Programmable networks enable the deployment of in-network ordering services that can improve the performance of consensus protocols. However, existing approaches either tolerate only crash faults or require additional replica coordination to handle Byzantine failures, limiting their performance gains. In this work, we present NsoBFT, a Byzantine fault-tolerant (BFT) consensus protocol that leverages secure in-network ordering through a USIG (Unique Sequential Identifier Generator)-based message ordering service deployed at the network layer. The ordering service is implemented using Network Functions Virtualization (NFV), allowing it to operate as a virtualized network function that provides isolation and flexible integration into modern infrastructures. By combining NFV-based in-network sequencing with a USIG-protected sequencer, NsoBFT eliminates the extra replica confirmation step required by prior BFT in-network protocols, shortening the critical execution path while preserving consensus safety and liveness properties. We implemented NsoBFT and evaluated it against NeoBFT and NOPaxos, two representative in-network consensus protocols. Experimental results show that NsoBFT consistently achieves lower latency and higher throughput than NeoBFT, significantly narrowing the performance gap between crash-tolerant and Byzantine-tolerant in-network consensus.
This work presents PULSE-FHIR, an open-source web platform that integrates Fast Healthcare Interoperability Resources (FHIR)-compliant electrocardiogram (ECG) recordings with temporally synchronized visualization of clinical and AI-generated annotations. PULSE-FHIR supports multi-lead display, including 12-lead ECG recordings, flexible temporal navigation, and the overlay of beat-level annotations from reference databases, such as MIT-BIH, or neural network outputs. Additionally, it interfaces with FHIR servers, such as HAPI-FHIR, using SMART on FHIR when authentication and authorization are required. Precise alignment of raw signals and annotation markers enables direct comparison of AI classifications with reference labels, supporting ECG review and validation workflows in Internet of Medical Things (IoMT) ecosystems. The manuscript details the system architecture, FHIR representations for ECG signals and annotations, the implementation of the visualization interface, and experimental validation using MIT-BIH recordings. Results demonstrate effective temporal synchronization and rendering latencies below 200 ms for complete one-minute ECG records. The evaluation also indicates the importance of incremental retrieval strategies for handling long-term ECG recordings. A QR code links to a demonstration video and the project source code on GitHub.
The rapid development of geographic localization techniques has provided the means for generating and increasingly collecting massive georeferenced datasets. Geographic Information Systems (GIS) have been used to process and represent this kind of data visually on maps in the search for a better understanding of the possible relationships between geographic areas and the variables of the represented application domain. The objective of its use is to reduce the time it takes to analyze information, in addition to enabling access to spatial patterns that would be difficult to identify with simplified data listings or graphs (e.g., bars, lines, etc.) that are not very informative in a spatial context. Some modern versions of GIS have sought the support of data mining, among which Association Rule Mining (ARM) is a good option due to its relative simplicity and transparency. The combination of GIS and ARM can support the development of more sophisticated analyses, improve spatial analysis, and enable the identification of patterns of interest and the extraction of valuable knowledge. However, there are still a few works in the literature that propose this type of solution, not exploring the full potential that this approach could offer to visually communicate patterns and trends that are difficult to identify through conventional data analysis methods, such as graphs and tables. To address this gap, the objective of this work is to introduce MVAR-Geo, a method for visualizing association rules in georeferenced data. To demonstrate how this proposal can benefit the analysis of association rules with georeferenced attributes, we apply the method in three case studies to conduct spatial analyses of data from different application domains using levels of geographic granularity in accordance with each context analyzed. The case studies demonstrated that MVAR-Geo significantly enhances the ability to interpret spatial data by enabling the integration of ARM with GIS. The use of interest measures from association rules in geographic visualizations, particularly the lift measure, has proven essential for distinguishing genuinely strong associations from those expected by chance, improving the understanding of how specific geographic areas can influence the chances of occurrence of certain events. The proposed method also favors identifying spatial patterns involving agglomerations, anomalies, and geographic subspaces , which corroborates the potential contribution of this work.
