Capítols de llibre
http://hdl.handle.net/2117/3928
2024-03-29T10:14:13ZChapter 1.6: Early warning signals of Earth system tipping points
http://hdl.handle.net/2117/399119
Chapter 1.6: Early warning signals of Earth system tipping points
Boulton, Chris A.; Buxton, Joshua E.; Arellano Nava, Beatriz; Bathiany, Sebastian; Blaschke, Lana; Boers, Niklas; Dakos, Vasilis; Dylewsky, Daniel; Kéfi, Sonia; López Martínez, Carlos; Parry, Isobel; Ritchie, Paul; van der Bolt, Bregje; van der Laan, Larissa; Weinans, Els
This chapter focuses on the methods used to predict the movement of parts of the Earth system towards tipping points. It begins by introducing the theory of critical slowing down (CSD), a general phenomenon of slowing recovery from perturbations that happens in many systems being forced slowly towards a tipping point. Then, it describes the various methods that can be used to estimate the occurrence of CSD and the approach of a tipping point, beginning with methods based on changes over time in the system, spatial changes, or changes in network structure, up to more advanced modelling techniques, including AI. These ‘early warning signals’ (EWS) can be used on data from a number of different sources, be these models, field experiments or remotely sensed data from satellites. The chapter considers various case studies that use real-world observations, to show how these methods are being used to predict losses in resilience in these systems. Finally, it explores limitations and potential solutions in the field of EWS, looking ahead to advances in data availability and what this could mean for predicting the movement towards tipping in these systems in the future
2024-01-11T07:58:42ZBoulton, Chris A.Buxton, Joshua E.Arellano Nava, BeatrizBathiany, SebastianBlaschke, LanaBoers, NiklasDakos, VasilisDylewsky, DanielKéfi, SoniaLópez Martínez, CarlosParry, IsobelRitchie, Paulvan der Bolt, Bregjevan der Laan, LarissaWeinans, ElsThis chapter focuses on the methods used to predict the movement of parts of the Earth system towards tipping points. It begins by introducing the theory of critical slowing down (CSD), a general phenomenon of slowing recovery from perturbations that happens in many systems being forced slowly towards a tipping point. Then, it describes the various methods that can be used to estimate the occurrence of CSD and the approach of a tipping point, beginning with methods based on changes over time in the system, spatial changes, or changes in network structure, up to more advanced modelling techniques, including AI. These ‘early warning signals’ (EWS) can be used on data from a number of different sources, be these models, field experiments or remotely sensed data from satellites. The chapter considers various case studies that use real-world observations, to show how these methods are being used to predict losses in resilience in these systems. Finally, it explores limitations and potential solutions in the field of EWS, looking ahead to advances in data availability and what this could mean for predicting the movement towards tipping in these systems in the futureIonospheric scintillation models: An inter-comparison study using GNSS data
http://hdl.handle.net/2117/389930
Ionospheric scintillation models: An inter-comparison study using GNSS data
Camps Carmona, Adriano José; Molina Ordóñez, Carlos; González Casado, Guillermo; Juan Zornoza, José Miguel; Lemorton, Joël; Fabbro, Vincent; Mainvis, Aymeric; Barbosa, José; Orús Pérez, Raul
Existing climatological ionosphere models, for example, GISM, SCIONAV, WBMOD, and STIPEE, have known limitations that prevent their wide use. In the framework of ESA study “Radio Climatology Models of the Ionosphere: Status and Way Forward” their performance was assessed using experimental observations of ionospheric scintillation collected over the past years to evaluate their ability to properly support future missions, and eventually indicate their weaknesses for future improvements. Model limitations are more important in terms of the intensity scintillation parameter (S4). To improve them, the COSMIC model has been fit (scaling factor and offset) to the measured data, and it became the one better predicting the intensity scintillation in a statistical sense.
