paper, we present RouteNet-Fermi, a custom GNN model that shares the same goals as Queuing Theory, while being considerably more accurate in the presence of realistic traffic models. The proposed model predicts accurately the delay, jitter, and packet loss of a network. We have tested RouteNet-Fermi in networks of increasing size (up to 300 nodes), including samples with mixed traffic profiles — e.g., with complex non-Markovian models — and arbitrary routing and queue scheduling configurations. Our experimental results show that RouteNet-Fermi achieves similar accuracy as computationally-expensive packet-level simulators and scales accurately to larger networks. Our model produces delay estimates with a mean relative error of 6.24% when applied to a test dataset of 1,000 samples, including network topologies one order of magnitude larger than those seen during training. Finally, we have also evaluated RouteNet-Fermi with measurements from a physical testbed and packet traces from a real-life network., This work was supported in part by the Spanish I+D+i Project TRAINER-A funded by the Ministry of Science and Innovation (MCIN)/State Investigation Agency (AEI)/10.13039/501100011033 under Grant PID2020-118011GB-C21, in part by the Catalan Institution for Research and Advanced Studies (ICREA), and in part by the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund., Peer Reviewed, Postprint (author's final draft)

In the paper, we propose two models of Artificial Intelligence (AI) patents in European Union (EU) countries addressing spatial and temporal behaviour. In particular, the models can quantitatively describe the interaction between countries or explain the rapidly growing trends in AI patents. For spatial analysis Poisson regression is used to explain collaboration between a pair of countries measured by the number of common patents. Through Bayesian inference, we estimated the strengths of interactions between countries in the EU and the rest of the world. In particular, a significant lack of cooperation has been identified for some pairs of countries. Alternatively, an inhomogeneous Poisson process combined with the logistic curve growth accurately models the temporal behaviour by an accurate trend line. Bayesian analysis in the time domain revealed an upcoming slowdown in patenting intensity., The research was supported in part by PL-Grid Infrastructure, POWER 2014–2020 program and the Polish Ministry of Science and Higher Education with the subvention funds of the Faculty of Computer Science, Electronics and Telecommunications of AGH University., Peer Reviewed, Postprint (published version)

Current trends in networking propose the use of Machine Learning (ML) for a wide variety of network optimization tasks. As such, many efforts have been made to produce ML-based solutions for Traffic Engineering (TE), which is a fundamental problem in Internet Service Provider (ISP) networks. Nowadays, state-of-the-art TE optimizers rely on traditional optimization techniques, such as Local search, Constraint Programming, or Linear programming. In this paper, we present MAGNNETO, a distributed ML-based framework that leverages Multi-Agent Reinforcement Learning and Graph Neural Networks for distributed TE optimization. MAGNNETO deploys a set of agents across the network that learn and communicate in a distributed fashion via message exchanges between neighboring agents. Particularly, we apply this framework to optimize link weights in Open Shortest Path First (OSPF), with the goal of minimizing network congestion. In our evaluation, we compare MAGNNETO against several state-of-the-art TE optimizers in more than 75 topologies (up to 153 nodes and 354 links), including realistic traffic loads. Our experimental results show that, thanks to its distributed nature, MAGNNETO achieves comparable performance to state-of-the-art TE optimizers with significantly lower execution times. Moreover, our ML-based solution demonstrates a strong generalization capability to successfully operate in new networks unseen during training., This publication is part of the Spanish I+D+i project TRAINER-A (ref. PID2020-118011GBC21), funded by MCIN/AEI/10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA), the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia, and the European Social Fund., Peer Reviewed, Postprint (author's final draft)

Wireless communications at the chip scale emerge as a interesting complement to traditional wire-based approaches thanks to their low latency, inherent broadcast nature, and capacity to bypass pin constraints. However, as current trends push towards massive and bandwidth-hungry processor architectures, there is a need for wireless chip-scale networks that exploit and share as many channels as possible. In this context, this work addresses the issue of channel sharing by exploring the design space of multi-channel Medium Access Control (MAC) protocols for chip-scale networks. Distinct channel assignment strategies for both random access and token passing are presented and evaluated under realistic traffic patterns. It is shown that, even with the improvements enabled by the multiple channels, both protocols maintain their intrinsic advantages and disadvantages., This work is supported by the European Commission under H2020 grant WiPLASH (GA 863337) and HE grant WINC (GA 101042080)., Peer Reviewed, Postprint (author's final draft)

Petroleum substances, as archetypical UVCBs (substances of unknown or variable composition, complex reaction products, or biological substances), pose a challenge for chemical risk assessment as they contain hundreds to thousands of individual constituents. It is particularly challenging to determine the biodegradability of petroleum substances since each constituent behaves differently. Testing the whole substance provides an average biodegradation, but it would be effectively impossible to obtain all constituents and test them individually. To overcome this challenge, comprehensive two-dimensional gas chromatography (GC × GC) in combination with advanced data-handling algorithms was applied to track and calculate degradation half-times (DT50s) of individual constituents in two dispersed middle distillate gas oils in seawater. By tracking >1000 peaks (representing ∼53–54% of the total mass across the entire chromatographic area), known biodegradation patterns of oil constituents were confirmed and extended to include many hundreds not currently investigated by traditional one-dimensional GC methods. Approximately 95% of the total tracked peak mass biodegraded after 64 days. By tracking the microbial community evolution, a correlation between the presence of functional microbial communities and the observed progression of DT50s between chemical classes was demonstrated. This approach could be used to screen the persistence of GC × GC-amenable constituents of petroleum substance UVCBs.

