About

Pau Closas is an Associate Professor at the Department of Electrical and Computer Engineering at Northeastern University, Boston, MA, since 2016. He holds MS and PhD degrees in Electrical Engineering from the Universitat Politècnica de Catalunya (UPC), obtained in 2003 and 2009 respectively, and an MS degree in Advanced Mathematics and Mathematical Engineering from UPC since 2014. His primary research interests include statistical and array signal processing, estimation and detection theory, stochastic filtering, robust statistics, and game theory, with applications to positioning systems such as GNSS and indoor technologies, wireless communications, and mathematical biology. Closas has a distinguished record of contributions to navigation systems and signal processing, recognized through awards such as the EURASIP Best PhD Thesis Award, the Duran Farell Award for Technology Research, the ION Early Achievement Award, and the Harry Rowe Mimno Award. He has served in organizational roles for major conferences and is a guest editor for several IEEE publications. Currently, he is the Editor-in-Chief of the IEEE Aerospace and Electronic Systems Magazine. His research involves practical problems in satellite-based navigation, indoor positioning, wireless communications, and computational biology, leading projects funded by agencies like the National Science Foundation and DARPA.

Research topics

• Computer Science
• Artificial Intelligence
• Machine Learning
• Systems engineering
• Physics
• Real-time computing
• Remote sensing
• Algorithm
• Engineering
• Geography
• Control engineering
• Meteorology
• Telecommunications
• Psychology
• Computer vision

Selected publications

• DPMM-CFL: Clustered Federated Learning via Dirichlet Process Mixture Model Nonparametric Clustering

  2026-04-21

  articleOpen accessSenior author

  Clustered Federated Learning (CFL) improves performance under non-IID client heterogeneity by clustering clients and training one model per cluster, thereby balancing between a global model and fully personalized models. However, most CFL methods require the number of clusters K to be fixed a priori, which is impractical when the latent structure is unknown. We propose DPMM-CFL, a CFL algorithm that places a Dirichlet Process (DP) prior over the distribution of cluster parameters. This enables nonparametric Bayesian inference to jointly infer both the number of clusters and client assignments, while optimizing per-cluster federated objectives. This results in a method where, at each round, federated updates and cluster inferences are coupled, as presented in this paper. The algorithm is validated on benchmark datasets under Dirichlet and class-split non-IID partitions.

  PublisherOA PDFDOI

• Bayesian Jammer Localization with a Hybrid CNN and Path-Loss Mixture of Experts

  2026-04-21

  articleOpen accessSenior author

  Global Navigation Satellite System (GNSS) signals are vulnerable to jamming, particularly in urban areas where multipath and shadowing distort received power. Previous data-driven approaches achieved reasonable localization but poorly reconstructed the received signal strength (RSS) field due to limited spatial context. We propose a hybrid Bayesian mixture-of-experts framework that fuses a physical path-loss (PL) model and a convolutional neural network (CNN) through log-linear pooling. The PL expert ensures physical consistency, while the CNN leverages building-height maps to capture urban propagation effects. Bayesian inference with Laplace approximation provides posterior uncertainty over both the jammer position and RSS field. Experiments on urban ray-tracing data show that localization accuracy improves and uncertainty decreases with more training points, while uncertainty concentrates near the jammer and along urban canyons where propagation is most sensitive.

  PublisherOA PDFDOI

• From the Editor-in-Chief: Launching 2025

  IEEE Aerospace and Electronic Systems Magazine · 2025-01-01

  article1st authorCorresponding
  PublisherDOI

• Advances in Anti-Deception Jamming Strategies for Radar Systems: A Survey

  arXiv (Cornell University) · 2025-03-01

  preprintOpen accessSenior author

  Deception jamming has long been a significant threat to radar systems, interfering with search, acquisition, and tracking by introducing false information that diverts attention from the targets of interest. As deception strategies become more sophisticated, the vulnerability of radar systems to these attacks continues to escalate. This paper offers a comprehensive review of the evolution of anti-deception jamming techniques, starting with legacy solutions and progressing to the latest advancements. Current research is categorized into three key areas: prevention strategies, which hinder the ability of jammers to alter radar processing; detection strategies, which alert the system to deception and may classify the type of attack; and mitigation strategies, which aim to reduce or suppress the impact of jamming. Additionally, key avenues for further research are highlighted, with a particular emphasis on distributed, cognitive, and AI-enabled radar systems. We envision this paper as a gateway to the existing literature on anti-deception jamming, a critical area for safeguarding radar systems against evolving threats.

