Doctoral School

Reading date: 11/11/2024

• NAVARRO BARBOZA, HECTOR: Experimentally constrained organic aerosol chemical and absorption properties in the atmospheric chemistry MONARCH model
  
  Author: NAVARRO BARBOZA, HECTOR
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ENVIRONMENTAL ENGINEERING
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 25/09/2024
  Reading date: 11/11/2024
  Reading time: 10:00
  Reading place: Place: ETSECCPB UPC, Campus Nord Building C1. Classroom: 002 C/Jordi Girona, 1-3 08034 Barcelona
  Thesis director: JORBA CASELLAS, ORIOL | PANDOLFI, MARCO
  Committee:
       PRESIDENT: SICARD, MICHAEL
       SECRETARI: BADIA I MORAGAS, ALBA
       VOCAL: PIÑEIRO IGLESIAS, MARIA
  Thesis abstract: Organic aerosols (OA) are a significant component of particulate matter (PM) in the atmosphere, accounting for 20-90% of total PM in Europe. OA plays a fundamental role in air quality, human health, and climate change. This Ph.D. thesis aimed to advance the understanding of OA characterization and optical properties, particularly focused on light-absorbing OA, termed brown carbon (BrC). This research used detailed observations from experimental campaigns and modeling techniques. Observational data were provided by the EGAR-IDAEA group, and the modeling part was developed at the BSC-AC group. The Multiscale Online Nonhydrostatic AtmospheRe CHemistry (MONARCH) model was further developed and used for all studies in this thesis.In the first study, we analyzed the chemical composition of coarse and fine PM simulated by the MONARCH model across three different environments in northeastern Spain (Barcelona urban, Montseny regional and Montsec remote sites). The urban site was characterized by mixed biogenic and anthropogenic sources, the regional site was primarily dominated by biogenic, and the remote site showed low concentrations except during long-range transport dust episodes. Sea salt aerosols played a fundamental role in the urban environment, while mineral dust had the largest impact during spring. In the second study, analysis of carbonaceous aerosols (black carbon, BC, and OA) using three emission inventories in the MONARCH model highlighted the importance of model resolution and detailed emission estimates. Traffic was the main contributor to BC on the urban site, with significant contributions from residential wood combustion (RWC) and shipping. Discrepancies were found in spatiotemporal disaggregation and the treatment of condensables in RWC emissions, highlighting uncertainties in OA due to emission characterization and limited secondary aerosol production in the model.The third study examined the variability of OA refractive imaginary index (k) from five sources using data from 12 European monitoring stations. The model performed well in simulating OA mass concentrations, identifying residential emissions as a major source during the colder months and secondary OA (SOA) during warmer periods, with a significant contribution of shipping emissions near ports. Optimizing k values at 370 nm revealed significant variability influenced by emission sources and environmental conditions.In the fourth study, we implemented a BrC parameterization in MONARCH, including a photobleaching effect. The BrC absorption of an annual simulation was evaluated with the European observations used in the third study. Results showed underestimations of the model BrC absorption, likely due to uncertainty in the BrC emission estimates. Through sensitivity runs perturbing BrC sources the model bias improved significantly. BrC contributes to light absorption, especially during winter as a result of increased biomass burning and biofuel emissions, highlighting important implications for climate modeling.The work carried out in this Ph.D. advances the understanding of OA and BrC and highlights the need for better representation of OA emission, formation processes, and optical properties in atmospheric models. These results will aid modelers in quantifying radiative effects of OA and their climate implications.
  

• NAVARRO MUÑOZ, ANTONI: Enhancing HPC efficiency: adaptive resource management and scheduling through online monitoring and prediction systems
  
