Miguel Elias M. Campista (UFRJ) will present a webinar on November 26

The topic of the talk is “Federated Vehicle Learning: From Theory to Practice.”

Connected vehicles generate large volumes of sensitive data, such as images, location, and driver behavior. Centralizing this data can compromise privacy and overload the network. This lecture presents federated learning as an alternative for personalizing in-vehicle applications without the data leaving its sources—the clients—and discusses the rationale behind this approach for the Internet of Things in vehicles.

The discussion is based on the AVADiP project, which addresses challenges from end to end: from the installation of cameras inside vehicles and hardware for training and inference, through the generation and transmission of data over mobile networks, to federated training and the selection and management of participating clients.
The talk also addresses two open challenges. The first is the need for source labeling when using supervised models; the second is the automation of model discovery using large language models (LLMs). Throughout the talk, the identification of vulnerable road users (VRUs), such as pedestrians, cyclists, and motorcyclists, serves as a case study for driver assistance. The presentation concludes with perspectives on the future development of AVADiP and its application in other areas, particularly in smart cities.

The speaker

Miguel Elias M. Campista earned his master’s and doctoral degrees in electrical engineering from the Federal University of Rio de Janeiro. In 2009, he served as an Adjunct Professor in the Department of Telecommunications Engineering at the Federal University of Fluminense, and he is currently an Associate Professor in the Department of Electronic and Computer Engineering at the Polytechnic School of UFRJ and a Full Professor in the Electrical Engineering Program at COPPE/UFRJ.

Miguel spent 2012 as a postdoctoral fellow at Sorbonne Université; he was named FAPERJ’s “Young Scientist of Our State” twice in a row (2015–2018, 2018–2021), and is an affiliate member of the Brazilian Academy of Sciences (ABC) (term 2018–2022), coordinator of the Special Committee on Computer Networks and Distributed Systems (CE-RESD) of the SBC (2021–2023), a member of the board of directors of the National Laboratory for Computer Networks (LARC) (2019–2025) and member of the Board of Directors (CADM) of the National Education and Research Network (RNP) (2021–2024).

He is currently the coordinator of the Electrical Engineering Program at COPPE/UFRJ, Associate Editor of Springer’s journal *Annals of Telecommunications*, a CNPq PQ-B fellow, a senior member of the IEEE, and a “Scientist of Our State” recipient from FAPERJ. Prof. Miguel coordinates research projects funded by national funding agencies such as CNPq, CAPES, and FAPERJ, as well as through partnerships with industry. Prof. Miguel has served as a visiting professor at international institutions such as the Sorbonne University and the University of La Rochelle, both in France. His primary field of expertise is computer science, and his areas of interest include computer networks, data and network science, cybersecurity, and applied machine learning.

 

Masking by clustering and labeling with LLMs for sharing network incident datasets

How can cybersecurity data be shared to expand knowledge about threats and vulnerabilities without exposing information that could compromise the security of an infrastructure? This cybersecurity dilemma inspired the work of Breno Manhaes (UFRJ), Guilherme Thomaz (UFRJ), and Miguel Elias Mitre Campista (UFRJ), published in the paper titled “Masking by Clustering and Labeling with LLMs for Sharing Network Incident Datasets” (SBRC 2026).

Sharing information about digital threats and vulnerabilities is one way to strengthen internet security. When different organizations have access to this data, they can identify problems, recognize patterns, and develop more effective solutions. But there is a challenge: how can this information be shared without exposing details that could compromise the security of a network or organization?

Organizations often avoid sharing their security logs for fear of exposing information about their infrastructure or violating data protection laws, such as the LGPD in Brazil and the GDPR in Europe. Studies show that legal restrictions and the lack of privacy safeguards are among the main obstacles to the exchange of information about cyber threats. On the other hand, vulnerability scanning tools, such as OpenVAS and Nessus, generate reports containing a great deal of information in plain text. While this text is important for understanding the context of a security issue, it may contain sensitive data, such as internal IP addresses, software versions, usernames, and folder and file paths. This information could make it possible to identify the organization involved.

