Doctoral School

Reading date: 17/06/2026

• MARTÍ ESPELT, ANIOL: Physically consistent wireless communications with statistical channel state information
  
  Author: MARTÍ ESPELT, ANIOL
  Programme: DOCTORAL DEGREE IN SIGNAL THEORY AND COMMUNICATIONS
  Department: Department of Signal Theory and Communications (TSC)
  Mode: Normal
  Deposit date: 20/05/2026
  Reading date: 17/06/2026
  Reading time: 11:00
  Reading place: Aula de Teleensenyament, Edifici B3, Campus Nord, Barcelona
  Thesis director: RIBA SAGARRA, JAUME | LAMARCA OROZCO, M. MERITXELL
  Thesis abstract: The deployment of 5G and 6G wireless networks enforces a paradigm shift in communication system design, driven by the move towards sub-THz frequencies and extremely large antenna arrays. These advancements leave traditional channel modeling—which often relies on far-field assumptions and neglects electromagnetic interactions between antenna elements—physically inconsistent. Furthermore, the overhead associated with acquiring instantaneous channel state information (CSI) in massive multiple-input multiple-output (mMIMO) systems presents a critical bottleneck, particularly for low-latency or high-mobility communications.This thesis addresses the aforementioned challenges by developing a framework for the design and analysis of noncoherent wireless communication systems that operate solely with statistical, rather than instantaneous, CSI. The core of this work is the establishment of a physically consistent channel model that accurately incorporates the effects of near-field spherical wavefronts and mutual coupling. We demonstrate that these complex physical phenomena can be effectively captured within a correlated Rayleigh fading model, providing a tractable yet realistic foundation for system analysis.Using this framework, we investigate the performance of one-shot, energy-based communication schemes, which are particularly well suited for low-latency applications. A key result is the existence of a fundamental error floor at high signal-to-noise ratio (SNR) for constellations with more than two energy levels when no CSI is available at the transmitter. However, we also prove that this error vanishes as the number of receiver antennas grows, highlighting the channel hardening benefits of massive arrays.A widely adopted receiver in energy-based noncoherent systems is the so-called energy detector. Although it is optimal under uncorrelated fading, its performance degrades significantly in correlated channels. To address this limitation, we introduce a novel class of quadratic detectors, including the best quadratic unbiased estimator (BQUE) as well as a practical implementation called assisted BQUE. These detectors leverage statistical CSI to achieve near-optimal performance. Furthermore, two strategies for enhancing reliability are proposed and evaluated: a constellation design methodology that minimizes the analytical symbol error rate by leveraging statistical CSI at the transmitter, and a permutational index modulation (PIM) scheme that introduces coding gain with minimal complexity.Finally, the thesis explores the impact of model mismatch, revealing that noncoherent systems exhibit greater robustness to mutual coupling than their coherent counterparts. We also demonstrate that wavefront curvature can be exploited well beyond the classical Fraunhofer distance. Moreover, we show that large antenna arrays enable the multiplexing and low-complexity detection of multiple users, even when employing noncoherent processing.
  

Reading date: 18/06/2026

• AREIAS FANZERES, LEONARDO: Sound-to-Image Translation Through Direct Cross-Modal Learning: An Exploratory and Architectural Study
  
  Author: AREIAS FANZERES, LEONARDO
  Programme: DOCTORAL DEGREE IN SIGNAL THEORY AND COMMUNICATIONS
  Department: Department of Signal Theory and Communications (TSC)
  Mode: Normal
  Deposit date: 16/04/2026
  Reading date: 18/06/2026
  Reading time: 11:00
  Reading place: Aula de Teleensenyament, edifici B3, Campus Nord, Barcelona
  Thesis director:
  Thesis abstract: Environmental sound conveys rich semantic and contextual information about events, objects, and spatial dynamics. However, prevailing computational approaches to environmental audio analysis, such as Acoustic Event Detection (AED), typically reduce this complexity to discrete textual labels. While effective for automated monitoring tasks, such representations oversimplify acoustic scenes and become inadequate when auditory information must be communicated across modalities. Sound-to-image (S2I) translation offers an alternative approach in which a model synthesizes images that visually depict sound-emitting sources and their surrounding environments.This thesis introduces and advances direct sound-to-image translation, a paradigm that establishes a connection between audio and visual modalities without relying on textual mediation, class supervision, or cluster-based alignment during training. The central hypothesis is that higher-level abstractions learned by deep neural networks provide a shared semantic space in which heterogeneous modalities can connect directly, enabling the generation of images that are interpretable and semantically coherent with the source sound. Such outputs are referred to as informative, meaning that they visually communicate meaningful aspects of the acoustic event.The first part of the thesis presents, to the best of our knowledge, the first study dedicated to direct S2I translation. A densely connected generative adversarial network (GAN), conditioned on audio embeddings, is developed to synthesize images directly from sound. Because multiple plausible images may correspond to a single acoustic event, translation quality cannot be evaluated through pixel-level reconstruction. To address this challenge, an informativity-based evaluation framework is proposed, employing classifiers to determine whether generated images are interpretable and semantically coherent with the source audio. Experiments reveal that, despite the inherent difficulty of the task, the model generalizes to unseen sounds and produces informative outputs for a meaningful portion of translations. Analysis further reveals that latent bottleneck dimensionality influences translation behavior, exposing a trade-off between pixel-space convergence and informativity.Building on this foundation, the second part investigates whether attention mechanisms can strengthen cross-modal alignment. Self-attention and cross-attention modules are integrated into the generator and evaluated across multiple configurations. Results show that attention improves translation performance when applied at early stages of the network, increasing informativity relative to the purely convolutional baseline while preserving the fully direct audio–visual connection.Together, these contributions formally define, validate, and extend direct S2I translation as a distinct research paradigm. Beyond architectural design and training strategies, this thesis advances methodological principles for quantitative evaluation of cross-modal translation in the absence of deterministic visual ground truth. The findings contribute to a broader understanding of multimodal representation learning and highlight the potential of direct S2I translation for applications in multimodal interaction and accessibility-oriented technologies, particularly for enhancing situational awareness in deaf and hard-of-hearing individuals.
  