  Author: NAVARRO MUÑOZ, ANTONI
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 22/07/2024
  Reading date: 11/11/2024
  Reading time: 10:30
  Reading place: Defensa Pública Sala E106 - Edifici C6 (FIB) - Campus Nord - Barcelona
  Thesis director: BELTRAN QUEROL, VICENÇ | AYGUADÉ PARRA, EDUARD
  Committee:
       PRESIDENT: PERICAS GLEIM, MIQUEL
       SECRETARI: PEÑA MONFERRER, ANTONIO JOSE
       VOCAL: DURAN GONZÁLEZ, ALEJANDRO
  Thesis abstract: High-Performance Computing (HPC) systems continuously evolve, driven by user needs and technology trends. Over the decades, research involving HPC systems has gone from exclusively prioritizing time-to-solution and performance to including energy efficiency and system throughput as equally important objectives. Although the approaches to tackle these objectives may vary, energy efficiency and system throughput are symbiotic, as often improving the latter enhances the former. Several software components seamlessly relate to these objectives, from applications, programming models, and runtime systems to job schedulers and operating systems.Nonetheless, parallel programming models aid HPC users in achieving these objectives by abstracting the intricacies of the underlying system. Hence, runtime systems are crucial in coping with the surfaced challenges. Runtime systems can gather precise and fine-grained profiling information and leverage it to implement advanced scheduling and resource management heuristics to optimize the execution of applications.However, current implementations are either too naive and static to cope with the irregularity and dynamism of today’s applications or introduce adverse effects in the form of overhead and complexity. To overcome these drawbacks, current runtime systems should utilize techniques that optimize system throughput adaptively through informed decisions rather than statically tailoring settings per execution.This thesis’s main objective is to design and develop a precise and low-overhead online monitoring and prediction infrastructure that provides all the necessary capabilities to enhance resource management and scheduling techniques for HPC systems. Our research finds that, based on the information provided by our monitoring infrastructures, creating adaptive techniques or enhancing existing ones can improve performance and energy efficiency compared to the static methods found in the literature. Furthermore, through a novel design, our monitoring infrastructures provide accurate and fine-grained metrics and predictions with negligible overhead while in an online operating mode.Finally, our contributions demonstrate how system throughput and energy efficiency can improve by leveraging detailed information from the runtime systems and system job schedulers.
  

• QIZILBASH, MASOOMA: Computational Fluid Dynamic (CFD) Study of Polymer Microencapsulation Processes
  
  Author: QIZILBASH, MASOOMA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN POLYMERS AND BIOPOLYMERS
  Department: Department of Chemical Engineering (EQ)
  Mode: Normal
  Deposit date: 26/07/2024
  Reading date: 11/11/2024
  Reading time: 10:00
  Reading place: EEBE - Sala Polivalent - Edifici A - Campus Diagonal-Besòs (Hora: 10h (hora local) / 12h. (a Jeddah))
  Thesis director: GUARDO ZABALETA, ALFREDO DE JESUS | DEL VALLE MENDOZA, LUIS JAVIER
  Committee:
       PRESIDENT: RAMIREZ RANGEL, ELIANA
       SECRETARI: MARTÍNEZ GOMARIZ, EDUARDO
       VOCAL: NASIR, RIZWAN
  Thesis abstract: Microencapsulation process of polymers and biopolymer is the most important technique in a polymer industry to make it unique and applicable in various fields of daily life. When it comes to the accuracy of the method only Computational Fluid Dynamic (CFD) simulations or numerical solution provides the authenticity of the method. CFD analysis gives a more schematic approach instead of any other Simulation method.In this thesis detailed use of governing equations of CFD is used to predict the accuracy of the microencapsulation method. The experimentation was done in the laboratory of the department of polymers and biopolymers under the supervision of Prof Luis J. Dell Valle. By using CFD the results are more precise, clear and accurate along with the zero minimum percentage of human error. After representing a literature review for the modeling of microencapsulation process in the first chapter, the methodology about how the process began and what parameters we followed to fulfill the requirement of the project. The effect of impeller speed 5000 rpm, 7500 rpm, 1000 rpm, 12000 rpm and 15000 rpm were determined on the number density functions and particle size reduction in the microencapsulation domain.Ansys. Fluent R2 2020 was used to model the experimental data. For the purpose of modeling the particle size distribution discrete method of population balance model is used. A total of five bins were selected to perform the size distribution analyze. The reason for selecting this small diameter range is the focus and accuracy of the solver. The bins should be more than enough to cover the distribution range and small enough to save computational time. Realizable k-e, Eulerian framework and appropriate bins initial values are used to predict the multiphase behavior. Under these conditions simulation converges very well and gives good agreement with the validation and experimental data.
  

• ROMÁN GARCÍA, FERNANDO: Contribucions per a l'impuls d'un ecosistema digital de mercat just, fiable i usable
  