2023-06-29T10:13:26ZCamps Carmona, Adriano JoséMolina Ordóñez, CarlosGonzález Casado, GuillermoJuan Zornoza, José MiguelLemorton, JoëlFabbro, VincentMainvis, AymericBarbosa, JoséOrús Pérez, RaulExisting climatological ionosphere models, for example, GISM, SCIONAV, WBMOD, and STIPEE, have known limitations that prevent their wide use. In the framework of ESA study “Radio Climatology Models of the Ionosphere: Status and Way Forward” their performance was assessed using experimental observations of ionospheric scintillation collected over the past years to evaluate their ability to properly support future missions, and eventually indicate their weaknesses for future improvements. Model limitations are more important in terms of the intensity scintillation parameter (S4). To improve them, the COSMIC model has been fit (scaling factor and offset) to the measured data, and it became the one better predicting the intensity scintillation in a statistical sense.Contrastive and attention-based multiple instance learning for the prediction of sentinel lymph node status from histopathologies of primary melanoma tumours
http://hdl.handle.net/2117/384779
Contrastive and attention-based multiple instance learning for the prediction of sentinel lymph node status from histopathologies of primary melanoma tumours
Hernández Pérez, Carlos; Combalia Escudero, Marc; Puig Sardá, Susana; Malvehy Guilera, Josep; Vilaplana Besler, Verónica
Sentinel lymph node status is a crucial prognosis factor for melanomas; nonetheless, the invasive surgery required to obtain it always puts the patient at risk. In this study, we develop a Deep Learning-based approach to predict lymph node metastasis from Whole Slide Images of primary tumours. Albeit very informative, these images come with complexities that hamper their use in machine learning applications, namely their large size and limited datasets. We propose a pre-training strategy based on self-supervised contrastive learning to extract better image feature representations and an attention-based Multiple Instance Learning approach to enhance the model’s performance. With this work, we quantitatively demonstrate that combining both methods improves various classification metrics and qualitatively show that contrastive learning encourages the network to output higher attention scores to tumour tissue and lower scores to image artifacts.
2023-03-09T11:14:14ZHernández Pérez, CarlosCombalia Escudero, MarcPuig Sardá, SusanaMalvehy Guilera, JosepVilaplana Besler, VerónicaSentinel lymph node status is a crucial prognosis factor for melanomas; nonetheless, the invasive surgery required to obtain it always puts the patient at risk. In this study, we develop a Deep Learning-based approach to predict lymph node metastasis from Whole Slide Images of primary tumours. Albeit very informative, these images come with complexities that hamper their use in machine learning applications, namely their large size and limited datasets. We propose a pre-training strategy based on self-supervised contrastive learning to extract better image feature representations and an attention-based Multiple Instance Learning approach to enhance the model’s performance. With this work, we quantitatively demonstrate that combining both methods improves various classification metrics and qualitatively show that contrastive learning encourages the network to output higher attention scores to tumour tissue and lower scores to image artifacts.Design of minimum nonlinear distortion reconfigurable antennas for next-generation communication systems
http://hdl.handle.net/2117/383897
Design of minimum nonlinear distortion reconfigurable antennas for next-generation communication systems
Ramírez Arroyave, Germán Augusto; Barlabe Dalmau, Antoni; Pradell i Cara, Lluís; Araque Araque, Javier Leonardo; Cetiner, Bedri Artug; Jofre Roca, Lluís
Nonlinear effects in the radio front-end can degrade communication quality and system performance. In this paper we present a new design technique for reconfigurable antennas that minimizes the nonlinear distortion and maximizes power efficiency through the minimization of the coupling between the internal switching ports and the external feeding ports. As a nonlinear design and validation instance, we present the nonlinear characterization up to 50 GHz of a PIN diode commonly used as a switch for reconfigurable devices in the microwave band. Nonlinear models are extracted through X-parameter measurements supported by accurate calibration and de-embedding procedures. Nonlinear switch models are validated by S-parameter measurements in the low power signal regime and by harmonic measurements in the large-signal regime and are further used to predict the measured nonlinearities of a reconfigurable antenna. These models have the desired particularity of being integrated straightforwardly in the internal multi-port method formulation, which is used and extended to