Container-based virtualization (CBV) is widely adopted in modern IT infrastructure, with technologies such as Docker and Kubernetes commonly used for application packaging and orchestration, respectively. Despite the growing body of research in this domain, no existing secondary study systematically maps CBV technologies across both IT domains and academic dimensions—comprising education, research, and outreach. This paper presents a Systematic Mapping Study (SMS) of 226 primary studies published between 2022 and 2024, identified through a hybrid strategy combining database searches across five digital libraries and forward/backward snowballing. Studies were classified using 11 IT domains and three academic dimensions, assessed through three quality indices (CVI, SCI, IRRQ), and organized into a dual-axis taxonomic structure. The results reveal a marked concentration around Docker (41.6%) and Kubernetes (29.6%), a dominance of IT Infrastructure as a research domain (75.66%), and an underrepresentation of education (11.06%) and outreach (5.88%) in the CBV literature. Six concrete research gaps are identified, including the need for alternative runtime evaluation, orchestration beyond Kubernetes, and empirical studies on CBV in educational contexts. The proposed taxonomy and identified gaps provide a structured foundation for researchers, educators, and practitioners navigating the evolving CBV landscape.
The global drone market has grown from US$4 billion in 2023 to a projected US$4.8 billion by 2029. Concomitant with this growth, integrating drones into Mobile Edge Computing (MEC) environments poses critical challenges in scalability, resource allocation, and latency-aware distributed processing, which directly affect overall system performance. Addressing these issues through physical prototyping is costly and complex, which motivates the use of analytical models that can predict system behavior under different conditions. In this work, we propose a performance evaluation approach based on Stochastic Petri Nets (SPNs) to model the admission, load balancing, and distributed task processing among drones acting as mobile edge nodes. The proposed model enables resource scaling and scalability analysis without physically implementing the architecture, reducing deployment risks and costs. Simulation results demonstrate that the model accurately captures key performance metrics, including Mean Response Time (MRT), throughput, and task drop rate, providing quantitative insights for designing efficient and resilient UAV-supported MEC systems.
This paper introduces PRISEC III, a dynamic cryptographic framework designed to balance security and performance in heterogeneous IoT environments. Unlike static, one-size-fits-all approaches, PRISEC III employs a rolebased multi-level model that adapts cryptographic strategies according to data sensitivity, device constraints, and network conditions. The framework integrates lightweight symmetric primitives for efficiency, robust asymmetric methods for secure key exchange, and hybrid schemes that combine multiple layers of protection. To enhance long-term resilience, PRISEC III also incorporates post-quantum options, ensuring preparedness against emerging cryptographic threats. Evaluation results demonstrate that the approach achieves scalable security while maintaining feasible computational costs, making it suitable for large-scale and resource-constrained IoT deployments.
Recent advances in drone logistics have led to research on integrated planning models. These models combine infrastructure design, energy constraints, and temporal scheduling. This paper introduces the Drone Delivery Location and Scheduling Problem with Recharging Stations (DDLSP-R). DDLSP-R advances previous facility-location models by explicitly incorporating time-slot scheduling and delivery due dates. In this problem, each customer must be served within a set time window. Deliveries completed after the due date incur a disutility penalty in the objective function. The proposed model jointly determines the locations of facilities and recharging stations, as well as the allocation and sequencing of drone deliveries. The goal is to minimize total energy consumption and lateness costs, while maximizing coverage and service feasibility. To enable large-scale instances, the approach uses a heuristic decomposition to address strategic location decisions and operational scheduling, each solved via mixed-integer programming within bounded computational time. Experimental results in realistic urban scenarios demonstrate that including recharging stations and temporal scheduling enhances delivery flexibility and expands coverage. These findings highlight the importance of integrating spatial and temporal decision-making for sustainable drone-based logistics in smart cities.
The popularity and development of Autonomous Electric Vehicles (AEVs) bring the need to ensure the safety and reliability of these systems. With the increasing adoption of such vehicles, the promise of reducing human errors and improving transportation efficiency is becoming increasingly attainable. However, failures in autonomous vehicle systems can lead to material damage and loss of human life. Minimizing these failures becomes critical to guarantee the safety of passengers and pedestrians. This work employs Reliability Block Diagram (RBD) and Stochastic Petri Net (SPN) models to evaluate the vehicle safety system of a Level 3 AEV. The proposed approach enables the analysis of availability and reliability metrics, identifying potential improvements in system components. The system’s reliability showed a predictable decline over time, allowing for the anticipation of maintenance and necessary enhancements. These results provide valuable insights for AEV developers regarding the optimization of vehicle safety and reliability.
