Publications

2026

1. Increasing the reliability of citizen science campaign data for deforestation detection in tropical forests

   Hugo Resende, Álvaro L. Fazenda, Fábio A.M. Cappabianco, and 1 more author

   Future Generation Computer Systems, 2026

   Abs doi: 10.1016/j.future.2025.108081

   In recent years, citizen science (CS) campaigns leveraging crowdsourcing have proven effective in generating large datasets across various fields such as environmental monitoring, and astronomy. However, the quality of volunteer-contributed data remains a challenge, as inconsistent responses often arise from inattentiveness and rapid analyses. To increase reliability in the generation of labeled datasets in citizen science campaigns, this paper proposes the combination of outlier detection techniques (Z-Score, Tukey and Median Absolute Deviation) to remove unreliable voluntary contributions, followed by exclusion of tasks with high Shannon entropy, that is, without consensus of volunteers. To validate this methodology, a case study was conducted using three CS campaigns from the ForestEyes project, which employs citizen science and machine learning to detect deforested areas. The results showed that applying those statistical techniques to filter contributions based on response time of the volunteers joining with median entropy filter led to a growth of up to 20% of accuracy in campaigns, highlighting the importance of integrating statistical techniques and variability to improve the CS results.

2. Citizen Science in Action: Building Trust in Deforestation Data Through Continuous Reliability Enhancements

   Hugo Resende, Álvaro L. Fazenda, Amanda A. S. de Oliveira, and 2 more authors

   IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2026

   doi: 10.1109/JSTARS.2026.3655185
3. Cycling towards equity: Financial incentives as a strategy for mobility justice

   Ana Yoon Faria de Lima, Frauke Behrendt, and Fabio Kon

   Transportation Research Interdisciplinary Perspectives, 2026

   doi: 10.1016/j.trip.2025.101794
4. Parallelization Strategies for the Feature Space Partition Algorithm Applied to Fault Detection and Stability Analysis in Smart Grids

   Saulo Andrade Almeida, Silvana Rossetto, and Carolina Gil Marcelino

   In Intelligent Systems, 2026

   Abs doi: 10.1007/978-3-032-15993-9_30

   This work presents the parallelization of the supervised learning algorithm Feature Space Partition (FSP), originally implemented sequentially in Python. To address computational challenges, multiple parallelization strategies were explored, including implementations based on CPU and GPU, while preserving the original algorithmic flow to ensure that classification performance remained unaffected. Among the approaches tested, CPU-based multiprocessing yielded the most effective results, achieving a reduction of 36.44% in execution time without compromising the predictive accuracy of the model. FSP has demonstrated high classification accuracy, ranging from 80% to 99% in identifying fault patterns in distribution networks and stability patterns in smart grid systems. This reinforces the viability of using existing Python-based parallel processing libraries to scale machine learning algorithms for critical electrical power grid applications.

5. Hybrid Architecture for Spectrum Access Systems With Permissioned Blockchain

   Alan Tony Souza Veloso, Jeffson Celeiro Sousa, Diego Medeiros De Abreu, and 2 more authors

   IEEE Access, 2026

   doi: 10.1109/ACCESS.2026.3660174
6. A Revolução das Tecnologias Quânticas: Caminhos para o Desenvolvimento Nacional

   Antônio Abelém

   Computação Brasil, Jan 2026

   Abs doi: 10.5753/compbr.2026.55.7501

   As tecnologias quânticas representam uma fronteira científica estratégica com potencial para transformar sistemas de comunicação, computação, segurança e sensoriamento. Avançar nesse campo no Brasil exige superar desafios estruturais relacionados ao desenvolvimento de hardware, à criação de algoritmos e ferramentas de software quântico, à construção de redes quânticas interoperáveis e ao fortalecimento da segurança resistente a ataques quânticos. Além disso, a formação de profissionais especializados permanece como um dos fatores mais críticos para consolidar o ecossistema nacional. Este artigo apresenta uma visão integrada desses desafios e discute caminhos para fortalecer a soberania tecnológica do Brasil e ampliar sua participação na economia quântica emergente.

2025

1. Assessing the potential of SAR images in the ForestEyes project’s citizen science campaings

   Vinícius D’Lucas Bezerra Queiroz, Hugo Resende, Fabio Augusto Faria, and 2 more authors

   In Anais do Simpósio Brasileiro de Sensoriamento Remoto (SBSR), Jan 2025

   accepted

   Abs

   Tropical forests play a crucial role in sustaining life on Earth, however, they have experienced a significant reduction in vegetation cover due to activities such as degradation and deforestation. A recent initiative to monitor these areas is the ForestEyes project, which combines citizen science and machine learning to track changes in tropical forests. The project currently uses images from optical sensors to create citizen science campaigns. However, the presence of clouds is a limiting factor in optical images and can prevent the detection of deforested areas. Thus, this article presents a preliminary evaluation of Synthetic Aperture Radar (SAR) images from Sentinel-1 and PALSAR-2/ALOS-2 sensors, exploring their viability for citizen science campaigns in the context of the ForestEyes project. Experimental results indicate that PALSAR-2 data is more effective, as it has been particularly advantageous in identifying deforested areas.

2. Unveiling the Impact of Sampling on Feature Selection for Performance Prediction in Configurable Systems

   João Marcello Bessa, Millena Cavalcanti, Mathieu Acher, and 2 more authors

   In 2025 International Conference on Software Reuse (ICSR), Apr 2025
3. Integrating Data Privacy and Energy Management in Smart Cities With Partial Sustainable IoT Networks

   Yuri Santo, Antônio Abelém, André Riker, and 1 more author

   IEEE Access, Apr 2025

   Abs doi: 10.1109/ACCESS.2025.3541213

   IoT networks are transitioning towards greener underlying technologies. In smart cities environments, while some devices are already sustainable with rechargeable batteries and energy harvesting hardware, many still rely on traditional non-sustainable components. This partial sustainable Internet of Things (IoT) network poses a research challenge, since the profile of sustainable nodes aims for neutral or self-powered operation, which may conflict with the energy-saving goals of traditional non-sustainable nodes. In this challenging scenario, it is essential to share Energy Information (EI), i.e. battery state-of-charge, among the nodes to efficiently manage energy resources. However, EI sharing is valuable data for attackers to perform energy-based attacks, exploiting the most vulnerable nodes in smart cities. To tackle these problems, this article presents the Dual Energy PROfile for interNet-of-thiNgs with Enhanced Security (DEPRONN-ES), which is a solution designed to support partially sustainable networks, i.e. traditional and sustainable nodes, and applies differential privacy to enhance data privacy levels of the EI sharing. The performed evaluation shows the proposed solution is able to reduce the success rate of attacks in 87% and its effectiveness in terms of consumed resources of the attacked node. Besides, DEPRONN-ES is able to support conflicting profiles to operate in distinct energy consumption levels. This framework utilized emulation of IoT devices and real data traces of indoor light intensity measurements provided by Columbia University as a harvesting source of the rechargeable batteries of sustainable nodes.

4. Exploring Superpixel Segmentation Methods in the Context of Citizen Science and Deforestation Detection

   Hugo Resende, Isabela Borlido, Victor Sundermann, and 4 more authors

   In IGARSS 2025 – 2025 IEEE International Geoscience and Remote Sensing Symposium, Aug 2025

   Abs doi: 10.1109/IGARSS55030.2025.11243539

   Tropical forests play a central role in the planet’s ecosystem, making the conservation of these biomes a global priority. However, ongoing deforestation and degradation represent a significant threat to its existence, which requires effective monitoring and the proposition of actions to mitigate the damage caused by these processes. In this sense, several government and private monitoring programs have been developed, which apply image processing, machine learning and citizen science (CS). Mainly in the context of CS, an important project is ForestEyes, which segments remote sensing images to identify deforested areas and subsequently submit these segments for analysis by non-specialized volunteers. As segmentation using techniques based on superpixels proves to be a viable solution for this important task. Therefore, this paper presents an analysis of 22 segmentation methods based on superpixels applied to remote sensing images, aiming to identify which of them are most suitable for generating better segments for CS campaigns. The results revealed that seven of the segmentation methods outperformed the baseline method (SLIC) currently employed in the ForestEyes project, indicating an opportunity for improvement in this important phase of campaign development.