account for the power induced on the switching elements. A new figure of merit for the design of reconfigurable antennas is introduced—the power margin, that is, the power difference between the fed port and the switching elements, which combined with the nonlinear load models directly translates into nonlinearities and power-efficiency-related metrics. Therefore, beyond traditional antenna aspects such as port match, gain, and beam orientation, switch power criteria are included in the design methodology. Guidelines for the design of reconfigurable antennas and parasitic layers of minimum nonlinearity are provided as well as the inherent trade-offs. A particular antenna design suitable for 5G communications in the 3.5 GHz band is presented according to these guidelines, in which the specific switching states for a set of target performance metrics are obtained via a balancing of the available figures of merit with multi-objective separation criteria, which enables good control of the various design trade-offs. Average Error Vector Magnitude (EVM) and power efficiency improvement of 12 and 6 dB, respectively, are obtained with the application of this design approach. In summary, this paper introduces a new framework for the nonlinear modeling and design of reconfigurable antennas and provides a set of general-purpose tools applicable in cases beyond those used as examples and validation in this work. Additionally, the use of these models and guidelines is presented, demonstrating one of the most appealing advantages of the reconfigurable parasitic layer approach, their low nonlinearity.
2023-02-22T14:51:01ZRamírez Arroyave, Germán AugustoBarlabe Dalmau, AntoniPradell i Cara, LluísAraque Araque, Javier LeonardoCetiner, Bedri ArtugJofre Roca, LluísNonlinear effects in the radio front-end can degrade communication quality and system performance. In this paper we present a new design technique for reconfigurable antennas that minimizes the nonlinear distortion and maximizes power efficiency through the minimization of the coupling between the internal switching ports and the external feeding ports. As a nonlinear design and validation instance, we present the nonlinear characterization up to 50 GHz of a PIN diode commonly used as a switch for reconfigurable devices in the microwave band. Nonlinear models are extracted through X-parameter measurements supported by accurate calibration and de-embedding procedures. Nonlinear switch models are validated by S-parameter measurements in the low power signal regime and by harmonic measurements in the large-signal regime and are further used to predict the measured nonlinearities of a reconfigurable antenna. These models have the desired particularity of being integrated straightforwardly in the internal multi-port method formulation, which is used and extended to account for the power induced on the switching elements. A new figure of merit for the design of reconfigurable antennas is introduced—the power margin, that is, the power difference between the fed port and the switching elements, which combined with the nonlinear load models directly translates into nonlinearities and power-efficiency-related metrics. Therefore, beyond traditional antenna aspects such as port match, gain, and beam orientation, switch power criteria are included in the design methodology. Guidelines for the design of reconfigurable antennas and parasitic layers of minimum nonlinearity are provided as well as the inherent trade-offs. A particular antenna design suitable for 5G communications in the 3.5 GHz band is presented according to these guidelines, in which the specific switching states for a set of target performance metrics are obtained via a balancing of the available figures of merit with multi-objective separation criteria, which enables good control of the various design trade-offs. Average Error Vector Magnitude (EVM) and power efficiency improvement of 12 and 6 dB, respectively, are obtained with the application of this design approach. In summary, this paper introduces a new framework for the nonlinear modeling and design of reconfigurable antennas and provides a set of general-purpose tools applicable in cases beyond those used as examples and validation in this work. Additionally, the use of these models and guidelines is presented, demonstrating one of the most appealing advantages of the reconfigurable parasitic layer approach, their low nonlinearity.Telecomunicacions
http://hdl.handle.net/2117/367805
Telecomunicacions
Berenguer Sau, Jordi
L’any 2019 es va caracteritzar per la planificació de l’execució del segon dividend digital, que, a més d’una replanificació de les freqüències dels canals de televisió digital terrestre, preveu, en un termini d’uns tres anys, la renovació total del parc de receptors, tot derivat dels esforços esmerçats per al desplegament per fases de
la cinquena generació de comunicacions mòbils, l’anomenada 5G.