Traditional anonymization techniques, such as deleting columns or replacing information with codes (hashes), can end up eliminating data that is important for analysis. Manual review of texts to remove sensitive details and replace them with more generic information, on the other hand, is time-consuming and labor-intensive, making it difficult to apply given the large number of records produced daily.

These dilemmas served as the starting point for the research by Manhaes, Thomaz, and Campista, conducted in partnership with the National Education and Research Network (RNP). The institution uses tools that perform security scans on its networks, identifying potential vulnerabilities and generating large volumes of data. This information could be useful to other researchers and professionals in the field of cybersecurity. However, before sharing this data, it is necessary to remove or modify elements that could reveal critical information about the network infrastructure.

O problema é que esse processo não é tão simples quanto retirar nomes, endereços ou identificações. Algumas informações aparentemente inofensivas podem, quando combinadas com outros dados disponíveis publicamente, permitir que alguém descubra detalhes que deveriam permanecer protegidos. Isso é especialmente difícil quando se trata de textos livres, ou seja, informações escritas sem um formato padronizado. Mesmo que não contenham uma identificação explícita, esses textos podem revelar características importantes sobre sistemas, equipamentos e vulnerabilidades.

Proteger sem perder informação

Anonimizar dados significa modificar ou retirar informações que possam identificar pessoas, organizações ou estruturas específicas. Mas existe uma questão importante: quanto mais informações são removidas, menor pode ser a utilidade dos dados para quem precisa analisá-los. Os pesquisadores precisam, portanto, encontrar um equilíbrio entre proteger informações sensíveis e preservar características que permitam seu estudo. Em outras palavras, não basta esconder os dados: é preciso fazer isso sem comprometer sua utilidade.

To address this challenge, the researchers developed an algorithm with a special focus on free-text fields. The solution uses a technique known as embeddings, which is widely employed in artificial intelligence systems that work with language. Put simply, this technique transforms text into numerical representations called vectors. These vectors make it possible to identify similarities in meaning between different texts: those that deal with similar topics tend to be close to one another in this mathematical space.

As a result, the algorithm can identify texts with similar characteristics, group them together, and assign more generic labels to these groups. Thus, instead of disclosing a very specific description—which could reveal details of a vulnerability—it is possible to replace it with a broader category, while preserving some of the meaning necessary for further analysis.

Desafios envolvidos

Uma das dificuldades da pesquisa foi avaliar a qualidade dos agrupamentos e dos rótulos gerados automaticamente. Como o método utiliza aprendizado não supervisionado, não existe uma classificação prévia considerada correta para servir de comparação. Os pesquisadores precisaram, então, desenvolver outras formas de verificar se os grupos identificados faziam sentido e se os rótulos representavam adequadamente o conteúdo reunido em cada um deles.

The evaluation work continues to be carried out in collaboration with RNP. The team is also investigating whether there are naturally identifiable patterns in the data that could serve as a benchmark for measuring the quality of the clusters. In addition, it plans to test the algorithm on public datasets.

Although it was developed to address a specific cybersecurity problem, the method may have applications in other contexts involving large volumes of textual information. Companies and institutions, for example, could use it to organize documents, classify PDF files, and identify similar content.
The research shows that sharing data is important for generating knowledge and strengthening digital security, but this sharing requires caution. The challenge is not simply to remove sensitive information, but to find ways to protect what needs to be preserved without discarding what may be useful for research.

ICoNIoT has reached the halfway point of its term with a positive evaluation from CNPq and numerous ongoing research partnerships Prof.

Dr. Nelson Fonseca provides a halfway review of the INCT ICoNIoT project

As the year comes to a close, Prof. Dr. Nelson Fonseca shares a message with INCT ICoNIoT researchers, reflecting on the project’s achievements and progress over the past two and a half years.

In this video, Prof. Nelson Fonseca notes that we are exactly halfway through the INCT ICoNIoT’s term, which is why he conducts a self-assessment and proposes a plan for the INCT’s future. The project was very well evaluated by CNPq, with emphasis on scientific output, the awards received, and the interaction among various universities. With 20 participating universities and approximately 90 researchers, it was possible to establish cooperative networks involving various joint projects that were only made possible thanks to the interaction facilitated by ICoNIoT.

