Doctoral School

Reading date: 23/03/2026

• KASULURU, VAISHNAVI: AI-Driven Network Service Management for Efficient and Sustainable Open-RAN Systems in 6G and Beyond
  
  Author: KASULURU, VAISHNAVI
  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 SIGNAL THEORY AND COMMUNICATIONS
  Department: Department of Signal Theory and Communications (TSC)
  Mode: Normal
  Deposit date: 24/02/2026
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: BLANCO BOTANA, LUIS | ZEYDAN, ENGIN
  Thesis abstract: This thesis presents an efficient and sustainable AI-driven resource management framework for next-generation Open Radio Access Network (O-RAN) in the context of the emerging 6G era. The framework operates in a cloud-native 6G environment and translates predictive intelligence into reliable, energy-aware orchestration. It combines advanced predictive modeling with optimization-based control to address challenges, like stochastic demand, multitenancy, and computational complexity in O-RAN. The predictive forecasting architecture is the core of the framework, quantifying uncertainty and interdependencies among network resources across multiple tenants. Probabilistic forecasting models generate distributions of future resource demands, enabling service providers to perform more informed and risk-aware resource orchestration in complex multitenant environments.Initially, the framework considered in this thesis considers univariate probabilistic estimators, including Simple-Feed-Forward (SFF), Deep Autoregressive Recurrent network (DeepAR), and Transformers, to predict individual resource demands and support effective provisioning in O-RAN. These models deliver efficient, agile, and uncertainty-aware resource predictions, which are integrated into a novel percentile-based orchestration strategy, Dynamic Percentile Adjustment Approach (DYNp). The proposed method dynamically adjusts the percentage to ensure efficient resource utilization in O-RAN systems. Selecting an appropriate percentile is critical for balancing resource waste and service reliability. However, univariate probabilistic estimates do not capture cross-resource interdependencies, leading to suboptimal decision-making. To address this limitation, the framework incorporates state-of-the-art multivariate probabilistic forecasting models such as Gaussian Process Vector Autoregression (GPVAR) and the Temporal Fusion Transformer (TFT). They jointly process multiple time series and provide robust estimates of future resource demands. These models effectively learn complex interdependencies among different resources and key parameter indicators across network slices and tenants. Furthermore, we have evaluated how low-rank approximation in GPVAR estimator enhances scalability and robustness by reducing the algorithm's training time. One of the main goals of this thesis is to achieve energy efficiency and effective resource management and sharing. By considering predictive intelligence together with power consumption, the proposed techniques proactively optimize the activation and deactivation of radio resources or radio units. This strategy significantly reduces power consumption while maintaining user experience and adhering to Service Level Agreement (SLA) guarantees. Furthermore, another relevant contribution of the thesis is the extension of the traditional cellular O-RAN architecture to include Cell-Free massive Multiple-Input and Multiple-Output (CF-mMIMO) networks, reflecting the architectural evolution for beyond 6G systems. This provides a scalable approach to ultra-dense, energy-efficient O-RAN deployments. Finally, the algorithm tools considered in the dissertation are implemented as modular applications to facilitate deployment across O-RAN. The cloud-native implementation of the forecasting and orchestration pipeline is a notable achievement. Each module has been containerized using Docker, and its functionality is exposed via Representational State Transfer (REST) APIs, such as Swagger. This enables the pipeline to operate as microservices, supporting flexible deployment, scalable execution and seamless integration within O-RAN. The thesis establishes a mathematical and architectural foundation for deploying AI-driven, sustainable, and energy-optimized O-RAN with uncertainty adaptation. It provides a basis for realizing intelligent, autonomous, and stable 6G networks and supports future research and industrial implementation of AI-powered O-RAN ecosystems.
  

• YARASCA AYBAR, CRISTIAN PEDRO: La arquitectura rural en el Qhapaq Ñan del Cusco: paisajes culturales y arquitecturas del borde
  