5. Performance Assessment of Optical and SAR Imagery for Superpixel-Based Deforestation Mapping in the ForestEyes Project

   Vinícius D’Lucas B. e Queiroz, Hugo Resende, Fabio A. Faria, and 1 more author

   In 2025 38th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), Sep 2025

   Abs doi: 10.1109/SIBGRAPI67909.2025.11223381

   This study investigates the potential of SAR data (Cand L-band), individually and combined with optical imagery (Landsat-8 OLI), for classifying deforested areas within the ForestEyes project. Six classification scenarios with varying data combinations were defined. Images were segmented using the Simple Linear Iterative Clustering (SLIC) algorithm, and Haralick texture descriptors were extracted from each superpixel. Classifications were performed using Support Vector Machine (SVM) and Random Forest (RF), with ground truth labels derived from PRODES (Monitoring Program of the Amazon Forest by Satellite). Performance evaluation was based on 100 classification iterations, from which 95% confidence intervals were calculated for confusion matrix-derived metrics, including recall and precision. Results showed that scenarios using optical data yielded the best classification outcomes, with high recall and precision values for the Recent Deforestation class. The inclusion of SAR data did not lead to statistically significant improvements and, increased Recent Deforestation false positives, especially in scenarios relying exclusively on SAR data. Confidence interval analysis confirmed the superior performance of optical imagery for detecting deforested areas in this context.

6. Assessing the Performance and Impact of an Open-Source RTI in Distributed Real-Time Military Simulation

   João Rosa, and Álvaro Fazenda

   In Anais do XXVI Simpósio em Sistemas Computacionais de Alto Desempenho, Bonito/MS, Sep 2025

   doi: 10.5753/sscad.2025.15637
7. Rotulagem Massiva de Imagens por Ciência Cidadã para Detecção de Desmatamento

   Hugo Resende, Álvaro Fazenda, and Fabio Faria

   In Anais do XVI Workshop de Computação Aplicada à Gestão do Meio Ambiente e Recursos Naturais, Maceió/AL, Sep 2025

   doi: 10.5753/wcama.2025.7514
8. MDMWARE: Model-Driven Domain-Specific Middleware as a Tool for IT/OT Integration in the Context of Smart Cities

   Paulo César F. Melo, and Fábio M. Costa

   In Proceedings of the International Workshop on Middleware for IT/OT Integration, Nashville, TN, USA, Sep 2025

   Abs doi: 10.1145/3774900.3776638

   The concept of smart cities entails the use of information technologies to access the resources of a city in a more intelligent way by providing specialized services to citizens and city officials. In this way, the city’s IT infrastructure should facilitate the development, deployment, and operation of smart city applications for diverse application domains. Building such applications is challenging since they are used on large-scale systems, have dynamic behavior, must access heterogeneous resources, and must deal with the semantics of different domains. We propose and evaluate an integrated approach for the development and execution of smart city applications in this context. To this end, we leverage a model-driven approach for domain-specific middleware, adding support for the development of applications in a platform-agnostic way using domain-specific building blocks. The proposed approach significantly reduces development complexity and enables the effective adaptation of domain-specific applications for different smart city platforms with adequate performance.

9. Efficient Service Placement and Node Allocation in Fog and Edge Computing

   Helberth Borelli, Humberto José Longo, Fábio M. Costa, and 1 more author

   In Proceedings of the 3rd International Workshop on Middleware for the Computing Continuum, Nashville, TN, USA, Sep 2025

   Abs doi: 10.1145/3774898.3778039

   Service placement and server allocation in fog and edge computing environments are challenging, especially for latency-sensitive applications, due to the limited resources and diverse service demands. This study proposes and evaluates three specialized heuristics that improve resource utilization while minimizing network latency and placement overhead. The heuristics are designed to: (i) minimize the number of allocated nodes, (ii) maximize resource utilization of active nodes, and (iii) reduce communication path lengths. To assess their effectiveness, the heuristics are compared against integer linear programming (ILP) models used as optimal baselines. The experimental results show that the heuristics achieve, on average, nearly 80% efficiency relative to their counterpart ILP solutions while drastically reducing execution time by several orders of magnitude, making them suitable for real-time and resource-constrained environments. These findings underscore the flexibility of the heuristics, offering a practical trade-off between solution quality and processing time in fog and edge computing scenarios.

10. Urban Digital Twins for Megalopolises: Requirements, Challenges and Opportunities

    Daniel Cordeiro, Jiefu Song, Kelly Braghetto, and 2 more authors

    In Anais do I Workshop em Engenharia de Software para Gêmeos Digitais, Recife/PE, Sep 2025

    doi: 10.5753/sedt.2025.14274
11. Automatically Augmenting GitHub Issues with Informative User Reviews

    Arthur Pilone, Marco Raglianti, Michele Lanza, and 2 more authors

    In 2025 IEEE International Conference on Software Maintenance and Evolution (ICSME), Sep 2025

    doi: 10.1109/ICSME64153.2025.00045
12. Compatibilização de setores censitários baseada em grafos para análise histórica de processos urbanos em escala local

    Pedro R. Mendonça, and Fabio Kon

    In Anais do IX Workshop de Computação Urbana, Natal/RN, Sep 2025

    doi: 10.5753/courb.2025.7950
13. Avaliação da Privacidade Diferencial Aplicada ao Aprendizado Federado Através de Ataques de Inversão de Modelo

    Breno V. Manhães, João Pedro M. F. Ferreira, Guilherme A. Thomaz, and 1 more author

    In Anais do XXV Simpósio Brasileiro de Cibersegurança (SBSeg 2025), Foz do Iguaçu/PR, Sep 2025

    Abs doi: 10.5753/sbseg.2025.11428

    Este trabalho propõe uma metodologia prática para avaliar a eficácia de mecanismos de Privacidade Diferencial (DP) no Aprendizado Federado (FL) diante de ataques de inversão de modelo. Para isso, foi adotado um cenário adversarial baseado no paradigma de segurança cibernética equipe vermelha/equipe azul (RT/BT), onde a equipe azul implementa a proteção de DP e a equipe vermelha realiza ataques para recuperar dados de um cliente específico a partir do modelo global. Foram implementados dois tipos de ataques — um baseado em gradientes e outro ingênuo — aplicados em diferentes intensidades de ruído gaussiano. Os experimentos demonstraram que ruídos baixos já são suficientes para mitigar significativamente os ataques de inversão, conforme evidenciado tanto por métricas quantitativas (SSIM, PSNR e MSE) quanto pela análise visual das imagens reconstruídas. Por outro lado, observou-se que a acurácia do modelo global se mantém estável até níveis moderados de ruído, com impacto limitado no desempenho. Esses resultados indicam que é possível alcançar um bom equilíbrio entre privacidade e utilidade do modelo em cenários práticos.

14. Blockchain na Era Quântica: Uma Rapid Review sobre Avanços, Lacunas e Direções Futuras

    Mateus Bastos, Alan Veloso, Diego Abreu, and 3 more authors

    In Anais Estendidos do XIV Latin-American Symposium on Dependable and Secure Computing, Valparaíso/Chile, Sep 2025

    doi: 10.5753/ladc_estendido.2025.16894
15. A two-stage Q-learning routing approach for quantum entanglement networks

    Diego Abreu, Arthur Pimentel, David Moura, and 2 more authors

    Annals of Telecommunications, Sep 2025

    doi: 10.1007/s12243-025-01090-4
16. Towards Blind Quantum Machine Learning in Entanglement Networks

    Diego Medeiros Abreu, and Antônio Abelém

    In Proceedings of the 2nd Workshop on Quantum Networks and Distributed Quantum Computing, Coimbra, Portugal, Sep 2025

    Abs doi: 10.1145/3749096.3750027

    Blind Quantum Computation (BQC) enables clients to delegate quantum computations to a quantum server while maintaining the privacy of their data and algorithms, even when the server is untrusted. In this work, we extend BQC frameworks to Quantum Machine Learning (QML) by implementing a network of entangled clients and a quantum server. Specifically, we explore the integration of Variational Quantum Classifiers (VQC) and Quantum Convolutional Neural Networks (QCNNs) within this paradigm. Our proposed model allows clients to perform classical preprocessing and optimization locally while leveraging the quantum server for computationally expensive quantum tasks. The entanglement-based network is managed by a controller that dynamically allocates resources according to the BQC protocol, allowing secure and efficient execution. We present simulation results indicating the feasibility of this approach, including an analysis of network efficiency and resource consumption, alongside the F1 score on QML benchmark datasets.