Prof. Nelson highlights activities aimed at strengthening these partnerships, citing the weekly internal newsletter sent to researchers, the seminars coordinated by Prof. Carlos Kamienski, and the workshops—both in-person at the SBRC and online, which take place at the end of the year. He also highlights the ongoing efforts of the research area coordinators, whose constant dedication creates opportunities for researchers to get to know one another and collaborate on joint projects.

He thanks everyone—the coordinators and the steering committee—for their support, and reminds INCT researchers that there are calls for proposals every year in which they can participate. Prof. Nelson also asks that anyone with suggestions share them, so that we can carry out even more joint initiatives and projects during the second half of the INCT’s term.

Chaos-K8s: Systematic Assessment of Availability in Kubernetes Clusters

In recent years, the reliance on digital services in critical sectors such as e-commerce, financial services, e-government, and healthcare applications has made availability a requirement as important as performance and security. Recent outages of major cloud platforms and widely used services such as Amazon Web Services (AWS) affected approximately 500 companies due to the disruption—some of which in Brazil included iFood, Mercado Livre, Itaú, and others. Even though they were short-lived, these outages caused global downtime and resulted in millions in losses, highlighting how just a few minutes of disruption can compromise entire business chains and user trust. In this scenario, availability and recovery time metrics have evolved from being merely operational indicators to becoming strategic inputs in capacity planning and the definition of service level agreements (SLAs).

This paper, authored by Jonas César Gomes Nunes (UFPI), Iure de Sousa Fé (UFPI), Lucas Lopes (UFPI), José Miqueias (UFPI), Elias P. Duarte Jr. (UFPR), and Francisco Airton Silva (UFPI), introduces Chaos-K8s, an innovative tool developed to systematically assess availability in Kubernetes clusters.

The solution focuses on automating infrastructure discovery and the empirical calibration of recovery times, allowing administrators to test failures in the control plane, nodes, and pods (a pod is a small part of an application containing everything it needs to function). By performing controlled simulations, the system helps quantify how different configurations impact the end-user experience and service resilience.

The study validates the tool’s effectiveness through mathematical modeling, demonstrating a high degree of statistical accuracy between practical tests and theoretical results. Ultimately, the work offers a structured methodology for companies to predict outages and optimize the robustness of their cloud operations.

Automatic Discovery and Calibration

The tool automatically maps the cluster components (control plane components such as etcd and kube-apiserver, worker nodes, and deployed applications) to generate an initial configuration without manual effort. Unlike other tools that use generic values, Chaos-K8s can perform a calibration phase where it injects controlled failures to measure the actual Mean Time to Recovery (MTTR) of each component in the specific environment. This ensures that availability calculations reflect the reality of the evaluated infrastructure.

The system acts as an external orchestrator that interacts with the cluster to apply failures at different layers: to Pods—terminating processes to force a restart by Kubernetes; Worker Nodes—restarting critical services such as kubelet and containerd via SSH; and the Control Plane—interrupting core processes such as etcd and kube-apiserver.

Monitoring and Availability Criteria

During testing, Chaos-K8s uses a monitoring module that continuously checks whether the system meets the availability criteria defined by the administrator (for example, the minimum number of healthy replicas of a pod).

If the number of replicas falls below this threshold, the system is considered unavailable.

Validation via Mathematical Modeling

To ensure accuracy, the empirical results obtained by Chaos-K8s are validated using Stochastic Petri Nets (SPN). This modeling allows for a comparison of the data collected in real-world tests with analytical values, providing a robust statistical basis (with 95% confidence in the case study) for availability estimates. Thus, Chaos-K8s not only injects “chaos” but also provides an integrated workflow to quantify the impact of architectural decisions—such as the number of pod replicas—on the system’s overall resilience.