Tesi amb menció de Doctorat Internacional, (English) In recent years, several industry sectors have adapted the Digital Twin (DT) paradigm to improve the performance of physical systems. This paradigm consists of leveraging computational methods to build high-fidelity virtual representations of a physical system or entity. The virtual replica accurately simulates or models the behavior of the physical system without altering its behavior in the real world. Since its inception, the DT has attracted the interest of both academia and industry which can be observed by the growing number of publications, processes, standards and concepts. The networking community has adapted the DT paradigm with the objective of achieving efficient control and management in modern communication networks. In this context, the Network Digital Twin (NDT) is a renovated concept of classical network modeling tools whose goal is to build accurate data-driven network models. NDTs can be applied to many fundamental networking applications. For example, the NDT allows network operators to design novel network optimization solutions, to perform troubleshooting, what-if analysis, or to plan network upgrades taking into account the network’s expected user growth. Since the interaction between the network operator with the NDT does not require access to the real-world network, the aforementioned processes can be carried out in real-time, without jeopardizing the physical network. This dissertation aims to develop new efficient real-time optimization mechanisms leveraging NDTs. Existing network optimization techniques can be generally divided among optimizer-based solutions (e.g., CP, ILP), heuristics and Machine Learning-based (ML) solutions. Optimizer-based solutions are computationally intensive and they suffer from scalability issues where the optimization time and the problem instance size scale at different speeds. The methods based on heuristics are solutions designed by human experts, making strong assumptions and simplifications on th, (Català) En els últims anys, diversos sectors industrials han adaptat el paradigma del "Digital Twin" (DT) per millorar el rendiment dels sistemes físics. Aquest paradigma consisteix en aprofitar mètodes computacionals per construir representacions virtuals d'alta fidelitat d'un sistema o entitat física. La rèplica virtual simula o modela amb precisió el comportament del sistema físic sense alterar el seu comportament en el món real. Des de la seva creació, el DT ha suscitat l'interès tant de l'acadèmia com de la indústria, el que es pot observar pel creixent nombre de publicacions, processos, estàndards i conceptes. La comunitat de xarxes ha adaptat el paradigma del DT amb l'objectiu d'aconseguir un control i una gestió eficients en les xarxes de comunicació modernes. En aquest context, el "Network Digital Twin" (NDT) és un concepte renovat de les eines de modelatge de xarxes clàssiques que té com a objectiu construir models de xarxes precisos basats en dades. Els NDT es poden aplicar a moltes aplicacions fonamentals de xarxes. Per exemple, els NDT permeten als operadors de xarxes dissenyar noves solucions d'optimització de xarxes, realitzar resolució de problemes, anàlisi de supòsits o planificar actualitzacions de xarxa tenint en compte el creixement esperat dels usuaris de la xarxa. A més, els processos esmentats es poden dur a terme en temps real sense posar en perill la xarxa física. Aquesta dissertació té com a objectiu desenvolupar nous mecanismes d'optimització en temps real eficients aprofitant els NDTs. Les tècniques d'optimització de xarxa existents es poden dividir en solucions basades en optimitzadors, solucions basades en heurístiques i solucions basades en aprenentatge automàtic (ML). Les solucions basades en optimitzadors són intensives computacionalment i pateixen problemes d'escalabilitat, on el temps d'optimització i la mida de la instància del problema escalen a diferents velocitats. Els mètodes basats en heurístiques són solucions dissenyades p, Postprint (published version)

We propose a graph neural network (GNN)-based method to predict the distribution of penalties induced by outages in communication networks, where connections are protected by resources shared between working and backup paths. The GNN-based algorithm is trained only with random graphs generated on the basis of the Barabási–Albert model. However, the results obtained show that we can accurately model the penalties in a wide range of existing topologies. We show that GNNs eliminate the need to simulate complex outage scenarios for the network topologies under study—in practice, the entire time of path placement evaluation based on the prediction is no longer than 4 ms on modern hardware. In this way, we gain up to 12 000 times in speed improvement compared to calculations based on simulations., This work was supported by the Polish Ministry of Science and Higher Education with the subvention funds of the Faculty of Computer Science, Electronics and Telecommunications of AGH University of Science and Technology (P.B., P.C.) and by the PL-Grid Infrastructure (K.R.)., Peer Reviewed, Postprint (published version)

El TFG que se presenta a continuación tiene como objetivo principal la creación de una propuesta de intervención educativa para un aula de 4º curso de educación primaria centrada en el uso de los materiales reales para la enseñanza de la primera lengua extranjera (inglés). El aula mencionada cuenta con un gran número de adaptaciones curriculares tanto significativas como no significativas y varios casos de escolarización tardía y absentismo escolar, es por ello por lo que a través de la mencionada propuesta se pretende crear un hilo conductor que permita trabajar a toda la clase en una misma dirección y a un ritmo similar, teniendo en cuenta las circunstancias de la clase y los ritmos de aprendizaje. Previo a la construcción de la construcción de la propuesta se ha llevado a cabo un proceso de búsqueda y análisis de referencias bibliográficas que fundamenten la propuesta y todo lo que se va a incluir en la misma., Grado en Educación Primaria

Analog In-Memory Computing (AIMC) is emerging as a disruptive paradigm for heterogeneous computing, potentially delivering orders of magnitude better peak performance and efficiency over traditional digital signal processing architectures on Matrix-Vector multiplication. However, to sustain this throughput in real-world applications, AIMC tiles must be supplied with data at very high bandwidth and low latency; this poses an unprecedented pressure on the on-chip communication infrastructure, which becomes the system's performance and efficiency bottleneck. In this context, the performance and plasticity of emerging on-chip wireless communication paradigms provide the required breakthrough to up-scale on-chip communication in large AIMC devices. This work presents a many-tile AIMC architecture with inter-tile wireless communication that integrates multiple heterogeneous computing clusters, embedding a mix of parallel RISC-V cores and AIMC tiles. We perform an extensive design space exploration of the proposed architecture and discuss the benefits of exploiting emerging on-chip communication technologies such as wireless transceivers in the millimeter-wave and terahertz bands., This work was supported by the WiPLASH project (g.a. 863337), founded from the European Union’s Horizon 2020 research and innovation program., Peer Reviewed, Postprint (author's final draft)

The proliferation of emergent network applications (e.g., telesurgery, metaverse) is increasing the difficulty of managing modern communication networks. These applications entail stringent network requirements (e.g., ultra-low deterministic latency), which hinders network operators to manage their resources efficiently. In this article, we introduce the network digital twin (NDT), a renovated concept of classical network modeling tools whose goal is to build accurate data-driven network models that can operate in real-time. We describe the general architecture of the NDT and argue that modern machine learning (ML) technologies enable building some of its core components. Then, we present a case study that leverages a ML-based NDT for network performance evaluation and apply it to routing optimization in a QoS-aware use case. Lastly, we describe some key open challenges and research opportunities yet to be explored to achieve effective deployment of NDTs in real-world networks., This publication is part of the Spanish I+D+i project TRAINER-A (ref.PID2020- 118011GB-C21), funded by MCIN/ AEI/10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA) and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund., Peer Reviewed, Postprint (author's final draft)