17. Sincronização Descentralizada entre Sistema de Acesso ao Espectro via Blockchain

    Alan Veloso, Jeffson Sousa, Diego Abreu, and 1 more author

    In Anais do LII Seminário Integrado de Software e Hardware, Maceió/AL, Sep 2025

    doi: 10.5753/semish.2025.9038
18. Uma Arquitetura Distribuída para Sistemas de Acesso ao Espectro com Blockchain

    Alan Veloso, Jeffson Sousa, Diego Abreu, and 1 more author

    In Anais do III Colóquio em Blockchain e Web Descentralizada, Maceió/AL, Sep 2025

    doi: 10.5753/cblockchain.2025.9032
19. Abordagens de Monitoramento Estruturado em Camadas para Redes Blockchain

    Mateus Bastos, Alan Veloso, Jeffson Sousa, and 2 more authors

    In Anais do III Colóquio em Blockchain e Web Descentralizada, Maceió/AL, Sep 2025

    doi: 10.5753/cblockchain.2025.8833
20. Aspectos Técnicos Avançados de Blockchain no Gerenciamento de Espectro

    Alan Veloso, Mateus Bastos, Jeffson Sousa, and 2 more authors

    In Anais do III Colóquio em Blockchain e Web Descentralizada, Maceió/AL, Sep 2025

    doi: 10.5753/cblockchain.2025.8824
21. Two-Stage Deep Q Learning Routing in Entanglement Networks

    Diego Abreu, Arthur Pimentel, Christian Esteve Rothenberg, and 1 more author

    In 2025 IEEE Symposium on Computers and Communications (ISCC), Sep 2025

    doi: 10.1109/ISCC65549.2025.11326361
22. A Privacy-Preserving and Secure Monitoring Framework for Permissioned Blockchain Networks

    Jeffson C. Sousa, Alan Veloso, Mateus Bastos, and 2 more authors

    In 2025 7th International Conference on Blockchain Computing and Applications (BCCA), Sep 2025

    doi: 10.1109/BCCA66705.2025.11229760
23. Performance Evaluation of Decentralized Digital Identity Contracts on Ethereum-Based Blockchain Networks

    Jeffson C. Sousa, Bruno Evaristo, Antonio Mateus, and 3 more authors

    In 2025 7th International Conference on Blockchain Computing and Applications (BCCA), Sep 2025

    doi: 10.1109/BCCA66705.2025.11229679
24. A Decentralized Marketplace with Self-Sovereign Identities for Telecommunication Services

    Jeffson C. Sousa, Bruno Evaristo, Vinicius Duarte, and 3 more authors

    In 2025 7th International Conference on Blockchain Computing and Applications (BCCA), Sep 2025

    doi: 10.1109/BCCA66705.2025.11229744
25. Governança Digital que se Explica: Interação Segura com Regras Programáveis

    Bruno Evaristo, Jeffson Celeiro, and Antônio Abelém

    In Anais Estendidos do XXIV Simpósio Brasileiro sobre Fatores Humanos em Sistemas Computacionais, Belo Horizonte/MG, Sep 2025

    doi: 10.5753/ihc_estendido.2025.13214
26. Ataque Adversarial de Evasão a Sistema de Detecção de Intrusão e Métodos de Defesa em Redes Open RAN

    Victor Dias, Murilo Silva, Matheus Gomes, and 3 more authors

    In Anais do XXV Simpósio Brasileiro de Cibersegurança, Foz do Iguaçu/PR, Sep 2025

    doi: 10.5753/sbseg.2025.11492
27. D-MTS: Um Sistema Baseado em Blockchain para Compartilhamento Eficiente e Transparente de Infraestrutura de Redes de Telecomunicações

    Jeffson Sousa, Alan Veloso, Bruno Evaristo, and 3 more authors

    In Anais do LII Seminário Integrado de Software e Hardware, Maceió/AL, Sep 2025

    doi: 10.5753/semish.2025.9357
28. Dynamic Network Slicing for User Mobility Support in 5G Networks

    Diogo M. Goncalves, Luiz F. Bittencourt, and Edmundo R. M. Madeira

    In 2025 IEEE Network Operations and Management Symposium (NOMS), Sep 2025

    Abs doi: 10.1109/NOMS57970.2025.11073669

    Network slicing has been a key supporting technology for resource management in modern computer networks, from 5G to 6G and beyond. Network slicing creates virtual networks over one common physical infrastructure. The dynamic resource allocation for services and slices to serve demand variations caused by user mobility is a challenge in 5G networks. Based on the described scenario, this thesis developed new resource management mechanisms for the dynamic network slicing allocation problem to support mobile users in 5G/edge computing infrastructures. The contributions of this work include: 1) one evaluation of the performance of dynamic network slicing in user mobility support scenarios; 2) a mathematical model which describes the network slices reconfiguration delays; 3) two heuristic solutions for improving the network slice reconfiguration process; and 4) a network simulator for evaluating the proposed solutions. The simulator is available as free software and has been adopted as the validation environment by the scientific community. The results of this work indicate the positive impact of slices’ dynamic resource allocation in dealing with demand variations triggered by mobile users. However, the frequency of those reconfigurations is crucial to the success of such a solution.

29. Federated Urban Flow: A Model-Centric Benchmark for Reproducible Evaluation of Deep Learning in Federated Traffic Forecasting

    Matteus V. S. Silva, Carlos A. Astudillo, and Luiz F. Bittencourt

    In 2025 3rd International Conference on Federated Learning Technologies and Applications (FLTA), Oct 2025

    Abs doi: 10.1109/FLTA67013.2025.11336506

    Urban traffic forecasting is a fundamental component of Smart Cities, yet centralized Machine Learning (ML) approaches present notable challenges regarding data privacy and communication overhead. Federated Learning (FL) offers a privacy-preserving alternative to run distributed learning at the edge, but a methodological gap exists: the lack of a standardized software benchmark to systematically evaluate and compare the performance of different deep learning architectures (e.g., RNNs vs. Transformers), impeding reproducible and trustworthy research. This paper introduces Federated Urban Flow, a modular, open-source software benchmark designed specifically to probe this “model-to-application” axis. We demonstrate the benchmark’s utility through a comprehensive case study involving 21 experimental scenarios. We evaluate on the public METRLA dataset using two aggregation rules (FedAvg, FedProx). The case study reveals that (1) contextual features are key for accuracy improvement, reducing prediction error by over 70%; (2) the Transformer architecture excels in challenging longhorizon forecasts; and (3) there are clear diminishing returns in scaling clients and communication rounds. The benchmark provides a useful tool for the community to advance research in decentralized, edge-based AI for Smart Cities.

30. Multiple-agent opportunistic data gathering in IoT networks using group selection

    Edvar Afonso, and Miguel Elias M. Campista

    In 2025 IEEE Latin-American Conference on Communications (LATINCOM), Nov 2025

    Abs doi: 10.1109/LATINCOM67778.2025.11345438 personal copy

    The Internet of Things (IoT) is based on data collection for future processing and decision-making. In multihop Low-Power and Lossy Network (LLN) scenarios, it is important to collect data from sensors and IoT devices with efficient energy consumption. This paper explores the concept of multi-mobile agents (MMA) combined with proactive caching to collect data in an IoT network. In the proposed architecture, the group of itineraries is computed by a mechanism that uses the Hamming distance technique to partition a sensor network into a number of data-centric clusters. In addition, the data collection mechanism is modeled as a knapsack problem that improves the energy efficiency with MA payload variation. The simulations show lower energy consumption from network devices as the payload and the number of MMA increase.