Future Developments

As part of future work, we intend to expand the evaluation of Chaos-K8s in larger-scale enterprise environments, considering real-world applications and more complex workloads. We aim to investigate its behavior during prolonged runs, analyzing availability, reliability, recovery time, and energy consumption, in order to assess its scalability and applicability in real production scenarios.

ICoNIoT’s proposal for the São Paulo School of Advanced Science is approved

This week, we were thrilled to receive the news that ICoNIoT’s proposal for the São Paulo School of Advanced Science (ESPCA), titled “São Paulo School of Advanced Science in Artificial Intelligence for Communication Networks,” has been approved. The project is coordinated by researcher Carlos Kamienski (UFABC), coordinator of Education and Outreach at INCT ICoNIoT. The courses are scheduled to take place between September 13 and 22, 2027.

The review report approving the School highlights the quality of the proposed content, which is multidisciplinary and spans the fields of computer networks, telecommunications, and AI. The reviewers stated that the project has great potential to contribute to increasing the visibility of research and to encourage participation in doctoral programs and postdoctoral fellowships in the State of São Paulo, especially for students from other countries and Brazilian states. They also emphasized that the proposal offers a framework capable of providing significant benefits to participants, students, and researchers in the state of São Paulo. The topics to be addressed by the speakers were deemed current and relevant.

The review highlighted that the scientists responsible for the courses (speakers/guest lecturers) are internationally renowned in their fields, and that the proposed group of scientists (speakers/guest lecturers) reflects a good balance between foreign and Brazilian scholars.

Student Participation

Each ESPCA is expected to host up to approximately 100 students, and priority will be given to proposals that demonstrate significant participation by students from outside Brazil. At least 50% of participants are expected to come from other countries.

Participating students will have the opportunity to present the results of their ongoing research in poster sessions included in the program.

Dissemination

It was emphasized that the proposal offers new ways to disseminate information and ideas in a manner that could not be achieved through traditional communication channels such as scientific publications and presentations at scientific events. In this regard, the focus is on the group activities and discussions, the Student Lightning Talks, the field trips, and the roundtable discussions with industry representatives, as outlined in the project. These include guided field trips to major centers, research institutions, academic institutions, and companies relevant to the topics covered. The following field trips are planned: 1) Telecom Operator; 2) Samsung / Motorola R&D Center; 3) Smart City / Urban Operations Center (COP-SP).

The report also highlighted that in-person discussions and hands-on activities allow for interaction and the exchange of information and ideas in ways that are not possible through usual communication channels, such as scientific publications and presentations at scientific events.

How ESPCA Works

FAPESP’s São Paulo School of Advanced Science (ESPCA) Program aims to support the organization of short-term courses that contribute to the advanced training of students and young researchers, promoting interaction with researchers of recognized excellence and expanding the international reach of research conducted in the state of São Paulo.

The ESPCA program has already funded dozens of São Paulo Schools of Advanced Science, which have brought together thousands of students and researchers from various countries. ESPCA provides funding for the organization of short-term courses in advanced research across different fields of knowledge in the state of São Paulo.

 

Marcelo Fernandes (UFRN) will present a webinar on October 29

The topic of the presentation is “Intelligent Horizontal Scaling and Control-Aware Sizing in IIoT Edge Systems”

The growing adoption of distributed architectures in the Industrial Internet of Things (IIoT) has expanded the use of edge computing environments for running intelligent services and control systems. However, the variability of computational loads, resource constraints, and time requirements of these applications pose challenges for the dynamic management of infrastructure.

This talk will present the evolution of horizontal scaling strategies in Kubernetes environments, addressing intelligent mechanisms based on artificial intelligence techniques—including fuzzy logic, machine learning, and deep learning—to optimize resource allocation, predict computational workloads, and dynamically adapt to the demands of distributed systems. Next, we will discuss the specific challenges of running distributed control systems on edge infrastructures, highlighting the limitations of traditional scaling strategies based exclusively on CPU utilization. The presentation will delve into the concept of control-aware sizing, in which control quality (Quality of Control, QoC) metrics, temporal constraints, and computational costs are considered jointly to determine the appropriate number of shared controllers.