Elastic Optical Networks (EONs) enable fine-grained spectrum allocation that adapts to heterogeneous Internet services. Yet, fragmentation and the lack of contiguous free slots remain persistent causes of connection blocking and SLA violations. This article extends our SBRC paper from 2024 by providing a comprehensive journal version of APACHE—an analytical provisioning algorithm that (i) distinguishes blocking due to fragmentation from blocking due to true scarcity, (ii) incorporates Class of Service (CoS) differentiation into the RSA, and (iii) evaluates the economic impact of blocking via revenue loss. We detail the algorithmic design, simulation methodology, and an expanded performance assessment on the USA and RNP topologies. Results show that APACHE reduces circuit blocking up to 80% (USA) and 56% (RNP) compared to CFCoSP and TSSCF baselines, while consistently minimizing revenue loss under mixed-demand traffic profiles. The findings highlight how fragmentation-awareness and service-class prioritization jointly improve the efficiency and business viability of EON provisioning for next-generation Internet services.
Graph-based learning has become a cornerstone for analyzing complex relationships in domains such as social networks. However, most existing solutions rely on Graph Neural Networks (GNNs), which often require high computational resources, careful parameter tuning, and extensive training time. In this paper, we propose a decoupled learning framework for node classification that replaces end-to-end deep graph architectures with latent structural embeddings and conventional machine-learning models. Our method generates low-dimensional node embeddings and then applies classifiers such as logistic regression, KNN, and random forests to predict node categories. By structurally decoupling the embedding and classification stages, our approach achieves competitive performance while drastically reducing memory complexity and training time. Experimental results on multiple social network datasets demonstrate that our method offers a scalable, interpretable alternative to deep GNN models.
The introduction of GPS-equipped IoT devices onboard of urban buses allows the collection of data to monitor the fleet and assess the quality of service of this public transportation modality. The present work analyzes the GPS data from Rio de Janeiro buses, which includes executed trajectories, bus routes and identifiers to estimate per-route performance metrics with the collected data. In particular, the time between two buses, the number of bunched buses, the time spent on the inner fraction of the route, and the entropy are investigated as indicators of quality of service. The results reveal the unpredictability degree of each analyzed route. Lastly, we investigate the correlations between different metrics as a means of discovering relationships between metric performances. Our analysis shows that some routes exhibit greater regularity than others when individual metrics are considered. In addition, we observe a correlation between bus interval and bus bunching, with longer intervals reducing the likelihood of bunching. Also, we observed a correlation between entropy and time spent in the inner fraction of the route. Those results provide useful information in fleet planning activities for public transport.
Smart cities increasingly rely on AI-driven solutions to improve citizen safety. In coastal regions such as Rio de Janeiro, Brazil, maritime risks present significant challenges, as exemplified by the collapse of the Tim Maia Bike lane. This paper proposes a method for monitoring the coastal infrastructure using a custom segmentation model based on YOLO, aiming to reduce the need for continuous human supervision. However, the external placement of cameras introduces lighting and weather-related challenges, which complicate accurate segmentation. To address these issues, we investigate how domain-specific data augmentation techniques affect model performance under adverse visual conditions. As a case study, we apply this method to develop a system for the Tim Maia Bike lane, with the improved models achieving 97.6% mAP50-95 throughout the day. Furthermore, we analyze the correlation between our system’s outputs and environmental measurements obtained from an ocean buoy. Our findings highlight the potential for integrating AI-based monitoring into broader urban risk management frameworks to provide real-time protection for coastal infrastructure.