31. ns3-oran-customizable-db: Uma Ferramenta para Construção de Base de Dados Personalizável para Simulações Open RAN

    Felipe Táparo, Igor Moraes, and Miguel Campista

    In Anais Estendidos do XLIII Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos, May 2025

    Abs doi: 10.5753/sbrc_estendido.2025.7111

    Uma parte essencial no desenvolvimento de aplicações para o controle e diagnóstico de uma Rede de Acesso via Rádio (Radio Access Network – RAN) é a simulação. É imprescindível a avaliação das soluções propostas em um cenário simulado próximo à realidade a fim de evitar possíveis falhas na implementação. O objetivo deste trabalho é adicionar funcionalidades à extensão "ns3-oran" do simulador de redes "ns-3". As funcionalidades implementadas dão ao usuário a liberdade para utilizar a extensão sem restrições de quais dados podem ser enviados ao Near-RT RIC. A capacidade de coleta, armazenamento e análise de diversos tipos de dados é fundamental no desenvolvimento e implementação de soluções baseadas em aprendizado de máquina. Assim, as ferramentas desenvolvidas permitem a criação de cenários de simulações capazes de treinar modelos por aprendizado por reforço, utilizando métricas de desempenho que antes eram impossíveis de serem manipuladas pelo simulador.

32. NS3-Oran-Customizable-Db: A Tool for Building a Customizable Database for Open RAN Simulations

    Felipe G. Táparo, Igor M. Moraes, and Miguel Elias M. Campista

    In 13th Wireless Days Conference (WD 2025), Dec 2025

    Abs doi: 10.1109/WD67713.2025.11302525 personal copy

    An essential approach for evaluating the perfor mance of applications designed for control and diagnosis of Radio Access Networks (RANs) is simulation. Assessing proposed solutions in simulated environments that closely approximate real-world network behavior is crucial to mitigate potential risks during deployment. This work aims to extend the capabilities of the "ns3-oran" module in the "ns-3" network simulator. The new functionalities remove restrictions on the type of data that can be exchanged with the Near-RT RIC (Near-Real-Time RAN Intelligent Controller), thereby increasing flexibility for researchers and developers. The ability to collect, store, and analyze heterogeneous data is a key enabler for the design and integration of machine learning-based solutions into the RAN. With these enhancements, the simulator can support the creation of scenarios tailored to reinforcement learning (RL) training, leveraging performance metrics that were previously inaccessible or non-configurable within ns-3. As a result, the proposed "ns3-oran-customizable-db" provides a more powerful simulation environment for advancing AI-driven RAN control and optimization approaches.

33. Seleção do Modelo Federado Baseada em Previsão de Desempenho e Tempo de Comunicação

    Luana Gantert, and Miguel Elias M Campista

    In XLIII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2025), Sep 2025

    Abs doi: 10.14209/sbrt.2025.1571157155

    O aprendizado federado assume que clientes e servidores se conhecem previamente e que o processo de seleção é apenas relativo aos clientes. Este trabalho vai em uma direção diferente e propõe a seleção de servidores a partir de estimativas futuras de desempenho dos modelos em treinamento. Cada cliente inicia dois procedimentos de treinamento em paralelo com dois servidores com dois modelos distintos e calcula, através de regressão linear, qual deles deve ser mantido. A proposta compara o desempenho dos servidores a partir da métrica AUC-ROC em uma dada rodada r e, caso a diferença seja maior ou igual a β%, o cliente mantém o treinamento apenas com o de melhor desempenho. A proposta também considera a latência da comunicação cliente-servidor como critério de decisão, simulando a latência com valores provenientes de aplicações de monitoramento industrial. Os resultados demonstram redução da função de perda durante o treinamento dos modelos em relação à proposta utilizada como base. Além disso, ao variar o parâmetro β, o melhor cenário analisado apresenta redução de aproximadamente 10,4% da latência acumulada no treinamento com maior valor de r e um aumento de 12,5% na métrica AUC-ROC, considerando o cliente com aumento de desempenho mais expressivo.

34. Battery life optimization in LoRa networks using spreading factor reallocation

    Ian H. de Andrade, Luís Henrique M.K. Costa, and Rodrigo S. Couto

    Ad Hoc Networks, Oct 2025

    Abs doi: 10.1016/j.adhoc.2025.103911 personal copy

    This paper proposes a dynamic strategy for the allocation of resources used by end devices in LoRa networks, which employ chirp spread spectrum modulation. The proposed battery life optimization (BLO) strategy splits end devices into different spreading factor (SF) groups. The basic idea is to reduce the collisions between end devices using the same SF. Moreover, BLO also considers the current battery level of each end device, and periodically reallocates the SF groups to optimize the battery consumption of all nodes and extend the network lifetime. The main innovation of BLO is to consider in addition to the RSSI the air time of different SFs as a weighting factor in SF allocation. We compare BLO to state-of-the-art (SoA) SF-allocation strategies, achieving 77% improvement in successful message delivery compared to LoRaWAN’s ADR scheme. Furthermore, we obtain better energy efficiency with BLO. In a scenario with one gateway and 500 devices operating over 24 hours, the remaining energy with BLO is 10 and 3.6 times larger than with EXPLoRa-SF and EXPLoRa-AT SoA strategies, respectively.

35. Exploring Traffic Pattern Variability in Vehicular Federated Learning

    Giuliano Prestes Fittipaldi, Rodrigo Souza Couto, and Luís Henrique Maciel Kosmalski Costa

    Computer Communications, Oct 2025

    Abs doi: 10.1016/j.comcom.2025.108279 personal copy

    The emergence of software-defined vehicles has brought machine learning into the vehicular domain. To support these data-driven applications, techniques to incentivize users to share their vehicle data are crucial. Federated learning trains machine learning models in a distributed manner, leveraging client data without compromising its privacy. Nonetheless, in vehicular networks, the dynamic behavior of nodes affects client availability and the global model’s performance. Accordingly, this paper evaluates federated learning (FL) in a realistic vehicular network topology, accounting for real vehicle traffic in two Brazilian urban areas. The network simulation covers 3.7 km^2 with 1,290 vehicles per hour and road speeds, based on real data. Our paper provides a comprehensive analysis of the impact that different traffic behaviors can yield during the training phase of a federated learning model. We observe that there is a performance decay in urban areas with longer vehicle permanence. Interestingly, longer vehicle participation in FL training leads to a biased finalmodelwithreducedgeneralization.Wepropose a novel approach to verify vehicle variability over time, by using the Dice-Sørensen coefficient to compare the set of clients participating in different rounds of training. By maintaining the vehicle variability over the rounds we can reduce the effect of the bias on the model, and– with a 47% reduction of the communication overhead– achieve faster learning, higher convergence in the first 15 rounds, and an equivalent final accuracy. Additionally, we extend our analysis by conducting simulations under more extreme traffic scenarios across multiple datasets, using a MobileNetV3. The results confirm that sustaining high vehicle variability– in scenarios with a brief participation of vehicles in the training– yields comparable model performance while saving up to 83.5 GB in communication costs.

36. Melhoria da Segmentação de Imagens em Condições Adversas no Monitoramento de Infraestrutura Costeira

    Pedro Videira Rubinstein, Ricardo Lorente Kauer, Alexandre G. Cardeman, and 4 more authors

    In Workshop de Computação Urbana (CoUrb), May 2025

    Abs doi: 10.5753/courb.2025.9518

    Smart cities increasingly rely on AI-driven solutions. In coastal areas like Rio de Janeiro, maritime risks present significant challenges, as exemplified by the collapse of the Tim Maia Bike Path. This paper proposes a method for monitoring coastal infrastructure using a custom segmentation model based on YOLOsegmentation, reducing the need for constant human supervision. However, external cameraplacementintroduceschallengesmakingsegmentationdifficult. To address these challenges, we investigate the effects of data augmentation on AI model performance. As a case study, we apply this method to develop a system for the Tim Maia Bike Path, with models achieving 98% mAP50-95 throughout the day.

