Rafael Veiga

Rafael Veiga

Auto-generated profile, based on publicly available information

Rafael Veiga

Federal University of Para

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 May 2023
Verified
Conference Paper
Open access
PDF file
RVRafael Veiga
Rafael Veiga
CBCristiano Bonato Both
Cristiano Bonato Both
IMIago Medeiros
Iago Medeiros
DRDenis Rosário
Denis Rosário
Eduardo CerqueiraEduardo Cerqueira
Eduardo Cerqueira
A smartphone can collect behavioral data without requiring additional actions on the user’s part and without the need for additional hardware. In an active or continuous user authentication process, information from integrated sensors, such as touch, and gyroscope, is used to monitor the user continuously. These sensors can capture behavioral (touch patterns, accelerometer) or physiological (fingerprint, face) data of the user naturally interacting with the device. However, transferring data from multiple users’ mobile devices to a server is not recommended due to user data privacy concerns. This paper introduces an Federated Learning (FL) approach to define a user’s biometric behavior pattern for continuous user identification and authentication. We also evaluate whether FL can be helpful in behavioral biometrics. Evaluation results compare CNNs in different epochs using FL and a centralized method with low chances of wrong predictions in user identification by the gyroscope.

8 March 2024 in Journal of the Brazilian Computer Society
Verified
Journal Article
Open access
PDF file
RVRafael Veiga
Rafael Veiga
JSJohn Sousa
John Sousa
RMRenan Morais
Renan Morais
LBLucas Bastos
Lucas Bastos
WJWellington Lobato Junior
Wellington Lobato Junior
Eduardo CerqueiraEduardo Cerqueira
Eduardo Cerqueira
There is a exponential growth of data usage, specially due to the proliferation of connected applications with personalized models for different applications. In this context, Federated Learning (FL) emerges as a promising solution to enable collaborative model training while preserving the privacy and autonomy of participating clients. In a typical FL scenario, clients exhibit significant heterogeneity in terms of data distribution and hardware configurations. In this way, randomly sampling clients in each training round may not fully exploit the local updates from heterogeneous clients, resulting in lower model accuracy, slower convergence rate, degraded fairness, etc. In addition, malicious users could disseminate incorrect weights, which may decrease the accuracy of aggregated models and increase the time for convergence in FL. In this article, we introduce Resilience-aware Client Selection Mechanism for non-IID data and malicious clients in FL environment, called RICA. The proposed mechanism employs data size and entropy as criteria for client selection. In addition, RICA relies Centroid-Based Kernel Alignment (CKA) to identify and exclude potentially malicious clients. Our evaluation shows an improvement of 125% in Accuracy values in a scenario of malicious clients, which means the RICA+CKA demonstrates a more stable and resilient approach, reaching 90% accuracy in a few rounds compared to the default average approach, reached only around 30%. Therefore, results of the behavior of RICA+CKA in different datasets show the evaluation of different numbers of clients reaching around 90% while the other approach does not pass the 50% Accuracy.

4 March 2024 in Journal of Internet Services and Applications
Verified
Journal Article
Open access
PDF file
ABAlex Barros
Alex Barros
RVRafael Veiga
Rafael Veiga
RMRenan Morais
Renan Morais
Denis RosarioDenis Rosario
Denis Rosario
Eduardo CerqueiraEduardo Cerqueira
Eduardo Cerqueira
Integrating big data and deep learning across various applications has significantly enhanced intelligence and efficiency in our daily lives. However, it also requires extensive data sharing, raising significant communication and privacy concerns. In this context, Federated Learning (FL) emerges as a promising solution to enable collaborative model training while preserving the privacy and autonomy of participating clients. FL facilitates collaborative model training by enabling data to be trained locally on devices, eliminating the need for individual information sharing among clients. A client selection mechanism strategically chooses a subset of participating clients to contribute to the model training in each learning round. However, an efficient selection of clients to participate in the training process directly impacts model convergence/accuracy and the overall communication load on the network. In addition, FL faces challenges when dealing with non-Independent and Non-Identically Distributed (non-IID) data, where the diversity in data distribution often leads to reduced classification accuracy. Hence, designing an efficient client selection mechanism in a scenario with non-IID data is essential, but it is still an open issue. This article proposes a Model Efficiency through Selective Federated Learning Algorithm called MESFLA. The mechanism employs a Centered Kernel Alignment (CKA) algorithm to search for similar models based on data weight or similarity between models, i.e., grouping participants with comparable data distributions or learning objectives. Afterward, MESFLA selects the clients with more relevance in each group based on data weight and entropy. Our comprehensive evaluation across multiple datasets, including MNIST, CIFAR-10, and CIFAR-100, demonstrates MESFLA's superior performance over traditional FL algorithms. Our results show an accuracy improvement and a minor loss in each client aggregation of the new global model sent to clients with a difference of 3 rounds using the Data Weight in comparison with the other selection methods.