Experimental results will be presented, and prospects for the development of resource management mechanisms capable of integrating scalability, computational efficiency, and control requirements in IIoT systems will be discussed.

The speaker

The speaker, Marcelo Augusto Costa Fernandes, is an associate professor in the Department of Computer Engineering and Automation at the Federal University of Rio Grande do Norte (UFRN). He holds a Ph.D. in Electrical Engineering from the State University of Campinas (UNICAMP) and has served as a visiting researcher at King’s College London (United Kingdom), Harvard University (United States), and the University of Granada (Spain). He is currently the director of the Center for Artificial Intelligence and Data Science (nIACD) at the Digital Metropolis Institute (IMD/UFRN) and coordinator of the InovAI Lab. His research focuses on artificial intelligence, deep learning, digital signal processing, embedded systems, reconfigurable computing, and bioinformatics. He is an associate researcher at INCT ICoNIoT and participates in national and international projects focused on the development of intelligent systems and AI technologies.

CAIROS: Adaptive Control of Federated Learning in Wireless Networks

The paper by Lucas Airam C. de Souza (UFRJ), Nadjib Achir (INRIA Saclay, France), Miguel Elias Mitre Campista (UFRJ), and Luis Henrique Maciel Kosmalski Costa (UFRJ) was presented at SBRC 2026. The research was also featured in a mini-course at the event, which included a hands-on activity using a simulation of the mobility studied by the researchers.

The research addresses one of the challenges of Federated Vehicle Learning: the high mobility of users. As vehicles move, communication conditions can vary significantly, affecting the time required for local training and for transmitting updates to the server. When an update exceeds the time limit set for a training round, it may end up being discarded.

To address this problem, the researchers propose CAIROS, a strategy that uses a Long Short-Term Memory (LSTM) model to dynamically estimate the network and computing conditions of each client.
The proposal is to use information available during training—such as the maximum time limit set for the round, the time already elapsed, the time remaining, the network conditions experienced by the client, and the amount of available data—to estimate the time needed to complete processing and transmit the update. In this way, the system aims to prevent the client’s work from being lost due to exceeding the time limit.

In CAIROS, the LSTM model acts as a dynamic delay estimator. Its function is to predict the communication conditions for each vehicle client and estimate whether the time required to process and transmit the local update might exceed the limit set for the round.
Based on this prediction, the vehicle can decide whether to continue local training or send the parameters to the server early. The strategy thus aims to reduce the discarding of updates and make better use of the work performed by the clients.
In the experiments presented by the researchers, CAIROS increased training efficiency and achieved gains of up to 25% in the accuracy of the global model in scenarios with severe time constraints, compared to the traditional FedAvg algorithm.

 

General coordinator of INCT ICoNIoT attended the INCT 2026 Seminar

Prof. Nelson Fonseca presented ICoNIoT in Brasília at the event organized by CNPq

The INCT 2026 Seminar was held in Brasília from September 22 to 24. The event brought together coordinators and representatives from the currently active National Institutes of Science and Technology, as well as officials from CNPq, the Ministry of Science, Technology, and Innovation (MCTI), Capes, the National Council of State Research Support Foundations (Confap), and other partner institutions.

O coordenador geral do ICoNIoT, prof. Nelson Fonseca, esteve presente realizando apresentação sobre o INCT ICoNIoT. Ele falou para o público do seminário sobre a estrutura do ICoNIoT; as produções realizadas até o momento; os pesquisadores formados; a interação do INCT com o setor produtivo – cooperações e parcerias com empresas privadas, órgãos públicos, secretarias; as iniciativas de internacionalização do INCT ICoNIoT – que incluem universidades presentes em quase 20 países, e as contribuições do INCT à inovação em políticas públicas. 

Segundo o prof. Nelson, “o workshop organizado pelo CNPq teve como participantes os coordenadores dos INCT vigentes, os gestores e os funcionários do programa. O INCT ICoNIoT foi selecionado para apresentação e recebeu um retorno muito positivo pela sua atuação e excelência na sua área de atuação”. 