37. Sistema de Detecção de Usuários Vulneráveis da Via Robusto à Baixa Iluminação

    Vinicius Oliveira Avena, Rodrigo Souza Couto, Miguel Elias Mitre Campista, and 1 more author

    In Workshop de Computação Urbana (CoUrb), May 2025

    Abs doi: 10.5753/courb.2025.9064

    This work proposes a computer vision-based ADAS system designed to enhance the safety of Vulnerable Road Users (VRUs) in blind spot regions using only rear-view camera input. The method combines object detection and multi-object tracking to generate proximity alerts in real time. The system also integrates a fallback mechanism for low-light conditions, based on headlight detection. To support research in this domain, we introduce FilterLane-VRU, a new dataset composed of real-world rear-view urban traffic scenarios annotated with alert labels. The proposed pipeline offers a cost-effective and reliable solution for VRU protection in urban contexts.

38. Análise da Qualidade de Serviço do Sistema Público de Ônibus Urbanos através de Monitoramento por GPS

    Tiago R. Borzino, Fernando D. M. Silva, Aline C. Viana, and 1 more author

    In Workshop de Computação Urbana (CoUrb), May 2025

    Abs doi: 10.5753/courb.2025.9544

    A introdução de dispositivos IoT equipados com GPS nos ônibus urbanos possibilita a coleta de dados para monitorar a frota e avaliar a qualidade do serviço prestado. Este trabalho analisa dados de GPS dos ônibus do Rio de Janeiro, incluindo trajetórias de viagens, linha e identificador do veículo. A partir dessas informações, são estimadas métricas de desempenho de cada linha. Em particular, são investigadas as métricas de intervalo de passagem, número de agrupamentos de ônibus, tempo para percorrer uma fração da trajetória e entropia como indicadores de qualidade de serviço. Os resultados demonstram a regularidade ou imprevisibilidade das diferentes linhas analisadas. Por fim, investiga-se a correlação entre as diferentes métricas a fim de mostrar se a piora ou melhora em um critério de desempenho reflete em algum dos outros.

39. DIGA: Definição e Implementação de Modelos com Aprendizado Generativo para Aplicações Inteligentes

    Lucas Airam Castro Souza, Guilherme Araujo Thomaz, Fernando Dias Mello Silva, and 4 more authors

    In Workshop de Gerência e Operação de Redes e Serviços (WGRS), May 2025

    Abs doi: 10.5753/wgrs.2025.7928

    Machine learning is already integrated into many everyday applications. However, selecting the appropriate model is challenging due to the numerous available options. This challenge becomes even more complex in vehicular networks, where applications require low latency and operate on heterogeneous devices. Thus, a mechanism that estimates inference time and selects the most suitable model for the target device becomes essential. This work proposes DIGA, a machine-learning model selection system designed for the automotive sector. Considering both device and candidate model characteristics, the system estimates the inference time required to process a sample and verifies whether the model meets the application’s latency criteria in a multi-device scenario. To estimate the inference time of each model on a device, this work compares a deterministic mathematical model with four different generative artificial intelligence models, using real execution times as a reference. The results show that deterministic mathematical models used in the literature underestimate inference time in 80% of cases. Therefore, this study highlights the need to consider model structure, as the number of floating-point operations derived mathematically is insufficient to accurately determine inference time.

40. Uma Nova Representação do Espaço de Soluções para Arrefecimento Simulado em Problemas de Alocação de Circuitos Virtuais

    Fernando D. M. Silva, and Luís Henrique M. K. Costa

    In Workshop de Gerência e Operação de Redes e Serviços (WGRS), May 2025

    Abs doi: 10.5753/wgrs.2025.8863

    The optimization of virtual circuits in networks is often solved by heuristics, such as Simulated Annealing (SA), due to the complex topology and demand patterns. To reduce SA’s execution time, other techniques have poor representation of the solution space or complex operations. We propose a solution space for SA that finds paths and optimizes the result simultaneously. The proposal represents the entire solution space, with a tradeoff of higher dimensionality. We compare with other heuristics and observe that the proposal finds viable solutions, although it requires more execution time. This representation can be used in scenarios where solutions based solely on κ-shortest paths are insufficient to achieve a near-optimal solution.

41. AGATA - Arquitetura para Gerenciamento Automático de Tarefas de Aprendizado Federado

    Guilherme Araujo Thomaz, Fernando Dias Mello Silva, Lucas Airam Castro Souza, and 2 more authors

    In Salão de Ferramentas do XLIII Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos (SBRC), May 2025

    Abs doi: 10.5753/sbrc_estendido.2025.7755

    O aprendizado federado protege a privacidade ao efetuar o treinamento do modelo nos dispositivos dos usuários. Entretanto, o operador em nuvem não possui acesso aos dispositivos, que estão muitas vezes indisponíveis, para experimentar com diferentes modelos e hiperparâmetros. Este trabalho propõe uma ferramenta que resolve esses problemas ao oferecer funcionalidades para a inicialização, finalização e troca automática de múltiplas tarefas de aprendizado federado. A implementação da AGATA é compatível com ferramentas populares, como o Flower, e se diferencia de outros sistemas por automatizar a troca de hiperparâmetros e códigos.

42. TOFL: Time Optimized Federated Learning

    Lucas Airam Castro Souza, Matteo Sammarco, Nadjib Achir, and 2 more authors

    In XXV Simpósio Brasileiro de Cibersegurança (SBSeg), Sep 2025

    Abs doi: 10.5753/sbseg.2025.10576

    Vehicular networks face cyber threats that can harm drivers, passengers, and pedestrians. In this scenario, federated learning is a possible solution to train models that detect threats without violating user privacy. However, federated learning is particularly sensitive to communication delays, which is a natural consequence of high mobility in vehicular networks. This problem is commonly ignored in the literature, which does not consider the possibility of network disconnections. This work proposes a client selection strategy designed to minimize the training time of a machine learning model for vehicular threat detection, considering the communication time that varies according to the movement of clients. The results demonstrate that TOFL, using only 20% of the total available clients, can reduce the time required to achieve high accuracy by up to 50% compared to state-of-the-art approaches, while reducing the resource consumption of client devices.

43. Bridging Domain Shifts Through Self-Contrastive Learning And Distribution Alignment

    Vinicius Avena, Rodrigo S. Couto, Luís Henrique M. K. Costa, and 1 more author

    In 2025 IEEE International Conference on Image Processing (ICIP), Oct 2025

    Abs doi: 10.1109/ICIP55913.2025.11084279 personal copy

    Deep learning models excel in computer vision tasks with large labeled datasets but often struggle with performance degradation under domain shifts to unlabeled target domains. Unsupervised Domain Adaptation (UDA) mitigates this challenge by transferring knowledge from a labeled source domain to an unlabeled target domain. In this paper, we propose Self-Contrastive Learning for Domain Adaptation (SCODA), a novel framework that combines self-supervised contrastive learning with distributional alignment techniques to learn domain-invariant representations. SCODA jointly optimizes domain alignment and classification by using contrastive loss to reduce feature gaps while preserving task performance through source supervision. Experiments on UDA benchmarks demonstrate that SCODA outperforms traditional baselines, achieving accuracy improvements without relying on target labels. These results highlight SCODA’s effectiveness in addressing complex domain shifts and its potential for real-world applications.

44. Mirror, Mirror on the Road, Is There a VRU too Close?

    Vinicius Avena, Rodrigo Souza Couto, Miguel Elias M. Campista, and 1 more author

    In IEEE 102nd Vehicular Technology Conference (VTC2025-Fall), Oct 2025

    Abs doi: 10.1109/VTC2025-Fall65116.2025.11310181 personal copy

    This work presents a proactive computer visionbased ADAS system designed to enhance VRU safety in risk zones at the rear of the vehicle and filter lanes. The system combines object detection, multi-object tracking, and risk assessment to generate real-time proximity alerts. Compared to a detection-only approach, the proposed method shows a 19% improvement in precision and a 10% increase in F1-score, while introducing a latency overhead of less than 1 ms. Evaluated on edge devices, the system maintains efficient performance across resourceconstrained platforms. To support research in this domain, we introduce FilterLane-VRU, a new dataset of rearview urban traffic scenarios with temporal proximity annotations. Results validate the feasibility of deploying the system in real-world ADAS applications, offering a reliable and cost-effective solution for VRU safety.