Network modeling is a critical component of Quality of Service (QoS) optimization. Current networks implement Service Level Agreements (SLA) by careful configuration of both routing and queue scheduling policies. However, existing modeling techniques are not able to produce accurate estimates of relevant SLA metrics, such as delay or jitter, in networks with complex QoS-aware queueing policies (e.g., strict priority, Weighted Fair Queueing, Deficit Round Robin). Recently, Graph Neural Networks (GNNs) have become a powerful tool to model networks since they are specifically designed to work with graph-structured data. In this paper, we propose a GNN-based network model able to understand the complex relationship between the queueing policy (scheduling algorithm and queue sizes), the network topology, the routing configuration, and the input traffic matrix. We call our model TwinNet, a Digital Twin that can accurately estimate relevant SLA metrics for network optimization. TwinNet can generalize to its input parameters, operating successfully in topologies, routing, and queueing configurations never seen during training. We evaluate TwinNet over a wide variety of scenarios with synthetic traffic and validate it with real traffic traces. Our results show that TwinNet can provide accurate estimates of end-to-end path delays in 106 unseen real-world topologies, under different queuing configurations with a Mean Absolute Percentage Error (MAPE) of 3.8%, as well as a MAPE of 6.3% error when evaluated with a real testbed. We also showcase the potential of the proposed model for SLA-driven network optimization and what-if analysis., This publication is part of the Spanish I+D+i project TRAINER-A (ref.PID2020-118011GB-C21), funded by MCIN/ AEI/, Spain10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA), Spain and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia, Spain and the European Social Fund., Peer Reviewed, Postprint (published version)

Network modeling is an essential tool for network planning and management. It allows network administrators to explore the performance of new protocols, mechanisms, or optimal configurations without the need for testing them in real production networks. Recently, Graph Neural Networks (GNNs) have emerged as a practical solution to produce network models that can learn and extract complex patterns from real data without making any assumptions. However, state-of-the-art GNN-based network models only work with traffic matrices, this is a very coarse and simplified representation of network traffic. Although this assumption has shown to work well in certain use-cases, it is a limiting factor because, in practice, networks operate with flows. In this paper, we present FlowDT a new DL-based solution designed to model computer networks at the fine-grained flow level. In our evaluation, we show how FlowDT can accurately predict relevant per-flow performance metrics with an error of 3.5%, FlowDT’s performance is also benchmarked against vanilla DL models as well as with Queuing Theory., This work has been supported by the Spanish Government through project TRAINER-A (PID2020-118011GB-C21) with FEDER contribution and the Catalan Institution for Research and Advanced Studies (ICREA)., Peer Reviewed, Postprint (author's final draft)

The Border Gateway Protocol (BGP) is central to the global connectivity of the Internet, enabling fast and efficient dissemination of routing information. Hence, detecting any anomaly concerning BGP announcements is of critical importance to ensure the continuous operation of Internet services. Typically, BGP anomaly detection algorithms have relied on features of the BGP messages, such as the average length of the AS_PATH attribute, the volume of messages, or the type of message (announcement or withdrawal). Even though these algorithms provide good performance, they do not take into account the BGP topology, that is, the graph of ASes created by the BGP announcements. In this paper we investigate if such topology can be useful to predict BGP anomalies. We leverage Graph Neural Networks (GNN), a subset of the Neural Network (NN) family that is designed to process graph-structured data. We propose a GNN model to detect BGP anomalies and study its generalization capability. We compare its performance with two baseline models: a Support Vector Machine (SVM) and a Multilayer Perceptron (MLP), two Machine Learning (ML) techniques used in state-of-the-art solutions. Our GNN model achieves an accuracy of 79.6% using a weakly supervised dataset of 300 anomalies and is able to outperform the two baseline models., This publication is part of the Spanish I+D+i project TRAINER-A (ref.PID2020-118011GB-C21), funded by MCIN/AEI/10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA)., Peer Reviewed, Postprint (author's final draft)

Emerging applications such as the metaverse, telesurgery or cloud computing require increasingly complex operational demands on networks (e.g., ultra-reliable low latency). Likewise, the ever-faster traffic dynamics will demand network control mechanisms that can operate at short timescales (e.g., sub-minute). In this context, Traffic Engineering (TE) is a key component to efficiently control network traffic according to some performance goals (e.g., minimize network congestion).This paper presents Routing By Backprop (RBB), a novel TE method based on Graph Neural Networks (GNN) and differentiable programming. Thanks to its internal GNN model, RBB builds an end-to-end differentiable function of the target TE problem (MinMaxLoad). This enables fast TE optimization via gradient descent. In our evaluation, we show the potential of RBB to optimize OSPF-based routing (˜25% of improvement with respect to default OSPF configurations). Moreover, we test the potential of RBB as an initializer of computationally-intensive TE solvers. The experimental results show promising prospects for accelerating this type of solvers and achieving efficient online TE optimization., This work was supported by the Polish Ministry of Science and Higher Education with the subvention funds of the Faculty of Computer Science, Electronics and Telecommunications of AGH University and by the PL-Grid Infrastructure. Also, this publication is part of the Spanish I+D+i project TRAINER-A (ref. PID2020-118011GB-C21), funded by MCIN/ AEI/10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA) and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund., Peer Reviewed, Postprint (author's final draft)

The next generation of wireless communications within the framework of 6G will be operational at the low THz frequency band. Although THz systems will dramatically enhance several performance indicators such as the data rate, spectral efficiency, and latency, exploiting such technology is challenging. Electromagnetic waves confront severe propagation losses including atmospheric attenuation and diffraction. Thus, such communications are limited to line-of-sight scenarios. In 5G networks, Reconfigurable Intelligent Surfaces (RISs) are intro-duced to solve this issue by redirecting the incident wave toward the receiver and implement virtual-line-of-sight communications. In this paper, we aim to employ this paradigm for 6G networks and design a graphene-based RIS optimized to perform at multiple low atmospheric attenuation channels. We investigate the performance of this multi-wideband design through numerical and analytical analysis., Peer Reviewed, Postprint (author's final draft)

Network modeling is a fundamental tool in network research, design, and operation. Arguably the most popular method for modeling is Queuing Theory (QT). Its main limitation is that it imposes strong assumptions on the packet arrival process, which typically do not hold in real networks. In the field of Deep Learning, Graph Neural Networks (GNN) have emerged as a new technique to build data-driven models that can learn complex and non-linear behavior. In this paper, we present RouteNet-Erlang, a pioneering GNN architecture designed to model computer networks. RouteNet-Erlang supports complex traffic models, multi-queue scheduling policies, routing policies and can provide accurate estimates in networks not seen in the training phase. We benchmark RouteNet-Erlang against a state-of-the-art QT model, and our results show that it outperforms QT in all the network scenarios., This publication is part of the Spanish I+D+i project TRAINER-A (ref.PID2020-118011GB-C21), funded by MCIN/ AEI/10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA) and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund., Peer Reviewed, Postprint (author's final draft)