  Author: YARASCA AYBAR, CRISTIAN PEDRO
  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 ARCHITECTURAL DESIGN
  Department: Department of Architectural Design (PA)
  Mode: Normal
  Deposit date: 24/02/2026
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: CUROS VILA, JUAN
  Thesis abstract: The Qhapaq Ñan (Andean Road System) is the largest pre-Hispanic infrastructure in the Americas, with an estimated length of over 30,000 km, and currently occupies the territory of six South American countries (Peru, Bolivia, Ecuador, Chile, Argentina, Ecuador, and Colombia). Peruvian territory is home to most (25,000 km) of the total Andean Road System. Despite being built in the 15th century during the Inca Empire, this great road network continues to be used by many Peruvian Andean communities for their daily activities of communication, herding, and cultural festivities. For this reason, the Qhapaq Ñan was declared a World Heritage Site by UNESCO in 2014, for its exceptional transnational cultural value.The Cusco region (located in the southern Peruvian Andes) contains approximately half (13,000 km) of the total length of the Qhapaq Ñan in Peru. The sections of the Inca Trail in Cusco are found in diverse geographical contexts, starting from the main city and connecting various rural settlements in heterogeneous Andean ecosystems. These pre-Hispanic roads in rural areas are part of the social, cultural, and spatial fabric of Cusco communities, forming cultural landscapes in constant motion and rural architecture. The immediate architecture of the Qhapaq Ñan consists of roadways, side walls, stone alignments, drainage systems, bridges, stairways, and tunnels, which shape and protect Andean roads. These elements, grouped under the name of edge architecture, pass through Andean communities and are associated with their rural architecture, consisting of dwellings, agricultural plots, corrals, and archaeological sites. The edge and rural architecture associated with the Qhapaq Ñan is integrated into an Andean cultural landscape with symbolic, ritual, and natural value. However, modern dynamics such as tourist attractions, rural depopulation, and contemporary construction are part of this confrontation between historical heritage and current demands. Therefore, this thesis addresses the analysis of the architecture and cultural landscapes of the Peruvian Andes, focusing on the interactions between human communities, the built environment, the natural environment, and the Qhapaq Ñan.The thesis takes the “Cusco-La Raya” section as a case study because it is the longest and most culturally significant section in the Cusco region. Within this section, three sections declared World Heritage Sites are specifically studied: “Rumiqolqa-Chuspitakana,” “Santa Cruz de Occobamba,” and “Raqchi-Qquea.” Direct observation and recording were used through pedestrian tours of the sections studied, documentary reviews, geographic information systems, and interviews. Rural architecture associated with the Qhapaq Ñan is presented as a reflection of sociocultural and historical processes, highlighting its relevance for territorial sustainability in the contemporary context. This thesis presents the results through a series of diagrams, maps, aerial photographs, graphic dissections, and systematized architectural notes. Cultural landscapes are multifunctional spaces that combine agricultural, residential, and symbolic activities. Their preservation is essential for maintaining the balance between biodiversity and sustainable practices in rural communities. However, significant threats have been identified, such as uncontrolled urbanization, the loss of traditional knowledge, and the introduction of modern materials that alter the authenticity of rural structures. The study concludes that the preservation and revaluation of Andean architecture and cultural landscapes associated with the Qhapaq Ñan are essential to promote sustainability and social cohesion in the face of current challenges in the Peruvian Andes.
  

Reading date: 25/03/2026

• BARRERA GÓMEZ, JOSE ANTONIO: Extension of statistical methods for time series analysis with applications in environmental epidemiology
  