45. Lithium-Ion Battery Aging Estimation with Federated Deep Learning

    Luan L. Santos, Guilherme A. Thomaz, Lucas A. C. Souza, and 4 more authors

    In 10th Workshop on Communication Networks and Power Systems (WCNPS), Nov 2025

    Abs doi: 10.1109/WCNPS69127.2025.11295869 personal copy

    A reliable method for predicting the State of Health (SOH) of lithium-ion batteries is crucial for extending their lifespan. However, using traditional machine learning models for this task requires large, centralized datasets, which creates significant privacy issues. This work proposes a federated learning (FL) framework with a temporal autoencoder architecture that estimates SOH directly from partial charging cycles, removing the need for full capacity curves. Our systematic evaluation against a centralized baseline revealed a trade-off between accuracy and privacy. The centralized model achieved superior performance, while the FL model showed a 25.9% higher loss and a 10.9% higher Mean Absolute Error (MAE). However, increasing the number of clients in the FL setup slowed model training convergence. These results validate FL as a practical, privacy-preserving solution for SOH prediction, despite a reduction in performance.

46. Smart health practices: Strategies to improve healthcare effectiveness with digital twin technology

    Md. Armanul Hasan, Ridwan Mustofa, Niamat Ullah Hossain, and 1 more author

    Smart Health, Nov 2025

    doi: 10.1016/j.smhl.2025.100541
47. Planning and optimization of nitrogen fertilization in corn based on multispectral images and leaf nitrogen content using unmanned aerial vehicle (UAV)

    Diogo Castilho Silva, Beata Emoke Madari, Maria da Conceição Santana Carvalho, and 2 more authors

    Precision Agriculture, Nov 2025
48. Integrative approach for monitoring the toxicity of effluents, surface water, and soil in the Cerrado biome

    Igor Romeiro Santos, Marcelino Benvindo Souza, Daniela Pereira da Silva Alves, and 9 more authors

    Environmental Monitoring and Assessment, Nov 2025
49. Citizen Science in Action: Enhancing Accuracy In Deforestation Monitoring with ForestEyes

    Hugo Resende, Alvaro Luiz Fazenda, Amanda A. S. Oliveira, and 2 more authors

    SSRN – Social Science Research Network, Nov 2025

    doi: 10.2139/ssrn.5100126
50. ImprovMLCQ: A Feature-Enriched Dataset for Advancing Code Smell Detection

    Joanne Carneiro, Jessica Ribas, Amanda Santana, and 2 more authors

    In 2025 Ibero-American Conference on Software Engineering (CIbSE), May 2025
51. Performance and Energy Consumption Prediction of Scientific Workflows using Machine Learning

    Felipe Barbosa, Erick Damasceno, Fellipe Queiroz, and 4 more authors

    In Latin American High Performance Computing Conference, May 2025
52. Exploring Superpixel Segmentation Methods in the Context of Citizen Science and Deforestation Detection

    Hugo Resende, Isabela Borlido, Victor Sundermann, and 4 more authors

    In International Geoscience and Remote Sensing Symposium (IGARSS), Aug 2025
53. Federated Urban Flow: A Model-Centric Benchmark for Reproducible Evaluation of Deep Learning in Federated Traffic Forecasting

    Matteus VS Silva, Carlos A Astudillo, and Luiz F Bittencourt

    In 2025 3rd International Conference on Federated Learning Technologies and Applications (FLTA), Aug 2025

2024

1. How to Identify Good Superpixels for Deforestation Detection on Tropical Rainforests

   Isabela Borlido, Eduardo Bouhid, Victor Sundermann, and 4 more authors

   IEEE Geoscience and Remote Sensing Letters, Aug 2024

   Abs doi: 10.1109/LGRS.2024.3454973

   The conservation of tropical forests is a topic of significant social and ecological relevance due to their crucial role in the global ecosystem. Unfortunately, deforestation and degradation impact millions of hectares annually, requiring government or private initiatives for effective forest monitoring. However, identifying deforested regions in satellite images is challenging due to data imbalance, image resolution, low-contrast regions, and occlusion. Superpixel segmentation can overcome these drawbacks, reducing workload and preserving important image boundaries. However, most works for remote-sensing images do not exploit recent superpixel methods. In this work, we evaluate 16 superpixel methods in satellite images to support a deforestation detection system in tropical forests. We also assess the performance of superpixel methods for the target task, establishing a relationship with segmentation methodological evaluation. According to our results, ERS, GMMSP, and DISF perform best on undersegmentation error (UE), boundary recall (BR), and similarity between image and reconstruction from superpixels (SIRSs), respectively, whereas ERS has the best tradeoff with compactness index (CO) and Reg. In classification, SH, DISF, and ISF perform best on RGB, UMDA, and PCA compositions, respectively. According to our experiments, superpixel methods with better tradeoffs among delineation, homogeneity, compactness, and regularity are more suitable for identifying good superpixels for deforestation detection tasks.

2. A Satellite Band Selection Framework for Amazon Forest Deforestation Detection Task

   Eduardo Bouhid Neto, Fabio Augusto Faria, Amanda de Almeida Sales De Oliveira, and 1 more author

   In Proceedings of the Genetic and Evolutionary Computation Conference, Aug 2024

   Abs doi: 10.1145/3638529.3654000

   The conservation of tropical forests is a topic of significant social and ecological relevance due to their crucial role in the global ecosystem. Unfortunately, deforestation and degradation impact millions of hectares annually, necessitating government or private initiatives for effective forest monitoring. This study introduces a novel framework that employs the Univariate Marginal Distribution Algorithm (UMDA) to select spectral bands from Landsat-8 satellite optical sensor, optimizing the representation of deforested areas. This selection guides a semantic segmentation architecture, DeepLabv3+, enhancing its performance. Experimental results revealed several band compositions that achieved superior balanced accuracy compared to commonly adopted combinations for deforestation detection, utilizing segment classification via a Support Vector Machine (SVM). Moreover, the optimal band compositions identified by the UMDA-based approach improved the performance of the DeepLabv3+ architecture, surpassing state-of-the-art approaches compared in this study. The observation that a few selected bands outperform the total contradicts the data-driven paradigm prevalent in the deep learning field. Therefore, this suggests an exception to the conventional wisdom that ‘more is always better’.

3. Sampling Strategies Based on Wisdom of Crowds for Amazon Deforestation Detection

   Hugo Resende, Eduardo B. Neto, Fábio A. M. Cappabianco, and 2 more authors

   In 2024 37th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), Aug 2024

   Abs doi: 10.1109/SIBGRAPI62404.2024.10716332

   Conserving tropical forests is highly relevant socially and ecologically because of their critical role in the global ecosystem. However, the ongoing deforestation and degradation affect millions of hectares each year, necessitating government or private initiatives to ensure effective forest monitoring. In April 2019, a project based on Citizen Science and Machine Learning models called ForestEyes (FE) was launched with the aim of providing supplementary data to assist experts from government and non-profit organizations in their deforestation monitoring efforts. Recent research has shown that labeling FE project volunteers/citizen scientists helps tailor machine learning models. In this sense, we adopt the FE project to create different sampling strategies based on the wisdom of crowds to select the most suitable samples from the training set to learn an SVM technique and obtain better classification results in deforestation detection tasks. In our experiments, we can show that our strategy based on user entropy-increasing achieved the best classification results in the deforestation detection task when compared with the random sampling strategies, as well as, reducing the convergence time of the SVM technique.