• HAULANI EL HARAK, HANAN: ESTUDIO DE LAS CAUSAS Y PROPUESTA DE MEDIDAS PARA REDUCIR LA TASA DE ABANDONO EN USUARIOS DE LENTES DE CONTACTO
  
  Author: HAULANI EL HARAK, HANAN
  Programme: DOCTORAL DEGREE IN OPTICAL ENGINEERING
  Department: Department of Optics and Optometry (OO)
  Mode: Normal
  Deposit date: 28/05/2026
  Reading date: 18/06/2026
  Reading time: 11:30
  Reading place: Auditori Joan Salvadó del Centre Universitari de la Visió. Passeig del 22 de Juliol, 660.08222 Terrassa
  Thesis director: TORRENTS GÓMEZ, AURORA | GISPETS PARCERISAS, JOAN
  Committee:
       PRESIDENT: GIRÁLDEZ FERNÁNDEZ, MARIA JESÚS
       SECRETARI: CARDONA TORRADEFLOT, GENIS
       VOCAL: DURBÁN FORNIELES, JUAN JOSÉ
  Thesis abstract: IntroductionContact lens (CL) use has become a widely accepted and effective optical option. However, a substantial proportion of users discontinue wear, often early and unplanned, due to unresolved difficulties or unmet expectations. CL dropout is a multifactorial phenomenon mainly associated with ocular discomfort, dryness, handling difficulties, and insufficient clinical follow-up, and it remains a clinical challenge despite technological advances.Work performedThis thesis analyzes CL dropout and develops prevention strategies through three main approaches: a literature review, three observational studies conducted in clinical, academic, and optical practice settings, and the design and implementation of a structured clinical guideline for fitting and follow-up applied at the CUV.ResultsDropout is confirmed as a frequent and consistent phenomenon across settings, with rates comparable to those reported in the literature. Dryness, discomfort, and foreign body sensation are the main associated factors, together with initial handling difficulties and lack of follow-up. Visual quality is not a determining factor. Structured clinical support, patient education, and standardized protocols are associated with lower dropout rates, whereas in optical practice settings, access barriers and deficiencies in the fitting process are identified. Environmental conditions may act as modulating factors.ConclusionsOcular discomfort and dryness remain the main causes of dropout, often linked to insufficient follow-up. The implementation of a structured fitting and follow-up guideline may reduce dropout and improve satisfaction and long-term retention. This thesis provides evidence on the situation in Spain and proposes a model applicable to clinical practice.
  

• KHAN, SHERAZ AHMED: A posteriori error estimates for the finite element approximation of flow problems based on the variational multiscale concept
  