  Author: BARRERA GÓMEZ, JOSE ANTONIO
  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 STATISTICS AND OPERATIONS RESEARCH
  Department: Department of Statistics and Operations Research (EIO)
  Mode: Normal
  Deposit date: 23/12/2025
  Reading date: 25/03/2026
  Reading time: 11:00
  Reading place: Sala d'actes de la FMECampus Diagonal Sud, Edifici U. C. Pau Gargallo, 14 08028 Barcelona
  Thesis director: BASAGAÑA FLORES, XAVIER | GINEBRA MOLINS, JOSEP
  Thesis abstract: Part of research in environmental epidemiology focuses on the assessment of associations betweenthe exposure to environmental factors and health outcomes based on aggregated longitudinal data in a given population. To model such associations, time series analysis is typically used, in which informational units are time points (e.g. days or weeks). An aggregated measure of the outcome at each time point (e.g. yearly mean cognitive test score or daily mortality count) is linked to an aggregated measure of exposure to some environmental factor of interest at the same time point (e.g. yearly mean air pollution level or daily mean ambient temperature) using a suitable regression (e.g. linear or Poisson) model. In this context, this thesis develops two studies.The first study deals with collinearity. Distributed lag models (DLNMs) have been increasingly used to model delayed effects of environmental factors on health. DLNMs include as predictors the same exposure measured at different time points. Those lagged variables are often highly correlated resulting in correlations between the estimated regression coefficients corresponding to different lags, which can lead to unreliable results. We first illustrate such problems and then propose a visual diagnosis tool to assess consequences of such collinearity. Essentially, new values of the outcome are simulated under an alternative hypothetical effect of the exposure of interest. Then, the original model is fitted again but now using the simulated data. Finally, both original and new results are compared graphically to assess if unexpected results obtained in the original analysis could be driven by collinearity. The tool is implemented in the R package collin. We provide illustrative examples and a user’s guide.The second study extends the Poisson regression model in multi-zone time series analysis for a count outcome. Those models need to control for trends and seasonality, which can be done by including time-stratum indicators (e.g. unique combinations of year, month and day of week). That implies having to include in the model a typically high number of nuisance parameters that can cause computational issues in the estimation process. This problem can be avoided with the conditional Poisson regression model, by conditioning by the sum of the outcome event counts in each stratum, which results in a multinomial regression model. By doing this, the nuisance parameters do not need to be estimated while the model provides relative risks (e.g. change in the mean mortality for a given increase in air pollution concentration) that are adjusted for long-term trends and seasonality. In cases of data from different geographical zones, a two-stage modelling procedure is usually performed, first analysing each zone separately and then combining zone-specific results into a single overall measure using, for instance, meta-analysis. A one-stage analysis, by analysing simultaneously data from all geographic zones, could be performed by including a random effect at zone level. However, the available (frequentist) software for conditional Poisson regression does not allow including random effects. In this context, we propose and develop a one-stage modelling approach, which is computationally feasible, namely Bayesian conditional Poisson mixed model, to analyse time series data for a count outcome that analyses all zones simultaneously while maintaining the good properties of the two-stage analysis. Our approach is based on conditioning out by the sum of the outcome event counts in each zone-time stratum and the inclusion of a random effect to model zone-specific association of interest. In addition, our method allows for including a spatial structure of the random effects as well as considering potential overdispersion. In the study, we derive model equations and implement the modelling procedure in R. To facilitate usage, we develop illustrative examples and provide code and data.
  

• IAMPIERI, ARIANNA: Oriol Maspons, un archivio della memoria visiva del secondo ‘900. La fotografia come interpretazione del contesto abitato
  
  Author: IAMPIERI, ARIANNA
  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 THEORY AND HISTORY OF ARCHITECTURE
  Department: Department of History and Theory of Architecture and Communication Techniques (THATC)
  Mode: Normal
  Deposit date: 13/01/2026
  Reading date: 25/03/2026
  Reading time: 11:00
  Reading place: ETSAB (Escuela Técnica Superior de Arquitectura de Barcelona) - Planta Baja - Sala de GradosAv. Diagonal, 649-651 - 08028 - Barcelona
  Thesis director: PIZZA DE NANNO, ANTONIO | BERGERA SERRANO, JOSE IGNACIO
  Thesis abstract: Oriol Maspons was one of the leading figures of the photographic scene in Barcelona during the second half of the 20th century and a key participant in the renewal of photographic language, which at the time was still anchored in academic and traditionalist canons. Although his vast and eclectic work has been studied over time, his contribution to the representation of architectural and urban space remains largely unexplored, despite this field being a constant throughout his photographic career. This thesis aims to delve into Oriol Maspons’s perspective on the inhabited environment, thereby contributing to the fascinating study of the intersections between photography, architecture, and urban space.Maspons's interest in architecture and the city emerged from the early years of his amateur practice and solidified over time, as demonstrated by the numerous professional collaborations he undertook, sometimes independently, sometimes together with his associate Julio Ubiña, with whom he founded a photography studio in 1957. Maspons and Ubiña produced photo reports for several architects and editorial projects in the field, including Cuadernos de Arquitectura (the journal of the Col·legi d’Arquitectes de Catalunya - COAC), documenting major architectural works and the ongoing urban transformation; another significant project was the work he carried out for the volume Arquitectura Española Contemporánea by Lluís Domènech i Girbau.In addition to being a visual witness to the architectural production of his time, Maspons also turned his camera toward historical heritage, capturing modernist and Gothic architecture, the latter being the focus of the photographic book Arquitectura gótica catalana, featuring texts by Alexandre Cirici and graphic design by architects Oscar Tusquets and Lluís Clotet.In parallel, Maspons visually narrated his city, Barcelona, through tourist guides such as Això també és Barcelona (in collaboration with Ubiña) and Barcelona pam a pam, contributing to the dissemination of a previously unseen image of the city. During the Olympic period, he worked alongside other photographers such as Francesc Català-Roca, Colita, and Xavier Miserachs, offering valuable testimony to the radical architectural and urban transformations that reshaped Barcelona in the 1980s.These examples highlight both the scope of Maspons’s work and his versatility. A deeper study of his oeuvre, which deserves greater attention, represents a significant contribution to both the history of photography and the history of Spanish architecture in the second half of the 20th century, given Maspons’s important role in the architectural discourse of the time and in shaping the visual imagination of the Catalan capital.Through the analysis of existing bibliography, the study of unpublished archival material, and the examination of publications illustrated with his photographs, the aim is to offer a broader and more in-depth perspective on the subject. The objective is to expand existing knowledge with new findings, consolidate certain positions, develop critical reflections on Maspons’s work in architectural and urban representation, and open up new research fields.
  