Wide Area Networks (WAN) are a key infrastructure in today’s society. During the last years, WANs have seen a considerable increase in network’s traffic and network applications, imposing new requirements on existing network technologies (e.g., low latency and high throughput). Consequently, Internet Service Providers (ISP) are under pressure to ensure the customer’s Quality of Service and fulfill Service Level Agreements. Network operators leverage Traffic Engineering (TE) techniques to efficiently manage the network’s resources. However, WAN’s traffic can drastically change during time and the connectivity can be affected due to external factors (e.g., link failures). Therefore, TE solutions must be able to adapt to dynamic scenarios in real-time. In this paper we propose Enero, an efficient real-time TE solution based on a two-stage optimization process. In the first one, Enero leverages Deep Reinforcement Learning (DRL) to optimize the routing configuration by generating a long-term TE strategy. To enable efficient operation over dynamic network scenarios (e.g., when link failures occur), we integrated a Graph Neural Network into the DRL agent. In the second stage, Enero uses a Local Search algorithm to improve DRL’s solution without adding computational overhead to the optimization process. The experimental results indicate that Enero is able to operate in real-world dynamic network topologies in 4.5 s on average for topologies up to 100 links., This publication is part of the Spanish I+D+i project TRAINER-A (ref.PID2020-118011GB-C21), funded by MCIN/ AEI/ 10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA) and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund., Peer Reviewed, Postprint (published version)

The recent growth of emergent network applications (e.g., satellite networks, vehicular networks) is increasing the complexity of managing modern communication networks. As a result, the community proposed the Digital Twin Networks (DTN) as a key enabler of efficient network management. Network operators can leverage the DTN to perform different optimization tasks (e.g., Traffic Engineering, Network Planning).Deep Reinforcement Learning (DRL) showed a high performance when applied to solve network optimization problems. In the context of DTN, DRL can be leveraged to solve optimization problems without directly impacting the real-world network behavior. However, DRL scales poorly with the problem size and complexity. In this paper, we explore the use of Evolutionary Strategies (ES) to train DRL agents for solving a routing optimization problem. The experimental results show that ES achieved a training time speed-up of 128 and 6 for the NSFNET and GEANT2 topologies respectively., This publication is part of the Spanish I+D+i project TRAINER-A (ref.PID2020-118011GB-C21), funded by MCIN/ AEI/10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA) and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund., Peer Reviewed, Postprint (author's final draft)

The next generation of wireless networks is expected to tap into the terahertz (THz) band (0.1–10 THz) to satisfy the extreme latency and bandwidth density requirements of future applications. However, the development of systems in this band is challenging as THz waves confront severe spreading and penetration losses, as well as molecular absorption, which leads to strong line-of-sight requirements through highly directive antennas. Recently, reconfigurable intelligent surfaces (RISs) have been proposed to address issues derived from non-line-of-sight (non-LoS) propagation, among other impairments, by redirecting the incident wave toward the receiver and implementing virtual-line-of-sight communications. However, the benefits provided by a RIS may be lost if the network operates at multiple bands. In this article, the suitability of the RIS paradigm in indoor THz scenarios for 6G is assessed grounded on the analysis of a tunable graphene-based RIS that can operate in multiple wideband transparency windows. A possible implementation of such a RIS is provided and numerically evaluated at 0.65/0.85/1.05 THz separately, demonstrating that beam steering and other relevant functionalities are realizable with excellent performance. Finally, the challenges associated with the design and fabrication of multiwideband graphene-based RISs are discussed, paving the way to the concurrent control of multiple THz bands in the context of 6G networks., Peer Reviewed, Postprint (author's final draft)

Deep Reinforcement Learning (DRL) has shown a dramatic improvement in decision-making and automated control problems. Consequently, DRL represents a promising technique to efficiently solve many relevant optimization problems (e.g., routing) in self-driving networks. However, existing DRL-based solutions applied to networking fail to generalize, which means that they are not able to operate properly when applied to network topologies not observed during training. This lack of generalization capability significantly hinders the deployment of DRL technologies in production networks. This is because state-of-the-art DRL-based networking solutions use standard neural networks (e.g., fully connected, convolutional), which are not suited to learn from information structured as graphs. In this paper, we integrate Graph Neural Networks (GNN) into DRL agents and we design a problem specific action space to enable generalization. GNNs are Deep Learning models inherently designed to generalize over graphs of different sizes and structures. This allows the proposed GNN-based DRL agent to learn and generalize over arbitrary network topologies. We test our DRL+GNN agent in a routing optimization use case in optical networks and evaluate it on 180 and 232 unseen synthetic and real-world network topologies respectively. The results show that the DRL+GNN agent is able to outperform state-of-the-art solutions in topologies never seen during training., This publication is part of the Spanish I+D+i project TRAINER-A (ref. PID2020-118011GB-C21), funded by MCIN/ AEI/10.13039/501- 100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA) and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund. This work was also supported by the Polish Ministry of Science and Higher Education with the subvention funds of the Faculty of Computer Science, Electronics and Telecommunications of AGH University and by the PL-Grid Infrastructure., Peer Reviewed, Postprint (author's final draft)

The reconfigurable intelligent surface is a promising technology for the manipulation and control of wireless electromagnetic signals. In particular, it has the potential to provide significant performance improvements for wireless networks. However, to do so, a proper reconfiguration of the reflection coefficients of unit cells is required, which often leads to complex and expensive devices. To amortize the cost, one may share the system resources among multiple transmitters and receivers. In this paper, we propose an efficient reconfiguration technique providing control over multiple beams independently. Compared to time-consuming optimization techniques, the proposed strategy utilizes an analytical method to configure the surface for multi-beam radiation. This method is easy to implement, effective and efficient since it only requires phase reconfiguration. We analyze the performance for indoor and outdoor scenarios, given the broadcast mode of operation. The aforesaid scenarios encompass some of the most challenging scenarios that wireless networks encounter. We show that our proposed technique provisions sufficient improvements in the observed channel capacity when the receivers are close to the surface in the indoor office environment scenario. Further, we report a considerable increase in the system throughput given the outdoor environment., The work of Gabriele Gradoni was supported by the Royal Society under Grant INF\R2\192066. This work was supported in part by the European Commission through the H2020 RISE-6G Project under Grant 101017011 and in part by the EPSRC under Grant EP/V048937/1., Peer Reviewed, Postprint (author's final draft)