2022-05-27T13:30:20ZBerenguer Sau, JordiL’any 2019 es va caracteritzar per la planificació de l’execució del segon dividend digital, que, a més d’una replanificació de les freqüències dels canals de televisió digital terrestre, preveu, en un termini d’uns tres anys, la renovació total del parc de receptors, tot derivat dels esforços esmerçats per al desplegament per fases de
la cinquena generació de comunicacions mòbils, l’anomenada 5G.MRI brain tumor segmentation and uncertainty estimation using 3D-UNet architectures
http://hdl.handle.net/2117/360072
MRI brain tumor segmentation and uncertainty estimation using 3D-UNet architectures
Mora Ballestar, Laura; Vilaplana Besler, Verónica
Automation of brain tumor segmentation in 3D magnetic resonance images (MRIs) is key to assess the diagnostic and treatment of the disease. In recent years, convolutional neural networks (CNNs) have shown improved results in the task. However, high memory consumption is still a problem in 3D-CNNs. Moreover, most methods do not include uncertainty information, which is especially critical in medical diagnosis. This work studies 3D encoder-decoder architectures trained with patch-based techniques to reduce memory consumption and decrease the effect of unbalanced data. The different trained models are then used to create an ensemble that leverages the properties of each model, thus increasing the performance. We also introduce voxel-wise uncertainty information, both epistemic and aleatoric using test-time dropout (TTD) and data-augmentation (TTA) respectively. In addition, a hybrid approach is proposed that helps increase the accuracy of the segmentation. The model and uncertainty estimation measurements proposed in this work have been used in the BraTS’20 Challenge for task 1 and 3 regarding tumor segmentation and uncertainty estimation.
2022-01-20T08:46:18ZMora Ballestar, LauraVilaplana Besler, VerónicaAutomation of brain tumor segmentation in 3D magnetic resonance images (MRIs) is key to assess the diagnostic and treatment of the disease. In recent years, convolutional neural networks (CNNs) have shown improved results in the task. However, high memory consumption is still a problem in 3D-CNNs. Moreover, most methods do not include uncertainty information, which is especially critical in medical diagnosis. This work studies 3D encoder-decoder architectures trained with patch-based techniques to reduce memory consumption and decrease the effect of unbalanced data. The different trained models are then used to create an ensemble that leverages the properties of each model, thus increasing the performance. We also introduce voxel-wise uncertainty information, both epistemic and aleatoric using test-time dropout (TTD) and data-augmentation (TTA) respectively. In addition, a hybrid approach is proposed that helps increase the accuracy of the segmentation. The model and uncertainty estimation measurements proposed in this work have been used in the BraTS’20 Challenge for task 1 and 3 regarding tumor segmentation and uncertainty estimation.Security in transnational interoperable PPDR communications: threats, requirements and architecture solution
http://hdl.handle.net/2117/359533
Security in transnational interoperable PPDR communications: threats, requirements and architecture solution
Ferrús Ferré, Ramón Antonio; Sallent Roig, Oriol; Verkoelen, Cor; Fransen, Frank; Andersen, Keld; Bjerrum-Niese, Christian; Saijonmaa, Jaakko; Olivieri, Claudia; Duits, Michel; Galin, Anita; Pangallo, Franco; Proietti Modi, Debora
The relevance of cross border security operations has been identified as a priority at European level for a long time. A European network where Public Protection and Disaster Relief (PPDR) forces share communications processes and a legal framework would greatly enforce response to disaster recovery and security against crime. Nevertheless, uncertainty on costs, timescale and functionalities have slowed down the interconnection of national PPDR networks and limited the transnational cooperation of their PPDR forces so far. Currently, the European research project ISITEP is aimed at developing the legal, operational and technical framework to achieve a cost effective solution for PPDR interoperability across European countries. Inter alia, ISITEP project is specifying a new Inter-System-Interface (ISI) for the interconnection of current TETRA and TETRAPOL networks through Internet Protocol (IP) connectivity. This approach turns communications security as a central aspect. In this context, this paper describes the framework and methodology defined to carry out the development of the security requirements for the interconnection of PPDR networks via the new IP ISI and provides a discussion on the undertaken security risk and vulnerability analysis. Furthermore, an overview of the designed security architecture solution for network interconnection is provided.