  PublisherDOI

• A Bayesian Framework for Clustered Federated Learning

  IEEE Transactions on Pattern Analysis and Machine Intelligence · 2025-11-26

  articleSenior author

  One of the main challenges of federated learning (FL) is handling non-independent and identically distributed (non-IID) client data, which may occur in practice due to unbalanced datasets and use of different data sources across clients. Knowledge sharing and model personalization are key strategies for addressing this issue. Clustered federated learning is a class of FL methods that groups clients that observe similarly distributed data into clusters, such that every client is typically associated with one data distribution and participates in training a model for that distribution along their cluster peers. In this paper, we present a unified Bayesian framework for clustered FL which associates clients to clusters. Then we propose several practical algorithms to handle the, otherwise growing, data associations in a way that trades off performance and computational complexity. This work provides insights on client-cluster associations and enables client knowledge sharing in new ways. The proposed framework circumvents the need for unique client-cluster associations, which is seen to increase the performance of the resulting models in a variety of experiments.

  PublisherDOI

• Continuously Optimizing Radar Placement With Model-Predictive Path Integrals

  IEEE Transactions on Aerospace and Electronic Systems · 2025-01-13

  article

  Continuously optimizing sensor placement is essential for precise target localization in various military and civilian applications. While information theory has shown promise in optimizing sensor placement, many studies oversimplify sensor measurement models or neglect dynamic constraints of mobile sensors. To address these challenges, we employ a range measurement model that incorporates radar parameters and radar–target distance, coupled with model-predictive path integral control to manage complex environmental obstacles and dynamic constraints. We compare the proposed approach against stationary radars or simplified range measurement models based on the root-mean-squared error (RMSE) of the cubature Kalman filter estimator for the targets' state. In addition, we visualize the evolving geometry of radars and targets over time, highlighting areas of highest measurement information gain, demonstrating the strengths of the approach. The proposed strategy outperforms stationary radars and simplified range measurement models in target localization, achieving a 38%–74% reduction in the mean RMSE and a 33%–79% reduction in the upper tail of the 90% highest density interval over 500 Monte Carlo trials across all time steps.

  PublisherDOI

• Advances in Anti-Deception Jamming Strategies for Radar Systems: A Survey

  IEEE Aerospace and Electronic Systems Magazine · 2025-07-11 · 10 citations

  articleOpen accessSenior author

  Deception jamming has long been a significant threat to radar systems, interfering with search, acquisition, and tracking by introducing false information that diverts attention from the targets of interest. As deception strategies become more sophisticated, the vulnerability of radar systems to these attacks continues to escalate. This paper offers a comprehensive review of the evolution of anti-deception jamming techniques, starting with legacy solutions and progressing to the latest advancements. Current research is categorized into three key areas: prevention strategies, which hinder the ability of jammers to alter radar processing; detection strategies, which alert the system to deception and may classify the type of attack; and mitigation strategies, which aim to reduce or suppress the impact of jamming. Additionally, key avenues for further research are highlighted, with a particular emphasis on distributed, cognitive, and AI-enabled radar systems. We envision this paper as a gateway to the existing literature on anti-deception jamming, a critical area for safeguarding radar systems against evolving threats

  PublisherOA PDFDOI

• Cross Ambiguity Function Resolution Enhancement Using Neural Networks for Direct Position Estimation Computational Cost Reduction

  2025-04-28 · 1 citations

  articleSenior author

  Conventional methods for the Global Navigation Satellite System (GNSS) positioning involves a two-step (2SP) process, where intermediate measurements such as Doppler shift and time delay of received GNSS signals are computed and then used to for the receiver's position estimation. Alternatively, Direct Position Estimation (DPE) was proposed to estimate the position directly from the sampled signal without intermediate variables, providing high sensitivity and reliable operation in challenging scenarios. Similar to the conventional 2SP method, search methods are designed in the DPE approach to estimate position by searching indices corresponding to the peak in the joint Cross Ambiguity Function (CAF). Currently, the popular search methods for DPE includes the grid-based method and the accelerated random search (ARS) method, but both methods need a large set of candidate points to achieve an accurate position estimation and therefore they involve high computational costs. This paper proposes a neural network based search method, aiming at reducing the high computational cost in the DPE approach. To validate the proposed algorithm, a static receiver positioning scenario is simulated, and the positioning accuracy as well as the running time of each search method is compared.