The field of Reinforcement Learning (RL) has been receiving much attention during the last few years as a new paradigm to solve complex problems. However, one of the main issues with the current state of the art is their computational cost. Compared with other paradigms such as Supervised learning, RL requires constant interaction with the environment, which is both expensive and hard to parallelize. In this work we explore a more scalable alternative to conventional RL through the use of Evolution Strategies (ES). This consists in iteratively modifying the current solution by adding Gaussian noise to it, evaluating these modifications, and use their score to guide the improvement of the solution. The advantage of ES lies on that creating and evaluating these modifications can be parallelized. After introducing the network routing scenario, we used it to compare how ES performed against PPO, a RL policy gradient method. Ultimately ES took advantage of increasing its number of workers to eventually overtake PPO, training faster while also generating better results overall. However, it was also clear that for this to occur ES must have access to a considerable amount of hardware resources, hence being viable only within high perfomance computing environments.

Modern cloud orchestrators like Kubernetes provide a versatile and robust way to host applications at scale. One of their key features is autoscaling, which automatically adjusts cloud resources (compute, memory, storage) in order to adapt to the demands of applications. However, the scope of cloud autoscaling is limited to the datacenter hosting the cloud and it doesn't apply uniformly to the allocation of network resources. In I/O-constrained or data-in-motion use cases this can lead to severe performance degradation for the application. For example, when the load on a cloud service increases and the Wide Area Network (WAN) connecting the datacenter to the Internet becomes saturated, the application flows experience an increase in delay and loss. In many cases this is dealt with overprovisioning network capacity, which introduces additional costs and inefficiencies. On the other hand, thanks to the concept of "Network as Code", the WAN exposes a set of APIs that can be used to dynamically allocate and de-allocate capacity on-demand. In this paper we propose extending the concept of cloud autoscaling into the network to address this limitation. This way, applications running in the cloud can communicate their networking requirements, like bandwidth or traffic profile, to a Software-Defined Networking (SDN) controller or Network as a Service (NaaS) platform. Moreover, we aim to define the concepts of vertical and horizontal autoscaling applied to networking. We present a prototype that automatically allocates bandwidth to the underlay network, according to the requirements of the applications hosted in Kubernetes. Finally, we discuss open research challenges., This work was supported by the Spanish MINECO under contract TEC2017-90034-C2-1-R (ALLIANCE), the Catalan Institution for Research and Advanced Studies (ICREA)., Peer Reviewed, Postprint (author's final draft)

WireGuard is a VPN protocol that has gained significant interest recently. Its main advantages are: (i) simple configuration (via pre-shared SSH-like public keys), (ii) mobility support, (iii) reduced codebase to ease auditing, and (iv) Linux kernel implementation that yields high performance. However, WireGuard (intentionally) lacks a control plane. This means that each peer in a WireGuard network has to be manually configured with the other peers’ public key and IP addresses, or by other means. In this paper we present an architecture based on a centralized server to automatically distribute this information. In a nutshell, first we manually establish a WireGuard tunnel to the centralized server, and ask all the peers to store their public keys and IP addresses in it. Then, WireGuard peers use this secure channel to retrieve on-demand the information for the peers they want to communicate to. Our design strives to: (i) offer a key distribution scheme simpler than PKI-based ones, (ii) limit the number of public keys sent to the peers, and (iii) reduce tunnel establishment latency by means of an UDP-based protocol. We argue that such automation can help the deployment in enterprise or ISP scenarios. We also describe in detail our implementation and analyze several performance metrics. Finally, we discuss possible improvements regarding several shortcomings we found during implementation., This work was partially supported by the Spanish MINECO under contract TEC2017-90034-C2-1-R (ALLIANCE) and the Catalan Institution for Research and Advanced Studies (ICREA)., Peer Reviewed, Postprint (author's final draft)

The last few years have seen an increasing wave of attacks with serious economic and privacy damages, which evinces the need for accurate Network Intrusion Detection Systems (NIDS). Recent works propose the use of Machine Learning (ML) techniques for building such systems (e.g., decision trees, neural networks). However, existing ML-based NIDS are barely robust to common adversarial attacks, which limits their applicability to real networks. A fundamental problem of these solutions is that they treat and classify flows independently. In contrast, in this paper we argue the importance of focusing on the structural patterns of attacks, by capturing not only the individual flow features, but also the relations between different flows (e.g., the source/destination hosts they share). To this end, we use a graph representation that keeps flow records and their relationships, and propose a novel Graph Neural Network (GNN) model tailored to process and learn from such graph-structured information. In our evaluation, we first show that the proposed GNN model achieves state-of-the-art results in the well-known CIC-IDS2017 dataset. Moreover, we assess the robustness of our solution under two common adversarial attacks, that intentionally modify the packet size and interarrival times to avoid detection. The results show that our model is able to maintain the same level of accuracy as in previous experiments, while state-of-the-art ML techniques degrade up to 50% their accuracy (F1-score) under these attacks. This unprecedented level of robustness is mainly induced by the capability of our GNN model to learn flow patterns of attacks structured as graphs., This publication is part of the Spanish I+D+i project TRAINER-A (ref. PID2020-118011GB-C21), funded by MCIN/AEI/10.13039/501100011033. This work is also partially funded by the Catalan Institution for Research and Advanced Studies (ICREA), and by the European Union’s Horizon 2020 research and innovation programme within the framework of the NGI-POINTER Project, funded under grant agreement No. 871528. This article reflects only the authors’ view; the European Commission is not responsible for any use that may be made of the information it contains., Peer Reviewed, Postprint (published version)

Recent years have seen the vast potential of graph neural networks (GNN) in many fields where data is structured as graphs (e.g., chemistry, recommender systems). In particular, GNNs are becoming increasingly popular in the field of networking, as graphs are intrinsically present at many levels (e.g., topology, routing). The main novelty of GNNs is their ability to generalize to other networks unseen during training, which is an essential feature for developing practical machine learning (ML) solutions for networking. However, implementing a functional GNN prototype is currently a cumbersome task that requires strong skills in neural network programming. This poses an important barrier to network engineers that often do not have the necessary ML expertise. In this article, we present IGNNITION, a novel open source framework that enables fast prototyping of GNNs for networking systems. IGNNITION is based on an intuitive high-level abstraction that hides the complexity behind GNNs, while still offering great flexibility to build custom GNN architectures. To showcase the versatility and performance of this framework, we implement two state-of-the-art GNN models applied to different networking use cases. Our results show that the GNN models produced by IGNNITION are equivalent in terms of accuracy and performance to their native implementations in TensorFlow., This open source project has received funding from the European Union’s Horizon 2020 research and innovation programme within the framework of the NGI-POINTER Project funded under grant agreement No. 871528. This article reflects only the author’s view; the European Commission is not responsible for any use that may be made of the information it contains. The work was also supported by the Spanish MINECO under contract TEC2017-90034-C2-1-R (ALLIANCE) and the Catalan Institution for Research and Advanced Studies (ICREA), Peer Reviewed, Postprint (author's final draft)