2022-01-13T08:37:51ZFerrús Ferré, Ramón AntonioSallent Roig, OriolVerkoelen, CorFransen, FrankAndersen, KeldBjerrum-Niese, ChristianSaijonmaa, JaakkoOlivieri, ClaudiaDuits, MichelGalin, AnitaPangallo, FrancoProietti Modi, DeboraThe relevance of cross border security operations has been identified as a priority at European level for a long time. A European network where Public Protection and Disaster Relief (PPDR) forces share communications processes and a legal framework would greatly enforce response to disaster recovery and security against crime. Nevertheless, uncertainty on costs, timescale and functionalities have slowed down the interconnection of national PPDR networks and limited the transnational cooperation of their PPDR forces so far. Currently, the European research project ISITEP is aimed at developing the legal, operational and technical framework to achieve a cost effective solution for PPDR interoperability across European countries. Inter alia, ISITEP project is specifying a new Inter-System-Interface (ISI) for the interconnection of current TETRA and TETRAPOL networks through Internet Protocol (IP) connectivity. This approach turns communications security as a central aspect. In this context, this paper describes the framework and methodology defined to carry out the development of the security requirements for the interconnection of PPDR networks via the new IP ISI and provides a discussion on the undertaken security risk and vulnerability analysis. Furthermore, an overview of the designed security architecture solution for network interconnection is provided.RocExs 2017 field trip: rockfall risk management in the Montserrat massif
http://hdl.handle.net/2117/354622
RocExs 2017 field trip: rockfall risk management in the Montserrat massif
Janeras Casanova, Marc; Corominas Dulcet, Jordi; Jara Salvador, José Antonio; Guinau, Marta; Aguasca Solé, Alberto; Blanch, Xabier; Paret, David; Ferré, Anna; Buxó, Pere
This field trip has been devised as the closure of this sixth edition of the RocExs workshop, on the Wednesday May 24th. According to the purposes of this interdisciplinary meeting, most of the suggested topics for the workshop are tried to be present during the field trip: Rockfall characterization, inventory and mapping; testing and modeling; hazard and risk analyses; monitoring; large rockfall cases; protective measures; risk mitigation and management. We hope that you will enjoy this trip, both from technical point of view and leisure, because Montserrat Mountain is a special place where earth and heaven meet configuring captivating scenery. The rockfall risk in Montserrat must be managed properly in order to keep on enjoy-ing this cultural and natural heritage.
2021-10-26T15:04:23ZJaneras Casanova, MarcCorominas Dulcet, JordiJara Salvador, José AntonioGuinau, MartaAguasca Solé, AlbertoBlanch, XabierParet, DavidFerré, AnnaBuxó, PereThis field trip has been devised as the closure of this sixth edition of the RocExs workshop, on the Wednesday May 24th. According to the purposes of this interdisciplinary meeting, most of the suggested topics for the workshop are tried to be present during the field trip: Rockfall characterization, inventory and mapping; testing and modeling; hazard and risk analyses; monitoring; large rockfall cases; protective measures; risk mitigation and management. We hope that you will enjoy this trip, both from technical point of view and leisure, because Montserrat Mountain is a special place where earth and heaven meet configuring captivating scenery. The rockfall risk in Montserrat must be managed properly in order to keep on enjoy-ing this cultural and natural heritage.Urban applications
http://hdl.handle.net/2117/343110
Urban applications
Colin-Koeniguer, Elise; Trouve, Nicolas; Yamaguchi, Yoshio; Huang, Yue; Ferro Famil, Laurent; Navarro Sánchez, Víctor Diego; López Sánchez, Juan Manuel; Monells Miralles, Daniel; Iglesias González, Rubén; Fabregas Canovas, Francisco Javier; Mallorquí Franquet, Jordi Joan; Aguasca Solé, Alberto; López Martínez, Carlos
The experimental result reported in this chapter review the application of (high resolution) Synthetic Aperture Radar (SAR) data to extract valuable information for monitoring urban environments in space and time. Full polarimetry is particularly useful for classification, as it allows the detection of built-up areas and to discriminate among their different types exploiting the variation of the polarimetric backscatter with the orientation, shape, and distribution of buildings and houses, and street patterns. On the other hand, polarimetric SAR data acquired in interferometric configuration can be combined for 3-D rendering through coherence optimization techniques. If multiple baselines are available, direct tomographic imaging can be employed, and polarimetry both increases separation performance and characterizes the response of each scatterer. Finally, polarimetry finds also application in differential interferometry for subsidence monitoring, for instance, by improving both the number of resolution cells in which the estimate is reliable, and the quality of these estimates.