Urban analysis often relies on census or census-like survey data, but the redesign of census tract layers often breaks historical comparability, thereby hampering longitudinal studies at the local scale. In this article, we propose an automated methodology for making census tracts comparable through the construction of a graph, using spatial join operations and non-spatial id comparisons. We adopt an aggregative heuristic to minimize the impact of the matching process on analytical possibilities. To validate the proposed methodology, we deploy it to match the most recent Brazilian censuses for the S & atilde;o Paulo, Vit & oacute;ria, and Recife metropolitan regions, as well as for the Brazilian Federal District. We also test our implementation for New York City and Buenos Aires. We then evaluate the results against available gold standard data and alternative metrics, concluding with a discussion of potential improvements. Our implementation produces comparability files efficiently and has the potential to enable studies that would otherwise be impractical.
Traditional radio access networks (RANs) are often characterized by rigid, proprietary architectures, high capital (CAPEX) and operational (OPEX) expenditures, and vendor lock-in. These limitations hinder cost-effective deployments in remote and underserved regions. In this paper, we evaluate the potential of Open RAN, a flexible and disaggregated networking paradigm leveraging virtualization and multi-vendor interoperability, to bridge the healthcare gap in Brazil through remote eHealth applications. Focusing on the mobile specialist practice (MSP) use case, we introduce a total cost of ownership model that integrates both CAPEX and OPEX across urban, suburban, and rural deployment scenarios, thereby addressing a critical gap in current network economic models. Additionally, we explore a detailed taxonomy of MSP sub-use cases that delineates the technical requirements for high-definition video streaming, haptic feedback, and remote diagnostics. We conducted laboratory experiments on a 5G Open RAN testbed using real ultrasound equipment in collaboration with Samsung, Benefic & ecirc;ncia Portuguesa (one of the largest private healthcare hubs in Latin America), and InovaHC (the innovation center of the region's largest hospital). These experiments demonstrated that our approach meets stringent performance indicators such as low latency, robust throughput, and reliable packet delivery under diverse backhaul conditions. The results further indicate that Open RAN can substantially reduce deployment costs while ensuring high-performance remote medical services. These findings provide valuable insights for advancing field trials, refining economic models, and guiding future policy initiatives.
Blockchain technology has emerged as a disruptive innovation, facilitating decentralized and secure transactions across various domains. However, performance and scalability are critical challenges, especially for enterprise applications. This study evaluates the performance of two Ethereum-based blockchain clients: Geth (Go Ethereum) and Hyperledger Besu, under different network conditions, including the use of a Load Balancer. We conducted an extensive benchmarking analysis using the Hyperledger Caliper tool to measure key performance metrics such as Additionally, we performed statistical analyses, including Percentage Difference, Coefficient of Variation, and Correlation between Metrics, to provide deeper insights. Our results indicate that Geth consistently outperforms Besu in terms of TPS and latency under high workloads, while Besu demonstrates greater stability and scalability as the number of nodes increases. The impact of the Load Balancer varies across metrics, improving throughput in some cases but increasing latency in others. These findings offer valuable insights for organizations selecting blockchain platforms for enterprise solutions, highlighting scenarios in which each client performs best.
Network Function Virtualization (NFV) is changing the way networks are built and maintained by replacing traditional middleboxes with software that runs on commodity hardware. This paradigm shift not only increases flexibility but can also reduce the cost of building and maintaining the network. The ETSI NFV-MANO reference model is currently widely adopted as the de facto NFV standard, with a large number of compliant platforms currently available. In this work, we propose Vines (Vines Is an NFV-MANO Extensible Solution), a CloudStack-based platform that supports NFV technology and is compliant with the ETSI NFV specifications. Vines includes an NFV Orchestrator (NFVO) that enables the creation of complex network services composed of multiple individual functions. Management & orchestration are holistic, in the sense that they provide a comprehensive set of functionalities for heterogeneous network functions and services. Moreover, nearly all NFV developments have been done on the OpenStack cloud platform. Vines opens up the possibility of using NFV solutions on Apache CloudStack, one of the most widely used cloud platforms worldwide. Experimental results are presented showing that Vines has the same performance level as the widely used OpenStack/Tacker.