The Border Gateway Protocol (BGP) is the inter-domain routing protocol that glues the Internet. BGP does not incorporate security and instead, it relies on careful configuration and manual filtering to offer some protection. As a consequence, the current inter-domain routing infrastructure is partially vulnerable to prefix and path hijacks as well as in misconfigurations that results in route leaks. There are many instances of these vulnerabilities being exploited by malicious actors on the Internet, resulting in disruption of services. To address this issue the IETF has designed RPKI, a centralised trust architecture that relies on Public Key Infrastructure. RPKI has slow adoption and its centralised nature is problematic: network administrators are required to trust CAs and do not have the ultimate control of their own critical Internet resources (e.g,. IP blocks, AS Numbers). In this context, we have built the Decentralised Internet Infrastructure (DII), a distributed ledger to securely store inter-domain routing information. The main advantages of DII are (i) it offers flexible trust models where the Internet community can define the rules of a consensus algorithm that properly reflects the power balance of its members and, (ii) offers protection against vulnerabilities (path hijack and route leaks) that goes well beyond what RPKI offers. We have deployed the prototype on the wild in a worldwide testbed including 7 ASes, we will use the testbed to demonstrate in a realistic scenario how allocation and delegation of Internet resources in DII work, and how this protects ASes against artificially produced path and prefix hijack as well as a route leak., This work was partially supported by the Spanish MINECO under contract TEC2017-90034-C2-1-R (ALLIANCE) and the Catalan Institution for Research and Advanced Studies (ICREA)., Peer Reviewed, Postprint (author's final draft)

During the last decade, Machine Learning (ML) has increasingly become a hot topic in the field of Computer Networks and is expected to be gradually adopted for a plethora of control, monitoring and management tasks in real-world deployments. This poses the need to count on new generations of students, researchers and practitioners with a solid background in ML applied to networks. During 2020, the International Telecommunication Union (ITU) has organized the "ITU AI/ML in 5G challenge", an open global competition that has introduced to a broad audience some of the current main challenges in ML for networks. This large-scale initiative has gathered 23 different challenges proposed by network operators, equipment manufacturers and academia, and has attracted a total of 1300+ participants from 60+ countries. This paper narrates our experience organizing one of the proposed challenges: the "Graph Neural Networking Challenge 2020". We describe the problem presented to participants, the tools and resources provided, some organization aspects and participation statistics, an outline of the top-3 awarded solutions, and a summary with some lessons learned during all this journey. As a result, this challenge leaves a curated set of educational resources openly available to anyone interested in the topic., This work has received funding from the European Union’s H2020 research and innovation programme within the framework of the NGI-POINTER Project funded under grant agreement No. 871528. This paper reflects only the authors’ view; the European Commission is not responsible for any use that may be made of the information it contains. This work was also supported by the Spanish MINECO under contract TEC2017-90034-C2-1-R (ALLIANCE), the Catalan Institution for Research and Advanced Studies (ICREA), and by FI-AGAUR grant by the Catalan Government. Salzburg Research is grateful for the support by the WISS 2025 (Science and Innovation Strategy Salzburg 2025) project ”IDALab Salzburg” (20204-WISS/225/197-2019 and 20102-F1901166-KZP) and the 5G-AI-MLab by the Federal Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology (BMK) and the Austrian state Salzburg., Peer Reviewed, Article escrit per 24 autors/autores: José Suárez-Varela (1), Miquel Ferriol-Galmés (1), Albert López (1), Paul Almasan (1), Guillermo Bernárdez (1), David Pujol-Perich (1), Krzysztof Rusek (1,2), Loïck Bonniot (3, 4), Christoph Neumann (3), François Schnitzler (3), François Taïani (4), Martin Happ (5, 6), Christian Maier (5), Jia Lei Du (5), Matthias Herlich (5), Peter Dorfinger (5), Nick Vincent Hainke (7), Stefan Venz (7), Johannes Wegener (7), Henrike Wissing (7), Bo Wu (8), Shihan Xiao (8), Pere Barlet-Ros (1), Albert Cabellos-Aparicio (1). 1- Barcelona Neural Networking center, Universitat Politècnica de Catalunya, Spain. 2- AGH University of Science and Technology, Department of Telecommunications, Poland. 3- InterDigital, France. 4- Univ. Rennes, Inria, CNRS, IRISA, France. 5- Salzburg Research Forschungsgesellschaft mbH, Austria. 6- IDA Lab, University of Salzburg, Austria. 7- Fraunhofer HHI, Germany. 8- Network Technology Lab., Huawei Technologies Co., Ltd., China., Postprint (author's final draft)

Graph Neural Networks (GNN) have recently exploded in the Machine Learning area as a novel technique for modeling graph-structured data. This makes them especially suitable for applications in the networking field, as communication networks inherently comprise graphs at many levels (e.g., topology, routing, user connections). In this demo, we will present IGNNITION, an open-source framework for fast prototyping of GNNs applied to communication networks. This framework is especially designed for network engineers and/or researchers with limited background on neural network programming. IGNNITION comprises a set of tools and functionalities that eases and accelerates the whole implementation process, from the design of a GNN model, to its training, evaluation, debugging, and integration into larger network applications. In the demo, we will show how a user can implement a complex GNN model applied to network performance modeling (RouteNet), following three simple steps., This open-source project has received funding from the European Union’s Horizon 2020 research and innovation programme within the framework of the NGI-POINTER Project funded under grant agreement No. 871528. This article reflects only the authors’ view; the EC is not responsible for any use that may be made of the information it contains. The work was also supported by the Spanish MINECO under contract TEC2017-90034-C2-1-R (ALLIANCE) and the Catalan Institution for Research and Advanced Studies (ICREA)., Peer Reviewed, Postprint (author's final draft)