2021-04-06T06:17:46ZColin-Koeniguer, EliseTrouve, NicolasYamaguchi, YoshioHuang, YueFerro Famil, LaurentNavarro Sánchez, Víctor DiegoLópez Sánchez, Juan ManuelMonells Miralles, DanielIglesias González, RubénFabregas Canovas, Francisco JavierMallorquí Franquet, Jordi JoanAguasca Solé, AlbertoLópez Martínez, CarlosThe experimental result reported in this chapter review the application of (high resolution) Synthetic Aperture Radar (SAR) data to extract valuable information for monitoring urban environments in space and time. Full polarimetry is particularly useful for classification, as it allows the detection of built-up areas and to discriminate among their different types exploiting the variation of the polarimetric backscatter with the orientation, shape, and distribution of buildings and houses, and street patterns. On the other hand, polarimetric SAR data acquired in interferometric configuration can be combined for 3-D rendering through coherence optimization techniques. If multiple baselines are available, direct tomographic imaging can be employed, and polarimetry both increases separation performance and characterizes the response of each scatterer. Finally, polarimetry finds also application in differential interferometry for subsidence monitoring, for instance, by improving both the number of resolution cells in which the estimate is reliable, and the quality of these estimates.Basic principles of SAR polarimetry
http://hdl.handle.net/2117/342850
Basic principles of SAR polarimetry
López Martínez, Carlos; Pottier, Eric
This chapter critically summarizes the main theoretical aspects necessary for a correct processing and interpretation of the polarimetric information towards the development of applications of synthetic aperture radar (SAR) polarimetry. First of all, the basic principles of wave polarimetry (which deals with the representation and the understanding of the polarization state of an electromagnetic wave) and scattering polarimetry (which concerns inferring the properties of a target given the incident and the scattered polarized electromagnetic waves) are given. Then, concepts regarding the description of polarimetric data are reviewed, covering statistical and scattering aspects, the latter in terms of coherent and incoherent decomposition techniques. Finally, polarimetric SAR interferometry and tomography, two acquisition modes that enable the extraction of the 3-D scatterer position and separation, respectively, and their polarimetric characterization, are described.
2021-03-30T11:08:33ZLópez Martínez, CarlosPottier, EricThis chapter critically summarizes the main theoretical aspects necessary for a correct processing and interpretation of the polarimetric information towards the development of applications of synthetic aperture radar (SAR) polarimetry. First of all, the basic principles of wave polarimetry (which deals with the representation and the understanding of the polarization state of an electromagnetic wave) and scattering polarimetry (which concerns inferring the properties of a target given the incident and the scattered polarized electromagnetic waves) are given. Then, concepts regarding the description of polarimetric data are reviewed, covering statistical and scattering aspects, the latter in terms of coherent and incoherent decomposition techniques. Finally, polarimetric SAR interferometry and tomography, two acquisition modes that enable the extraction of the 3-D scatterer position and separation, respectively, and their polarimetric characterization, are described.