4. Exploring emergent microservice evolution in elastic deployment environments

   Roberto Rodrigues-Filho, Iwens Sene, Barry Porter, and 3 more authors

   Journal of Systems and Software, Aug 2024

   Abs doi: 10.1016/j.jss.2024.112252

   Microservices have become an important technology to enable the dynamic composition of large-scale self-adaptive systems. Although modern microservice ecosystems provide a variety of autonomous adaptation mechanisms, when focusing on the microservice itself, they can only account for changes in the sheer increase in workload volume. On the other hand, when workload patterns change, efficient treatment requires the intervention of DevOps experts to manually evolve the internal architecture of services. Given the need to quickly adapt systems to respond to changes, solely relying on DevOps to react to workload pattern changes becomes a bottleneck for future systems. To address this issue, we advance the concept of emergent microservices, that autonomously adapt and evolve their internal architectural composition to better handle changes in the pattern of incoming requests without human intervention. We demonstrate the effectiveness of our approach by exploring this novel concept in the context of a microservice-based Smart City platform. Editor’s note: Open Science material was validated by the Journal of Systems and Software Open Science Board.

5. Performance Evaluation of Dense Linear Algebra Kernels using Chameleon and StarPU on AWS

   Vinicius Pinto, João Lima, Vanderlei Munhoz, and 3 more authors

   In Anais do XXV Simpósio em Sistemas Computacionais de Alto Desempenho, Aug 2024

   doi: 10.5753/sscad.2024.244405
6. Evolução histórica do desempenho energético de tarefas cotidianas em uma distribuição Linux

   Artur Launikas Cupelli, Fábio Nakano, Flávio Coutinho, and 1 more author

   In Anais Estendidos do XXV Simpósio em Sistemas Computacionais de Alto Desempenho, Aug 2024

   doi: 10.5753/sscad_estendido.2024.244345
7. The road to Sustainable DevOps

   Darwish Ahmad Herati, Maria Clara Aderne, and Fabio Kon

   In International Colloquium on Energy-Efficient and Sustainable Distributed Systems (E2SDS), Aug 2024
8. Smart University: A pathway for advancing Sustainable Development Goals

   Priscila Aparecida de Moraes, Flávia Pisani, and Juliana Freitag Borin

   Internet of Things, Oct 2024

   Abs doi: 10.1016/j.iot.2024.101246

   This paper aims to highlight the potential of smart university solutions in contributing to the advancement of the United Nations Sustainable Development Goals (SDGs). We surveyed solutions tailored to address each of the seventeen SDGs with an emphasis on identifying those adaptable within university settings. Our research reveals a diverse array of solutions within the university environment that align with the SDGs. Remarkably, we have identified initiatives relevant even to goals seemingly peripheral to the university’s primary mission, such as those addressing poverty eradication. On a further step, we have established a correlation between the smart university concept and the SDGs, effectively showcasing the practicality of utilizing existing solutions to enhance and accomplish all goals. We also propose definitions and a well-structured taxonomy for smart university and smart campus to support our study and guide future research.

9. Reaproveitamento de TV Boxes para Aplicação de Contagem de Pessoas na Borda em Cidades Inteligentes

   Gabriel Sato, Gustavo Luz, Luis Gonzalez, and 1 more author

   In Anais do VIII Workshop de Computação Urbana, Niterói/RJ, Oct 2024

   doi: 10.5753/courb.2024.3375
10. Repurposing of TV Boxes for a Circular Economy in Smart Cities Applications

    Gustavo P. C. P. Luz, Gabriel Massuyoshi Sato, and Luis Fernando Gomez Gonzalez al.

    Scientific Reports, Jul 2024

    Abs doi: 10.21203/rs.3.rs-4619533/v1

    In recent years, large quantities of illegal TV Box equipment have been confiscated in Brazil. According to news released in March 2024, an estimated 2.5 million TV Boxes were stored in the warehouses of the Federal Revenue Service. Typically, this equipment is destroyed, which not only incurs significant costs for the government but also generates substantial e-waste. Meanwhile, the advancement of smart city applications based on the Internet of Things (IoT) and machine learning has driven research in edge computing using hardware-constrained devices. This paper explores the feasibility of repurposing TV Boxes for edge computing in applications involving people counting in images collected by cameras. We developed a testbed consisting of 20 TV boxes to conduct a thorough evaluation of their resilience and carbon footprint compared to commonly used edge computing equipment. Our findings demonstrate that these repurposed devices can perform over five million inferences, outperforming commercially available devices in terms of carbon footprint using Brazilian energy matrix. This study underscores the innovative potential and environmental benefits of repurposing TV Boxes for smart city applications, especially when utilizing lightweight machine learning models.

11. “I am always tired”: Perceptions on Mental Health and Productivity Metrics

    Júlia Azevedo, Theo Sousa, Johny Arriel, and 3 more authors

    In Congresso Brasileiro de Software: Teoria e Prática (CBSoft), Sep 2024

    doi: 10.5753/cbsoft_estendido.2024.4105
12. Explorando o Potencial e a Viabilidade de LLMs Open-Source na Análise de Sentimentos

    Breno Braga Neves, Theo Sousa, Daniel Coutinho, and 2 more authors

    In Congresso Brasileiro de Software: Teoria e Prática (CBSoft), Sep 2024

    doi: 10.5753/cbsoft_estendido.2024.4106
13. Eyes on Code Smells: Analyzing Developers’ Responses During Code Snippet Analysis

    Vinícius Martins, Pedro Lopes Verardo Ramos, Breno Braga Neves, and 6 more authors

    In Simpósio Brasileiro de Engenharia de Software (SBES), Sep 2024

    doi: 10.5753/sbes.2024.3431
14. On the Investigation of Exception Pull Request Characteristics: Exploring the Apache Ecosystem

    João Correia, Daniel Coutinho, Alessandro Garcia, and 8 more authors

    In 2024 IEEE International Conference on Source Code Analysis and Manipulation (SCAM), Oct 2024

    doi: 10.1109/SCAM63643.2024.00023
15. Assisting the Requirements Definition and Modeling of IoT applications through the ReqM4IoT tool

    Rodrigo Nascimento, Bruno Silva, Guilherme Avelino, and 1 more author

    In Proceedings of the XXIII Brazilian Symposium on Software Quality, Oct 2024

    Abs doi: 10.1145/3701625.3701667

    The Internet of Things (IoT) represents a computational paradigm where connected devices share a common communication network to interact and provide various applications. In this context, the challenges for IoT and Software Engineering arise from the drive to create smart applications in various domains, such as smart cities, homes, and hospitals. Requirements analysis and software modeling are steps in building IoT systems that aim to ensure a good specification of the system. However, the literature lacks a well-established set of practices for these phases of IoT system development, especially regarding requirements analysis and system modeling. This results in a significant gap in knowledge and standardization of processes. Therefore, a solution that supports these stages of the IoT system development lifecycle is needed. This paper presents ReqM4IoT, a tool that integrates two previously developed techniques: ReqMLSCity, a specific technique for requirements analysis for IoT, and SysIoTML, a specific technique for software modeling for IoT. With this solution, the quality of the software can benefit significantly from using ReqM4IoT, as suggested by the initial evaluation conducted in this work. The evaluation involved five participants with different levels of experience but with a common knowledge base in Software Engineering and IoT. The experiment aimed to evaluate the feasibility, usefulness, and ease of use of the tool in the development of IoT applications. The results indicated that the tool is useful and the participants were able to effectively solve the proposed problem. Although the evaluation pointed out opportunities for improvement, these insights pave the way for further enhancement of ReqM4IoT.