Modern cloud orchestrators like Kubernetes provide a versatile and robust way to host applications at scale. One of their key features is autoscaling, that automatically adjusts cloud resources (compute, memory, storage) in order to dynamically adapt to the demands of the application. However, the scope of cloud autoscaling is limited to the datacenter hosting the cloud and it doesn't apply uniformly to the allocation of network resources. In I/O-constrained or data-in-motion use cases this can lead to severe performance degradation for the application. For example, when the load on a cloud service increases and the Wide Area Network (WAN) connecting the datacenter to the Internet becomes saturated, the application experiences an increase in delay and loss. In many cases this is dealt by overprovisioning network capacity, which introduces significant additional costs and inefficiencies. On the other hand, thanks to the concept of "Network as Code", the WAN today exposes a programmable set ofAPIs that can be used to dynamically allocate and deallocate capacity on-demand. In this paper we propose extending the concept of cloud autoscaling into the network to address this limitation. This way, applications running in the cloud can communicate their networking requirements, like bandwidth or traffic profile, to an SDN controller or Network as a Service (NaaS) platform. Moreover, we aim to define the concepts of vertical and horizontal autoscaling applied to networking. We present a prototype that automatically allocates bandwidth in the underlay of an SD-WAN, according to the requirements of the applications hosted in Kubernetes. Finally, we discuss open research challenges.

The digital transformation is pushing the existing network technologies towards new horizons, enabling new applications (e.g., vehicular networks). As a result, the networking community has seen a noticeable increase in the requirements of emerging network applications. One main open challenge is the need to accommodate control systems to highly dynamic network scenarios. Nowadays, existing network optimization technologies do not meet the needed requirements to effectively operate in real time. Some of them are based on hand-crafted heuristics with limited performance and adaptability, while some technologies use optimizers which are often too time-consuming. Recent advances in Deep Reinforcement Learning (DRL) have shown a dramatic improvement in decision-making and automated control problems. Consequently, DRL represents a promising technique to efficiently solve a variety of relevant network optimization problems, such as online routing. In this paper, we explore the use of state-of-the-art DRL technologies for real-time routing optimization and outline some relevant open challenges to achieve production-ready DRL-based solutions., This work has received funding from the European Union’s Horizon 2020 research and innovation programme within the framework of the NGI-POINTER Project funded under grant agreement No 871528. This paper reflects only the author’s view; the European Commission is not responsible for any use that may be made of the information it contains. This work was also supported by the Spanish MINECO under contract TEC2017-90034-C2-1-R (ALLIANCE), the Catalan Institution for Research and Advanced Studies (ICREA) and the Secretariat for Universities and Research of the Ministry of Business and Knowledge of the Government of Catalonia and the European Social Fund., Peer Reviewed, Postprint (author's final draft)

Leaving the current 4th generation of mobile communications behind, 5G will represent a disruptive paradigm shift integrating 5G Radio Access Networks (RANs), ultra-high-capacity access/metro/core optical networks, and intra-datacentre (DC) network and computational resources into a single converged 5G network infrastructure. The present paper overviews the main achievements obtained in the ALLIANCE project. This project ambitiously aims at architecting a converged 5G-enabled network infrastructure satisfying those needs to effectively realise the envisioned upcoming Digital Society. In particular, we present two networking solutions for 5G and beyond 5G (B5G), such as Software Defined Networking/Network Function Virtualisation (SDN/NFV) on top of an ultra-high-capacity spatially and spectrally flexible all-optical network infrastructure, and the clean-slate Recursive Inter-Network Architecture (RINA) over packet networks, including access, metro, core and DC segments. The common umbrella of all these solutions is the Knowledge-Defined Networking (KDN)-based orchestration layer which, by implementing Artificial Intelligence (AI) techniques, enables an optimal end-to-end service provisioning. Finally, the cross-layer manager of the ALLIANCE architecture includes two novel elements, namely the monitoring element providing network and user data in real time to the KDN, and the blockchain-based trust element in charge of exchanging reliable and confident information with external domains., This work has been partially funded by the Spanish Ministry of Economy and Competitiveness under contract FEDER TEC2017-90034-C2 (ALLIANCE project) and by the Generalitat de Catalunya under contract 2017SGR-1037 and 2017SGR-605., Peer Reviewed, Postprint (published version)

