Titular d'Universitat (Associate Professor, acredited as Full Professor)
Departament d'Arquitectura de Computadors (DAC)
Statistical Analysis of Networks and Systems (SANS) Research Group included in CDNS Research Group
Universitat Politècnica de Catalunya (UPC)
Campus Nord, Mòdul C6, despatx 214
Jordi Girona, 1-3
08034 BARCELONA, Spain
e-mail-1: jose dot maria dot barcelo at upc dot edu
e-mail-2 (alias of the previous one): joseb at ac dot upc dot edu
Tel: +34 93 40 54051
Fax: +34 93 40 17055
Biography
I am graduated in Telecommunication Engineering (equivalent to BSc and MSc) at Universitat Politècnica de Catalunya (UPC) and got the PhD degree in Telecommunication Engineering at UPC in 1998. I am also graduated in Mathematics (equivalent to BSc and MSc) at UNED.
I now am Associate Professor, and I am acredited as Full Professor (Catedrático de Universidad), at the Computer Architecture Department at UPC. I joined the Statistical Analysis of Networks and Systems (SANS) Research Group at the Computer Architecture Departament in 1993. The SANS research group, previously called CompNet , is a subgroup of the CNDS research group. From that year I have participated in several European projects such as EXPLOIT, BAF, EXPERT, NETPERF and MOEBIUS, WIDENS (FP6) and Networks of Excellence (NoE) EuroNGI (FP6), EuroNFI (FP6) and EuroNF (FP7) and H2020 CAPTOR.
In past years, I researched in Broadband Networks (my PhD was on ATM traffic control), TCP/IP protocol design, Internet Topology (BGPv4 and Internet taxonomy). I also worked in several areas related to ad hoc networks, such as ad hoc routing protocols, Wireless Sensor Networks (WSN), Smart Cities, Mobile Wireless Networks, Vehicular Networks and Delay/Disruptive Tolerant Networks. I have also worked on how to extract "human mobility patterns" (Smart Data) from mobility traces, e.g. from smartphone devices, IoT devices, etc., and on digital tracing contact protocols.
In recent years, I have been working on how to ensure the quality of data obtained in real air quality monitoring sensor networks, e.g. projects H2020 CAPTOR, SEMIOTIC (TIN2016-78473-C3-1-R), or IMAQ (PID2019-107910RB-I00). The problem arises when current low-cost sensors (LCSs) do not provide accurate data, due to the lack of calibration of low-cost air pollution sensors, or because measurements from LCSs have outliers, gaps (need for imputation of missing data), etc. LCS measurements have to follow a pipeline process to obtain good quality estimates. These sensors have to be calibrated using machine learning techniques. In addition, the sensors are quite sensitive to environmental conditions, so recalibration methods are needed. Some data are also lost or missing and signal reconstruction techniques need to be applied to predict gaps in the data, e.g., using the whole network and applying graphical signal processing (GSP) techniques. Other challenges in air quality monitoring sensor networks are the detection of outliers or the creation of proxies and virtual sensors. The techniques we use to ensure data quality are machine learning and deep learning techniques. Some of these tasks can be performed locally (in what we call the estimation layer) and others using neighbouring sensor nodes (in what we call the network-wide processing layer).
I am co-leading national project PID2022-138155OB-I00 "Data-driven techniques for improving data-quality in networks of IoT nodes (DIDEROT)" in Sept 2024 (ending in Sept 2027). DIDEROT project aims to investigate the application of data-driven techniques for obtaining maximum accuracy in the data captured by IoT devices equipped with low-cost sensors. The main outcome of the project will be techniques to optimizing data quality while reducing the acquisition, maintenance and operation costs of monitoring networks, with applications to a wide range of fields. Our research activities will produced beyond the state-of-the-art techniques applied to IoT monitoring networks by applying ML&AI techniques, such as deep learning, graphical signal processing, graph neural networks, sparse signal reconstruction and quantum algorithms, among others, to improve data quality at the capture, estimation and network-wide processing layers in IoT monitoring networks. Our research is based on the use of experimental data.
Now, our research group started regional project 2023 CLIMA 0097 "Under the skin of the city: Urban simulations for nature-based solutions (URBANAT)" in Feb 2024 (ending in Jan 2026). URBANAT project will develop a digital-twin (DT) of Barcelona city for the prediction of thermal stress and air pollution scenarios. URBANAT will leverage the use of deep neural networks (DNNs), trained and validated using high-fidelity datasets and in situ measurements, together with IoT technology, to provide reliable, real-time predictions of the urban environment.
I am also co-leading a research project as OPI (Public Research Organization) with REPSOL S.A (January/2023-June/2025) called "Quantum Cognitive Digital Industry" (CDTI 2022 missions) in which we analyze the quality of data in IoT platforms, at the edge, using quantum flavor techniques, and projects with MENSOFT in which we build a digital twin applied to industrial IoT applications, and with AKO in which we investigate smart alarm detection systems in temporal series data in IoT and device predictive maintenance systems.
If you are interested in following a PhD in our research group, please, send me an email (jose dot maria dot barcelo at upc dot edu). We regularly have access to PhD grants. We are looking for candidates with a solid background in applied mathematics, machine learning, statistics or signal processing techniques. Degrees we are considering include (but are not limited to): applied mathematics, physics, computer science, or electrical and telecommunication engineering. We are looking for candidates with a strong research orientation, and interested in the application of research to solve problems of practical interest. Contact: send CV and gradings of BSc and MSc to the attention of Prof. Jorge Garcia-Vidal and Prof. Jose M. Barcelo-Ordinas. Please send the email to both.
Clicking below in the different buttons, you may find more information about my research activities, current topics, publications, projects and services. Thanks for visiting my personal webpage.