  Author: KHAN, SHERAZ AHMED
  Programme: DOCTORAL DEGREE IN APPLIED MATHEMATICS
  Department: School of Mathematics and Statistics (FME)
  Mode: Normal
  Deposit date: 15/05/2026
  Reading date: 18/06/2026
  Reading time: 12:00
  Reading place: Sala d'Actes de l'FME, Edifici U, Campus Sud
  Thesis director: CODINA ROVIRA, RAMON
  Thesis abstract: Numerical methods have become popular for solving complex flow problems in recent decades. Finite element methods have attracted the attention of scientists and engineers by offering computational solutions to partial differential equations, widely used to address engineering problems. In fluid mechanics, the standard Galerkin finite element method exhibits unstable and oscillatory solutions when dealing with convection-dominated problems. To overcome these limitations, stabilized finite element methods have emerged that provide stable solutions by incorporating additional stabilization terms into the standard Galerkin formulation. Predicting and controlling the error of numerical approximations remains a computational challenge. A posteriori error estimation has made significant progress and has become an essential tool for finite element practitioners, assessing and controlling the accuracy of numerical solutions and guiding adaptive refinement strategies.The initial part of this study develops a posteriori error estimates for the convection-diffusion-reaction equation using the variational multiscale framework. We present the results of the a priori analysis and two strategies of the a posteriori error analysis for the orthogonal sub-grid scale method. Our proposal is to use a scaled norm of the sub-grid scales as an a posteriori error estimate in the stabilized norm of the problem, which provides control over the convective term. The error convergence analysis is conducted using the L2 norm and the stabilized norm. Numerical examples demonstrate the reliable performance of the proposed error estimator compared to other estimators belonging to the variational multiscale family.The second part develops a goal-oriented a posteriori error estimation framework for linear functionals in the stabilized finite element discretization of the stationary convection-diffusion-reaction equation. We propose an explicit approach using the orthogonal sub-grid scale method, based on evaluating the sub-grid scale in the quantity of interest functionals, and compare it with a duality-based method requiring the solution of an additional adjoint problem. The results indicate that both methods yield similar error estimates, whereas the VMS-based explicit approach is computationally less expensive. Numerical tests demonstrate the effectiveness of the proposed techniques in terms of the quantity of interest functionals.The third part is dedicated to developing an adaptive mesh refinement strategy guided by a posteriori error estimators for the transient Navier-Stokes equations of incompressible flows. A VMS-based a posteriori space error estimator is derived using the orthogonal sub-grid scale method, and an h-refinement strategy driven by the local error indicator is proposed and validated on benchmark flow problems.Overall, this thesis develops a posteriori error estimates for orthogonal sub-grid scale discretizations of flow problems within the variational multiscale framework. The proposed error estimation strategies are shown to be robust for applications in fluid mechanics.
  

• MAQUIRRIAIN ANTOÑANZAS, JAVIER: Algoritmos matheurísticos para la programación de mantenimientos preventivos
  
  Author: MAQUIRRIAIN ANTOÑANZAS, JAVIER
  Programme: DOCTORAL DEGREE IN SUPPLY CHAIN AND OPERATIONS MANAGEMENT
  Department: Department of Management (OE)
  Mode: Normal
  Deposit date: 05/05/2026
  Reading date: 18/06/2026
  Reading time: 10:30
  Reading place: Seminari de l'IOC, planta 11, ETSEIB.
  Thesis director: GARCÍA VILLORIA, ALBERTO | PASTOR MORENO, RAFAEL
  Thesis abstract: This doctoral thesis addresses the problem of cyclic preventive maintenance scheduling in industrial systems, which is NP-hard due to its combinatorial complexity. The problem consists of defining a maintenance policy P that schedules the maintenance of M machines over a cyclic sequence of T time periods, where at most one machine can be serviced in each period. The objective is to determine a policy that minimizes the total costs incurred. To this end, mathematical models and matheuristic algorithms are developed and evaluated.A cost structure combining linear operating costs and stepwise maintenance costs is proposed, providing a more realistic representation of behaviour observed in industrial environments. Based on this structure, two variants of the problem are defined: • Cyclic preventive maintenance scheduling with a predetermined cycle length (SECIMAP_Tpred), in which the maintenance cycle length T is fixed in advance.• Cyclic preventive maintenance scheduling with a variable cycle length (SECIMAP_Tvar), in which T is treated as a decision variable to be determined within the problem.For the SECIMAP_Tpred variant, an exact mathematical model (MMat) is formulated, together with a relaxed version (MPR) that enables the computation of lower bounds, which are subsequently used to assess the quality of the designed (mat)heuristic algorithms. Since MMat is computationally inefficient for industrial-sized instances, three matheuristics (MH1P, MH2P, and MHFI) integrating mathematical programming and heuristic strategies are developed. A sequential procedure (MHSEC) that executes the three matheuristics and retains the best solution obtained is also proposed, achieving an average deviation of at most 3.19% from the optimal solution. In addition, a matheuristic cocktail (COMH) is developed which, by applying an early termination technique based on adaptive prioritization of subalgorithms and cancellation of non-promising solutions, obtains the same results as MHSEC while reducing execution times by 53.37%.The study is then extended to the SECIMAP_Tvar variant, in which the cycle length T is optimized. For this case, a new mathematical model for the relaxed problem (MTI) is developed, together with four hybrid algorithms combining metaheuristic procedures and matheuristic methods (MTI+MHFI, ALG1P, ALG2P, and ALGFI), a sequential algorithm (ALGSEC), and an algorithm cocktail (COALG) designed to minimize computational times. The best resolution procedures for the variable-T case (ALGSEC and COALG) achieve an average deviation of at most 1.43% from the optimal solution. Furthermore, COALG attains the same solutions as ALGSEC while reducing execution times by 33.2% through the use of early termination and intelligent pruning techniques that prevent the exploration of solution configurations unable to improve upon the best solution found so far.The comparative analysis of the results obtained for both problem variants (SECIMAP_Tpred and SECIMAP_Tvar) shows that treating T as a decision variable expands the solution space and, consequently, enables the derivation of maintenance policies that achieve a 2.68% reduction in costs.Finally, the thesis presents the general conclusions, the main methodological contributions (including the introduction of stepwise cost structures or the systematic use of lower bounds computed via mathematical programming to guide heuristic search) and directions for future research.