Impacts and Contributions of ICoNIoT

The scientific and technological impacts of INCT ICoNIoT were also highlighted; the institute has been pushing the frontiers of knowledge in areas such as federated learning, 5G/6G and OpenRAN networks, cybersecurity, edge, fog, and cloud computing, IoT and vehicular systems, optical networks, and distributed AI. The Institute has contributed to the development of new methodologies, analytical models, experimental platforms, databases, and computational tools focused on network traffic analysis, anomaly detection, DDoS attack prediction, protection of IoT and IIoT devices, privacy preservation, and monitoring of distributed systems. These results have served as the foundation for new research projects, the training of highly qualified professionals, and the transfer of knowledge to strategic sectors of the digital economy.

The scientific outreach initiatives aimed at bridging the gap between science, technology, education, society, and the productive sector were highlighted. The strategy aims to help disseminate the work carried out at the INCT through communication in languages appropriate for both internal and external audiences; foster a culture of collaboration among researchers to ensure that the knowledge produced leads to new partnerships; support a strategy for training professionals in the INCT’s areas of focus; and help enhance other aspects of science and technology management at the INCT, such as technology transfer.

Check out some photos from the seminar:

Professors Edmundo Monteiro and Marília Curado, international collaborators at ICoNIoT, will lead seminars at the UFRGS Institute of Informatics on October 2

Prof. Edmundo will deliver a lecture titled “Telecommunications Network Management in the Age of AI,” and Prof. Marília Curado will present a lecture titled “Technology Transfer: From Scientific Innovation to Impact on Society.”

Prof. Edmundo is a full professor in the Department of Computer Engineering at the Faculty of Science and Technology of the University of Coimbra, in Portugal, and an international collaborator at INCT ICoNIoT. Prof. Marília is a full professor in the Department of Computer Engineering at the University of Coimbra and director of the Computer Science and Systems Laboratory at the Pedro Nunes Institute.

Read the details of the following seminars:

SemINF — Seminar Series of the UFRGS Institute of Informatics

Date: Friday, October 2, 2026
Location: Auditorium 0, Building 43412, Block 4, Vale Campus
Live stream: https://youtube.com/@seminariosINF

────────────────────
10:30 a.m. — Technology Transfer: From Scientific Innovation to Impact on Society
Speaker: Prof. Marília Curado

Abstract
This presentation highlights projects developed at the Pedro Nunes Institute’s Computer Science and Systems Laboratory, including intelligent computing solutions applied to areas such as sustainable tourism, resource management, 6G connectivity, and healthcare. It addresses technology transfer as a bridge between scientific research and tangible impact on society, highlighting how intelligent digital systems can address complex challenges, support decision-making processes, and foster new forms of collaboration among citizens, industry, institutions, and organizations.

Short Biography
Marília Curado is a full professor in the Department of Computer Engineering at the University of Coimbra and director of the Computer Science and Systems Laboratory at the Pedro Nunes Institute. Her research covers topics such as resource allocation and management in computer networks, resilience and energy efficiency in networks, as well as 5G and the Internet of Things. She also participates in national and international projects and is active in the scientific community dedicated to networks and communication systems.

11:15 a.m. — Telecommunications Network Management in the Age of AI
Speaker: Prof. Edmundo Monteiro
────────────────────

Abstract
This presentation explores the evolution of telecommunications network and service management, from the earliest manual configurations to today’s increasingly automated environments supported by artificial intelligence. It will cover advancements such as automation, software-defined networks, and telemetry, which have paved the way for the use of machine learning in predicting failures, optimizing performance, and enhancing security. The presentation concludes with a reflection on the transformation of network operations by AI and the new challenges associated with increasingly autonomous telecommunications ecosystems.

Short Biography
Edmundo Monteiro is a full professor in the Department of Computer Engineering at the Faculty of Science and Technology of the University of Coimbra and the director of the institution. With over 40 years of experience in teaching, research, and technology transfer, he works in the fields of computer science, telecommunications, computer networks, network and service management, and cybersecurity. He is the author and editor of six books and has over 200 scientific publications to his credit.