  PublisherDOI

• Direct Position Estimation Framework for Lunar Positioning, Navigation, and Timing

  2025-04-28

  articleSenior author

  With the increasing volume of lunar missions planned by international space agencies in the near future, precise and reliable Positioning, Navigation, and Timing (PNT) for lunar operations have become essential for successful launches and exploration. Recent research has focused on leveraging the Global Navigation Satellite System (GNSS) for lunar positioning. However, one of the primary challenges of employing GNSS for lunar PNT is the significantly low signal power, caused by the long propagation distance and the side-lobe beam pattern of GNSS signals. In this work, we introduce the Direct Position Estimation (DPE) framework for GNSS-based lunar positioning. DPE is a high-sensitivity receiver design which is capable of operating the weak signal for positioning. It has been shown that the DPE approach outperforms the traditional two-step (2SP) approach, especially when working under harsh environments. Besides, we present GNSSLunarSim, a simple, flexible, and controllable signal simulator designed to generate GNSS baseband signals on the Moon for PNT solution validation. Simulation results demonstrate that the DPE framework achieves higher positioning accuracy than the 2SP approach under various conditions, highlighting its strong potential for lunar PNT applications.

  PublisherDOI

• Collaborative Code-Based Differential GNSS for Multi-User Positioning

  2025-04-28 · 1 citations

  articleSenior author

  This paper introduces a collaborative differential GNSS (C-DGNSS) method designed to enhance positioning accuracy in scenarios with limited satellite visibility, relying solely on GNSS pseudorange measurements and requiring no inter-ranging information. C-DGNSS exploits the correlation among single difference observations from multiple users to a common base station, which are not leveraged by conventional DGNSS. The estimation problem is formulated as an iterative weighted least squares (WLS) estimator, known to be optimal under the Gaussian assumption, and its variance lower bound is derived from the Cramér-Rao bound (CRB). C-DGNSS is compared against conventional DGNSS and the ideal DGNSS bound, the latter assuming noiseless pseudorange measurements for the base station. Simulation results demonstrate that the proposed scheme outperforms standard DGNSS even with a single aiding user, approaching the ideal bound as the number of aiding users grows and their geometry improves. Accuracy gains of up to 28.5% are achieved.

  PublisherDOI

Recent grants

• SaTC: CORE: Small: Securing GNSS-based infrastructures

  NSF · $160k · 2018–2020

• CAREER: Secure and ubiquitous position, navigation and timing

  NSF · $500k · 2019–2026

• NSF-AoF: CIF: Small: Distributed AI for enhanced security in satellite-aided wireless navigation (RESILIENT)

  NSF · $600k · 2023–2026

Frequent coauthors

• Carles Fernández‐Prades

  Centre Tecnologic de Telecomunicacions de Catalunya

  106 shared

• Tales Imbiriba

  70 shared

• Haoqing Li

  62 shared

• Javier Arribas

  Centre Tecnologic de Telecomunicacions de Catalunya

  53 shared

• Jordi Vilà‐Valls

  Université de Toulouse

  52 shared

• Ricardo Augusto Borsoi

  Université de Lorraine

  49 shared

• Deniz Erdoğmuş

  Northeastern University

  39 shared

• Juan A. Fernandez–Rubio

  35 shared

Labs

• Information Processing LabPI

Education

• MSc Advanced Mathematics and Mathematical Engineering, Maths and Statistics

  Universitat Politècnica de Catalunya

  2014

• PhD Electrical Engineering, Signal Theory and Communications

  Universitat Politècnica de Catalunya

  2009

• MSc Electrical Engineering, Signal Theory and Communications

  Universitat Politècnica de Catalunya

  2003

Awards & honors

• EURASIP Best PhD Thesis Award 2014
• 9th Duran Farell Award for Technology Research
• 2016 ION Early Achievement Award
• 2023 Harry Rowe Mimno Award
• 2016 Best Paper Presentation Award, Institute of Navigation