  Author: ROMÁN GARCÍA, FERNANDO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN NETWORK ENGINEERING
  Department: Department of Network Engineering (ENTEL)
  Mode: Normal
  Deposit date: 14/10/2024
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: HERNANDEZ SERRANO, JUAN BAUTISTA
  Committee:
       PRESIDENT: HERNÁNDEZ GAÑÁN, CARLOS
       SECRETARI: LEÓN ABARCA, OLGA
       VOCAL: AGUDO RUIZ, ISAAC
  Thesis abstract: the modern digital economy, digital marketplaces are key platforms for interactions between providers and consumers, where trust has traditionally relied on reputation mechanisms such as reviews and ratings. While these mechanisms are effective, they are susceptible to manipulation and bias, and can pose challenges for new participants attempting to build a reputation. This reliance on reputation can undermine fairness and accessibility in data marketplaces. Distributed ledger technologies (DLTs) offer an alternative by shifting reputation to certainty, acting as distributed notaries that ensure the immutability, transparency, and reliability of stored data. However, the complexity of DLTs can hinder their adoption, especially for users without specialized technical skills. This thesis aims to address these barriers by simplifying the development and use of DLT-based applications in digital data marketplaces.The research introduces protocols and tools to enhance the usability of DLTs. Key contributions include: a protocol for secure pairing between decentralized applications (DApps) and wallets, enabling the safe exchange of cryptographic material; and a non-repudiation protocol for data exchanges, ensuring that both parties fulfil their obligations without relying on a centralized trusted third party (TTP).It also explores decentralized identity management through the use of self-sovereign identities (SSIs) with OpenID Connect (OIDC), enhancing user privacy and control over their own identities. Additionally, the research presents a modular wallet that can securely synchronize in the cloud, providing users with access to a wide range of cryptographic resources. The proposed solutions, validated through formal verification and performance evaluations, demonstrate robustness and efficiency. By reducing the technical complexity of DLT-based systems and making them more accessible, this thesis contributes to the development of a more secure, fair, and reliable digital marketplace ecosystem, where trust is built on certainties rather than reputation.
  

• VALLEJO MANCERO, BERNARDO JAVIER: Highly scalable hardware architecture for real-time execution of spiking neural networks applied to neural cognitive applications
  
  Author: VALLEJO MANCERO, BERNARDO JAVIER
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ELECTRONIC ENGINEERING
  Department: Department of Electronic Engineering (EEL)
  Mode: Article-based thesis
  Deposit date: 01/10/2024
  Reading date: 11/11/2024
  Reading time: 11:00
  Reading place: ETSETB , Aula de Graus (C4002), campus nord
  Thesis director: MADRENAS BOADAS, JORDI | ZAPATA RODRÍGUEZ, MIREYA PATRICIA
  Committee:
       PRESIDENT: ITUERO HERRERO, PABLO
       SECRETARI: MORENO AROSTEGUI, JUAN MANUEL
       VOCAL: MARGARIT TAULÉ, JOSEP MARIA
  Thesis abstract: This thesis contributes to the field of neuromorphic hardware. In particular, to the significant improvement of a scalable hardware architecture, named Hardware Emulator of Evolvable Neural Spiking Systems (HEENS), for the real-time execution of spiking neural networks (SNNs) in cognitive applications. SNNs are neural networks inspired by the biological activity of the brain, designed to process discrete temporal events, offering improved capabilities for handling temporal data as well as local plasticity, thus exhibiting greater energy efficiency compared to traditional neural networks.The HEENS architecture is implemented on several AMD (Xilinx) Zynq hardware platforms, which combine ARM processing cores with programmable logic (FPGAs), known for their high flexibility and parallelism in the execution of different neural models. One of the key achievements of this thesis is the optimization of the architecture to reduce latency, minimize resource usage, increase processing capacity, overcome previous architectural issues, and improve adaptability to various cognitive applications, such as sensory processing and pattern recognition.The results obtained can be divided into architectural contributions and experimental results in real applications. In the first part, this work improves the synaptic mapping capability, develops new system configuration mechanisms, and introduces interaction with external sensors.The enhanced system integrates an HDMI interface, demonstrating real-time visualization of neural activity and the neural and synaptic parameters of the SNN allows continuous monitoring without affecting system performance. This real-time monitoring capability has been tested in multiple experiments, where a precise representation of neural activity was observed on a 1 ms time scale, considered to be real-time, with support to other time scales, both faster or slower. The use of HEENS in physical sensor processing is highlighted, where it was proven that the architecture can adapt in real-time to changes in input signals, making it an ideal platform for applications in robotics and autonomous systems.Regarding testbenches and applications, significant results are presented in the implementation of cognitive applications using HEENS. In handwritten digit recognition, the architecture showed high accuracy using SNN models with synaptic plasticity, achieving good performance in terms of processing time and energy consumption compared to other existing solutions. Another key result is the hardware emulation of neuronal cultures, comparing the behavior of simulated neural networks in vitro, in silico, and in duris silico (physical hardware). The experiments showed that HEENS is capable of faithfully replicating the behavior and statistical properties observed in real neuronal cultures, opening new possibilities for research in neuroscience and biomedicine.Finally, the thesis concludes that the HEENS architecture not only offers a flexible and efficient environment for the research and development of SNN but also enables its implementation in real-world applications demanding real-time processing with good energy performance. The advances achieved in this work represent an important step toward the creation of next-generation scalable neuromorphic systems capable of efficiently emulating complex cognitive functions.