• KUMAR, SUNEEL: Leveraging Multi-Link Diversity for Reliable, Low-Latency Industrial Wi-Fi
  
  Author: KUMAR, SUNEEL
  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: 20/02/2026
  Reading date: 25/03/2026
  Reading time: 11:00
  Reading place: sala C4-001 de l'EETAC, campus Baix LlobregatPer videoconferència a l'enllaç: https://meet.google.com/vcm-onwi-xpc
  Thesis director: GARCIA VILLEGAS, EDUARDO | CAMPS MUR, DANIEL
  Thesis abstract: Industrial Internet of Things (IIoT) applications, such as motion control, robotics, and functional safety, are increasingly deployed over Wi-Fi to gain flexibility, mobility, and improved operational adaptability compared to wired alternatives. However, current Wi-Fi deployments struggle to deliver the strict latency, reliability, and availability guarantees demanded by these safety and time-critical workloads. This thesis addresses how Wi-Fi can be made SLA-aware, i.e., capable of meeting per-flow Service Level Agreements (SLAs) expressed in terms of delay bounds for heterogeneous industrial traffic. To this end, it investigates the combination of Time-Sensitive Networking (TSN) features, specifically Frame Replication and Elimination for Reliability (FRER) and adaptive redundancy, with emerging Wi-Fi 7/8 features such as Multi-Link Operation (MLO) and Multi-AP Coordination (MAP-Co), aiming to provide predictable and efficient wireless connectivity in demanding industrial environments.First, the thesis focuses on SLA-driven scheduling for Wi-Fi 7 MLO. An SLA-MLO scheduler is proposed that adjusts per-flow link-selection probabilities based on measured per-link delay and SLA deviation, targeting per-flow SLA compliance rather than only minimizing average delay. This is extended to a cooperative variant, C-SLA-MLO, in which multiple MLO stations share SLA status information to prioritize flows that are at risk of not meeting their SLAs. Finally, a Graph-driven Min-Max Link Scheduling (GMMLS) framework is presented for dense multi-AP deployments with overlapping basic service sets, coordinating MLO link usage across APs to minimize maximum SLA stress in the network. NS-3-based simulation results in industrial-like scenarios show that the proposed mechanisms significantly improve SLA compliance (up to ≈90% reduction in SLA deviation).However, these MLO-based mechanisms require MAC-level support in Wi-Fi chipsets and are therefore tightly coupled to vendor implementations and future Wi-Fi 7/8 hardware. To complement this long-term direction with a solution that can be deployed immediately on existing Wi-Fi infrastructure, the second part of the thesis turns to TSN mechanisms operating above the MAC layer.The second part of the thesis evaluates TSN FRER for safety-critical applications such as Open Platform Communications Unified Architecture (OPC UA) Safety traffic over Wi-Fi. Using an emulation framework that couples real OPC UA Safety stacks with Wi-Fi models, the thesis shows that FRER over disjoint WLAN paths can drastically reduce safety-message round-trip times and raise communication availability from nearly zero in a single-WLAN configuration to above 90% under high load, at the expense of significant spectrum overhead. As a next step to address this overhead, an SLA-Aware Adaptive Redundancy (SAAR) scheduler is introduced, which dynamically activates FRER only when required to protect the application’s SLA, defined in terms of delay bounds and loss/SLA violation constraints. SAAR uses application-level feedback (e.g., OPC UA Safety round-trip times and availability) and channel utilization to decide when and where to replicate packets, preserving safety-grade reliability while cutting replication overhead by nearly two orders of magnitude compared to static FRER. Overall, this thesis charts a path from today’s best-effort wireless links to tomorrow’s SLA-guaranteed industrial Wi-Fi, unifying safety-critical redundancy, multi-link operation, and multi-AP coordination into a coherent toolkit for heterogeneous IIoT requirements.