16. Quantumnetsec: Quantum machine learning for network security

    Diego Abreu, David Moura, Christian Rothenberg, and 1 more author

    Authorea Preprints, Oct 2024

    doi: 10.22541/au.172540693.31989215/v1
17. Application of electronic nose and machine learning used to detect soybean gases under water stress and variability throughout the daytime

    Paulo Sergio De Paula Herrmann, Matheus dos Santos Luccas, Ednaldo José Ferreira, and 1 more author

    Frontiers in Plant Science, Oct 2024
18. Federated learning for performance behavior detection in a fog-IoT system

    Franklin Magalhães Ribeiro Junior, and Carlos Alberto Kamienski

    Internet of Things, Oct 2024
19. Evapotranspiration measurements in pasture, crops, and native Brazilian Cerrado based on UAV-borne multispectral sensor

    Gabriella Santos Arruda Lima, Manuel Eduardo Ferreira, Jepherson Correia Sales, and 5 more authors

    Environmental Monitoring and Assessment, Oct 2024
20. Human digital twin: a survey

    Ying Lin, and  outros

    Journal of Cloud Computing, Oct 2024

    doi: 10.1186/s13677-024-00358-9
21. IoTDeploy: Deployment of IoT Smart Applications over the Computing Continuum

    Francis Borges Oliveira, Marco Di Felice, and Carlos Kamienski

    Internet of Things, Oct 2024
22. State Disaggregation for Dynamic Scaling of Network Functions

    Fabrício B. Carvalho, Ronaldo A. Ferreira, Ítalo Cunha, and 2 more authors

    IEEE/ACM Transactions on Networking, Oct 2024

    doi: 10.1109/TNET.2023.3282562
23. Uncovering BGP Action Communities and Community Squatters in the Wild

    Brivaldo Alves Silva, Adriano Bastos Carvalho, Italo Cunha, and 3 more authors

    Proc. ACM Meas. Anal. Comput. Syst., Dec 2024

    Abs doi: 10.1145/3700431

    The Border Gateway Protocol (BGP) offers several "knobs" to control routing decisions, but they are coarse-grained and only affect routes received from neighboring Autonomous Systems (AS). To enhance policy expressiveness, BGP was extended with the communities attribute, allowing an AS to attach metadata to routes and influence the routing decisions of a remote AS. The metadata can carry information to (e.g., where a route was received) or request an action from a remote AS (e.g., not to export a route to one of its neighbors). Unfortunately, the semantics of BGP communities are not standardized, lack universal rules, and are poorly documented. In this work, we design and evaluate algorithms to automatically uncover BGP action communities and ASes that violate standard practices by consistently using the information communities of other ASes, revealing undocumented relationships between them (e.g., siblings). Our experimental evaluation with billions of route announcements from public BGP route collectors from 2018 to 2023 uncovers previously unknown AS relationships and shows that our algorithm for identifying action communities achieves average precision and recall of 92.5% and 86.5%, respectively.

24. Multipurpose Quantum Network Simulators: A comparative study

    Diego Abreu, Arthur Pimentel, Polyana Moraes, and 3 more authors

    In Workshop de Redes Quânticas, Dec 2024

    doi: 10.5753/wqunets.2024.3223
25. qIDS: Sistema de Detecção de Ataques baseado em Aprendizado de Máquina Quântico Híbrido

    Diego Abreu, Christian R Esteve Rothenberg, and Antônio Abelém

    In Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos (SBRC), Dec 2024

    doi: 10.5753/sbrc.2024.1353
26. Reqroute: Protocolo de roteamento por reforço para redes de entrelaçamento quântico

    Diego Abreu, Arthur Pimentel, and Antônio Abelém

    In Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos (SBRC), Dec 2024

    doi: 10.5753/sbrc.2024.1450
27. MinIndy: Automating the Deployment and Management of Hyperledger Indy Networks

    Mateus Bastos, Alan Veloso, Jeffson Sousa, and 3 more authors

    In 2024 IEEE Latin-American Conference on Communications (LATINCOM), Dec 2024

    doi: 10.1109/LATINCOM62985.2024.10770675
28. IoTinuum: The IoT Computing Continuum

    Carlos Kamienski, Ivan Zyrianoff, Luiz Fernando Bittencourt, and 1 more author

    In 2024 20th International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT), Dec 2024
29. TÉCNICAS ALIADAS CONTRA ALAGAMENTOS: FERRAMENTA DE PREVISÃO DE PRECIPITAÇÃO, AGRICULTURA URBANA E JARDINS DE CHUVA

    Patricia Sousa Marques, Karla Emmanuela Ribeiro Hora, Manuel Eduardo Ferreira, and 1 more author

    In Workshop Internacional em Sustentabilidade, Indicadores e Gestão de Recursos Hídricos (Sustentare & WIPIS 2024), Dec 2024
30. A VARIABILIDADE FLUVIOMÉTRICA DA BACIA HIDROGRÁFICA DO RIO MEIA PONTE DURANTE O FENÔMENO ENOS DECENAL (2014-2024)

    Patricia Sousa Marques, Chrystiano Alves Souza, Wesley Melo Barbosa Menezes, and 2 more authors

    In Workshop Internacional em Sustentabilidade, Indicadores e Gestão de Recursos Hídricos (Sustentare & WIPIS 2024), Dec 2024
31. PREVISÃO DE NÍVEIS NO RIO ARAGUAIA UTILIZANDO INFORMAÇÕES DE PREVISÃO DE CHUVA

    Mauro C Trindade, Aldrei M. Veiga, Angel D. Chovert, and 1 more author

    In IV Encontro Nacional de Desastres, Dec 2024
32. CARACTERIZAÇÃO DAS ÁREAS DE RESTAURAÇÃO DO CERRADO DISPONÍVEIS NA PLATAFORMA ARATICUM

    Jairo Matos Rocha, Ana Karolyna Nunes Amaral, Victória Vasconcelos Freitas, and 2 more authors

    In XX Simpósio Brasileiro de Geografia Física Aplicada, Dec 2024
33. Alocação de Recursos em Redes de Distribuição Quântica de Chaves Definidas por Software

    Arthur Pimentel, Diego Abreu, and Antônio Abelém

    In Escola Regional de Informática Norte 2 (ERIN 2), Dec 2024

    doi: 10.5753/erin.2024.4535
34. Alocação de recursos em redes de distribuição quântica de chaves multiprotocolo

    Arthur Pimentel, Diego Abreu, and Antônio Abelém

    In Workshop de Pesquisa Experimental da Internet do Futuro (WPEIF), Dec 2024

    doi: 10.5753/wpeif.2024.3230
35. AFP: Um Escalonador de Requisições de Microsserviços Guiado por Feedback

    Mayco Berghetti, Fabrício Carvalho, and Ronaldo Ferreira

    In Anais do XLII Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos, Niterói/RJ, Dec 2024

    doi: 10.5753/sbrc.2024.1554
36. Abrindo a Caixa-Preta – Aplicando IA Explicável para Aprimorar a Detecção de Sequestros de Prefixo

    Adriano Carvalho, Brivaldo Silva Jr., Carlos Silva, and 1 more author

    In Anais do XXIV Simpósio Brasileiro de Segurança da Informação e de Sistemas Computacionais, São José dos Campos/SP, Dec 2024

    doi: 10.5753/sbseg.2024.240764
37. Applications and Challenges of Sensor Networks in the Amazon Rainforest: Toward Enhanced Environmental Monitoring and Conservation

    M. Turo, A. Branco, and M. Endler

    In Proceedings of the International Colloquium on Energy-Efficient and Sustainable Distributed Systems (E2SDS), Dec 2024

    Hybrid event, November 26–27, 2024
38. MinIndy: Um Framework para Automatizar a Implantação e o Gerenciamento de Redes Blockchain Hyperledger Indy

    Alan Veloso, Jeffson Sousa, Bruno Evaristo, and 3 more authors

    In Workshop em Blockchain: Teoria, Tecnologias e Aplicações (WBlockchain), Dec 2024

    doi: 10.5753/wblockchain.2024.329

2023

1. CharM – Evaluating a model for characterizing service-based architectures

   Thatiane Oliveira Rosa, Eduardo Martins Guerra, Filipe Figueiredo Correia, and 1 more author

   Journal of Systems and Software, Dec 2023

   doi: 10.1016/j.jss.2023.111826
2. O AVANÇO DA PASTAGEM SOBRE AS ÁREAS DE PRESERVAÇÃO PERMANENTE NA BACIA HIDROGRÁFICA DO RIO VERMELHO – ESTADO DE GOIÁS

   Arlete Gomes do Nascimento Vieira, and Manuel Eduardo Ferreira

   Sociedade e Território– Natal, Dec 2023
3. Satellite, UAV, and Geophysical Data to Identify Surface and Subsurface Hydrodynamics of Geographically Isolated Wetlands: Understanding an Undervalued Ecosystem at the Atlantic Forest-Cerrado Interface of Brazil

   Lucas Moreira Furlan, Manuel Eduardo Ferreira, César Augusto Moreira, and 3 more authors

   Remote Sensing, Dec 2023