24 May 2024
Verified
Conference Paper
Open access
PDF file
KVKleber Vilhena
Kleber Vilhena
CRCarlos Rocha
Carlos Rocha
RVRafael Veiga
Rafael Veiga
LBLucas Bastos
Lucas Bastos
Eduardo CerqueiraEduardo Cerqueira
Eduardo Cerqueira
DRDenis Rosário
Denis Rosário
O-RAN is an architecture that promotes interoperability and openness in 5G Radio Access Networks (RAN) using scheduling, disaggregation, and virtualization. RICs (RAN Intelligence Controllers) offer solutions such as Machine Learning (ML), traffic steering, anomaly detection, and QoS (Quality of Service) support. Novel intelligent handover strategies are critical to the success of 5G or even 6G O-RAN-based networks. This paper proposes and evaluates an intelligent handover algorithm for O-RAN environments. It leverages an LTE testbed featuring O-RAN architecture to assess downlink and uplink performance across various User Equipment (UE) scenarios. The proposed scheme was implemented and tested using ns-O-RAN, an O-RAN system integrated with the NS-3 simulator. Our simulator results demonstrate a throughput and delay enhancement compared to traditional handover methods across various scenarios involving 50 to 100 UEs.

19 May 2025
Verified
Conference Paper
Open access
PDF file
RMRenan Morais
Renan Morais
RVRafael Veiga
Rafael Veiga
LBLucas Bastos
Lucas Bastos
Denis RosarioDenis Rosario
Denis Rosario
ECEduardo Cerqueira
Eduardo Cerqueira
Traditional Machine Learning (ML) relies on centralized systems, requiring large amounts of sensitive data. In contrast, edge processing minimizes data transmission by performing computations locally, improving efficiency. In this context, Federated Learning (FL) is a decentralized machine learning approach that allows clients to train models collaboratively while keeping their data private. However, FL faces challenges such as high communication costs due to the large number of model updates between clients and the central server. In this paper, we introduce an efficient method for model compression for FL scenarios called HUFE-FL, which applies quantization to transform model parameters into discrete values, reducing their size, followed by Huffman encoding. This further compresses the model by efficiently encoding frequent symbols with fewer bits. Our main results show that HUFE-FL achieves up to 77.48% reduction in data transmission compared to traditional FL methods, with minimal loss in accuracy.

9 January 2026 in Journal of Internet Services and Applications
Verified
Journal Article
Open access
PDF file
RVRafael Veiga
Rafael Veiga
RMRenan Morais
Renan Morais
LBLucas Bastos
Lucas Bastos
Susana SargentoSusana Sargento
Susana Sargento
Denis RosarioDenis Rosario
Denis Rosario
Eduardo CerqueiraEduardo Cerqueira
Eduardo Cerqueira
Federated Learning (FL) is a machine learning training method that uses a collaborative model for training across a diverse set of clients, while preserving data privacy in accordance with the General Data Protection Regulation (GDPR) for clients' data. The classification and similarity within many clients can become an issue, often leading to decreased model accuracy and slower convergence, or reducing the number of parameters to the point of not learning anything during aggregation. However, the presence of non-IID data and malicious clients poses significant challenges to the significance of generalization models and distribution data. Malicious clients can perform poisoning attacks by sending harmful model updates that degrade the performance of the global model. This article introduces MONZA, a scoring system designed to detect and exclude malicious clients in FL environments. In a scenario where clients can also engage in various attacks, including model poisoning and data poisoning, this can lead to incorrect training. The proposed method uses cosine similarity to calculate client scores. It employs L2 normalization to identify biased models; in some cases, the similarity is not sufficient to classify a client, effectively filtering out malicious participants before aggregation and implementing a penalty with a quarentine method. Our evaluation shows that MONZA achieves an accuracy of 54.5% in a scenario with 30% malicious clients, while zPROBE (i.e., existing resilient methods) only reached an accuracy of 50%. Furthermore, MONZA reduces the simulation execution time by 66% and the computational effort to 83 MFLOP/s compared to zPROBE, which demonstrates to be a more efficient and resilient method. These results confirm that MONZA maintains the integrity of the model, making a security aggregation while minimizing resource consumption in malicious FL settings.