Overlay networks are a technique to build a new network on top of an existing one. They are a key tool to add functionality to existing networks, and are used in different layers of the Internet stack for a wide variety of purposes, like confidentiality, Quality of Service, virtual networking, etc. Specifically, network overlays in the IP networking layer are widely used in some of these use cases. However, these kind of overlay networks do not have as many functionalities as overlays in other layers. For example, thanks to the Zero Trust Networking paradigm it is possible to build secure overlay networks at L7 using HTTPS. Taking this into account, this thesis strives to add new features and improve on others of IP overlay networks, in order to support emerging challenges. This thesis focuses on three axes: security, trust, and deployment in enterprise scenarios. First, regarding security, we explore how to simplify the setup of secure tunnels over the Internet, without relying on external Public Key Infrastructure or proprietary solutions. To this purpose, we leverage WireGuard, a state of the art VPN protocol, and add a control plane on top of it to distribute encryption keys. In addition, we present the implementation of a prototype and a performance evaluation. Second, with respect to trust, we investigate how emerging blockchain technology can be used in distributed mapping systems. Mapping systems are a database used in some overlay network deployments to assist in the creation of tunnels, by storing overlay to underlay pairs of addresses. Mapping systems are not commonly used in scenarios with multiple administrative domains, due to configuration complexity and centralized control. We explore how some of the properties of blockchains, such as distributed control, or auditability, can help in building these type of mapping systems. We take into account both the policy aspects, that is, the advantages of a distributed trust scheme, and the technical ones, like, Les xarxes superposades o overlays són una tècnica per a construir una xarxa sobre una d'existent. Són una eina fonamental per a afegir funcionalitat a xarxes ja existents, i es fan servir en diferents capes de la pila de protocols d’Internet per a diferents objectius, com ara confidencialitat, Qualitat de Servei, virtualització de xarxes, etc. Específicament, les xarxes superposades a la capa de xarxa IP es fan servir àmpliament per alguns d’aquest casos d’ús. No obstant, aquest tipus de xarxes superposades no tenen tanta funcionalitat com overlays en altres capes. Per exemple, gràcies al paradigma Zero Trust és possible construir xarxes superposades segures a la capa 7 amb HTTPS. Tenint això en compte, aquesta tesi vol afegir noves funcionalitats i millorar-ne d’altres en xarxes IP superposades, amb l’objectiu d’afrontar nous reptes. Aquesta tesi es centra en tres eixos: seguretat, confiança i desplegament en escenaris de xarxes empresarials. En primer lloc, pel que fa a seguretat, explorem com simplificar la configuració de túnels segurs a través d’Internet, sense dependre duna infraestructura de clau pública externa o de solucions propietàries. Per a aquest objectiu, utilitzem WireGuard, un protocol VPN d’última generació i li afegim un pla de control per a distribuir les claus d’encriptació. A més, presentem la implementació d’un prototip i una avaluació del seu rendiment. En segon lloc, respecte la confiança, investiguem com les tecnologies emergents basades en cadenes de blocs (blockchain) es poden fer servir en sistemes de mapatge distribuïts. Els sistemes de mapatge són una base de dades que es fa servir en alguns desplegaments de xarxes superposades per a ajudar en la creació dels túnels; normalment guarden parelles d’adreces que tradueixen l’adreça de la xarxa superposada a la de la xarxa de sota. Els sistemes de mapatge no es solen utilitzar en escenaris amb múltiples dominis administratius, degut a la complexitat de la configuració i a la centralització, Las redes superpuestas u overlays son una técnica para construir una nueva red encima de una ya existente. Son una herramienta clave para añadir funcionalidad a redes existentes, y se usan en diferentes capas de la pila de protocolos de Internet para una gran variedad de propósitos, como confidencialidad, Calidad de Servicio, redes virtuales, etc. Específicamente, las redes superpuestas en la capa de red IP son ampliamente usadas para algunos de estos casos de uso. No obstante, este tipo de redes no disponen de tantas funcionalidades como redes superpuestas en otras capas. Por ejemplo, gracias al paradigma Zero Trust es posible construir redes superpuestas seguras en la capa 7 usando HTTPS. Teniendo esto en cuenta, esta tesis tiene como objetivo añadir nuevas funcionalidades y mejorar otras de las redes IP superpuestas, con el propósito de afrontar los nuevos retos que van apareciendo. Esta tesis se centra en tres ejes: seguridad, confianza y despliegue en escenarios empresariales. En primer lugar, y respecto a la seguridad, exploramos cómo simplificar la configuración de túneles seguros a través de Internet, sin usar una infraestructura de clave pública externa o soluciones propietarias. Para este objetivo, utilizamos WireGuard, un protocolo VPN de última generación, y le añadimos un plano de control para distribuir las claves de encriptado. Además, presentamos la implementación de un prototipo y una evaluación de rendimiento. En segundo lugar, en relación a la confianza, investigamos como las tecnologías emergentes basadas en cadena de bloques (blockchain) se pueden usar en sistemas de mapeado distribuidos. Los sistemas de mapeado son una base de datos que se utiliza en algunas redes superpuestas para ayudar en la creación de los túneles. Normalmente, estos sistemas guardan parejas de direcciones que traducen direcciones de la red superpuesta a la de la red subyacente. Los sistemas de mapeado no son muy utilizados en escenarios con múltiples dominios administrativ, Postprint (published version)

Sixth Generation (6G) of wireless networks will be even more heterogeneous and dense as compared to Fifth Generation (5G) and other legacy networks. Thus, the 6G architecture will need to be adapted to serve the ever-evolving capacity and quality of service requirements. To satisfy these ever-increasing demands, multiple enablers such as visible light communication, light fidelity, Reconfigurable Intelligent Surfaces (RISs), THz communications, etc., have been proposed. Specifically, RISs, through their programmable characteristics, can perform the fine-grained manipulation of the radio signals being generated by the myriad transmitter devices/access points for their corresponding receivers. Such manipulations include absorption of certain components of the impinging radio signals, as well as fine-grained control of these signals in terms of direction, polarization, phase, and power in a frequency-selective manner. An RIS consists of a device that controls the behavior of the Electromagnetic (EM) waves, alongside circuits that provide the tuning mechanism and the intelligence to control it. This device that controls the EM wave behavior can be realized using Metasurfaces (MSs), which are electromagnetically thin-film and planar artificial structures that enable the control of EM fields in engineered and even atypical ways. Hence, the MS is a component of the RIS. On a more granular level, an MS is composed of subwavelength building blocks known as unit cells or meta-atoms. The design of these unit cells depends on the required EM functionality, reconfigurability, or accuracy. The promises of the RIS paradigm, therefore, come at the expense of a non-trivial complexity in the MS. On the one hand, the performance of an RIS depends on the size of the unit cells, the number of unit cell states, or the size of the whole MS. On the other hand, there are costs and energy overheads associated with the fabrication and operation of RISs that also scale with the aforementioned, La sisena generació (6G) de xarxes sense fils seran encara més denses i heterogènies que les xarxes de cinquena generació. Per tant, la arquitectura de les xarxes 6G necessitaran adaptar-se a necessitats de capacitat i qualitat de servei en constant canvi. Per servir aquestes necessitats creixents, s’han proposat diverses tecnologies com comunicacions a l’espectre visible, Light Fidelity (LiFi), superfícies reconfigurables intel·ligents (RIS, per les sigles en anglès), comunicacions a la banda dels THz, etc. En concret, les RIS han despertat interès degut a la seva capacitat de manipular de forma precisa i programàtica les senyals de ràdio generades per diversos transmissors o punts d’accés i dirigir-les a múltiples receptors. Les manipulacions esmentades inclouen l’absorció de certs components de les ones incidents, o el control de la seva direcció, polarització, fase, o amplitud de forma selectiva en freqüència. En aquest context, les metasuperfícies (MSs) són part essencial de les RIS com a estructures artificials i planars que implementen la funció electromagnètica u òptica que la RIS necessita, que pot ser fins i tot anti-natural i atípica. Així doncs, els RIS consisteixen de dispositius que controlen les ones electromagnètiques (que podrien ser MS) co-integrades amb dispositius que doten a la MS de reconfigurabilitat i intel·ligència. Les MS, perla seva banda, estan composades de elements menor de la longitud d’ona que es coneixen coma “unit cells” o “meta-atoms”. El problema radica en que la promesa de les RIS, però, comporta una complexitat a nivell de MS que no és trivial. Per una banda, el rendiment de la RIS depèn de la mida de les unit cells, el número d’estats que pugui prendre la unit cell, o la mida de la MS completa. D’altra banda, hi ha problemes de costos i consum energètic associat amb la fabricació i operació de les RIS, que també depenen dels factors abans esmentats. Aquesta tesi té com a objectiu adreçar aquests problemes proposant un mètode pe, Postprint (published version)