The Internet of Things (IoT) drives technological advances across various sectors by enabling seamless communication among smart devices. However, significant challenges remain regarding the integrity and reliability of the data stored by these devices. Traditional blockchain solutions, such as those based on Proof of Work (PoW), are generally unsuitable for IoT applications due to their high computational resource demands. Although approaches combining multiple consensus algorithms have emerged as alternatives to optimise performance and security, determining the best combination for each scenario remains an open problem. This paper proposes a strategy based on Genetic Algorithms (GAs) to adaptively select and combine consensus algorithms, thus improving blockchain efficiency in IoT environments. The approach was evaluated on a test blockchain, OmniBlock, implemented using a Directed Acyclic Graph (DAG) and designed specifically for evaluation purposes in IoT applications. OmniBlock supports multiple consensus algorithms, including Proof of Authority (PoA), Proof of Stake (PoS), Proof of Work (PoW), Practical Byzantine Fault Tolerance (PBFT), Raft, and others. The consensus algorithm combination is chosen based on performance attributes. All combinations of consensus algorithms evaluated in this work were suggested by GAs; the practical feasibility of each is analyzed empirically. Experimental results indicate that the evolutionary optimization-based strategy performs better across most of the combinations suggested by the GAs.
This paper proposes an approach to evaluate the availability and reliability of drone surveillance systems using complementary modeling techniques. Resilient system architecture with drone and battery redundancy is analyzed using two modeling strategies: (i) an analytical model based on Continuous-Time Markov Chains (CTMC), which yields closed-form availability equations, and (ii) a numerical model employing Stochastic Petri Nets (SPN) to handle more complex redundancy scenarios. Both models consider key factors such as battery charging/discharging times, drone failure and repair rates, and replacement operations. Sensitivity analyses highlight battery-related parameters as critical to system performance. Case studies show that optimizing component parameters can yield up to 97% availability, while redundancy alone can provide 91%. Combined strategies can achieve up to 99.89% availability. For long missions (30 hours), reliability analysis indicates that 15-20 redundant batteries and charging times below 36 minutes are needed to maintain over 80% reliability. For shorter missions, discharge times over 144 minutes are beneficial. This integrated modeling approach provides a robust framework for dependability assessment, guiding the design of resilient and cost-effective drone surveillance systems for mission-critical applications.
Protection mechanisms, also known as security mechanisms, in automotive systems are proactive components that continuously monitor vehicle signals to detect early signs of potential faults. For autonomous vehicles, it is essential that safety models, such as Responsibility-Sensitive Safety (RSS), which governs longitudinal and lateral safety, account for these mechanisms to enable timely and effective countermeasures against imminent actuation failures. A typical example is the proactive application of braking to increase longitudinal distance and mitigate the risk of losing braking capability. In this paper, we present a data-centric approach for modeling protection mechanisms using the SmartData framework, which facilitates the automatic derivation of safety properties for real-time formal verification via a Safety Enforcement Unit (SEU). We introduce extensions to RSS proper response strategies, enabling them to anticipate potential actuation constraints by leveraging shared internal states of protection mechanisms and a predictive time-to-trigger metric. We formally demonstrate that our approach preserves compliance with the original RSS safety guarantees by extending its inductive proof structure. Furthermore, we validate the feasibility of our solution through empirical evaluation, showing that the embedded formal verification can automatically extract properties from publish-subscribe message systems and operate at runtime with minimal overhead (less than 1% of platform processing capacity). Finally, we integrate our approach with RSS and a representative protection mechanism within the CARLA simulator to showcase its effectiveness in a realistic autonomous driving environment.
Open source dependencies are the leading source of vulnerabilities in applications and are often exploited in software supply chain attacks. Efforts to assess vulnerabilities are employed during DevSecOps pipelines in order to keep a system compliant with security regimes. However, current strategies for continuous compliance are limited to preventing issues before deployment, and thus do not address changes in dynamic aspects such as newfound vulnerabilities, let alone how to respond to such incidents. In this work, we leverage zero-trust to enable continuous, post-deployment vulnerability compliance assessment, isolating workloads that fail to meet a minimum security posture. This approach balances exploitation prevention with application availability - a fundamental trade-off for critical use cases. The solution is built on top of SPIRE, a robust open-source identity provider based on workload attestation, and implements a custom plugin that responds to compliance violations driven by dynamic aspects exposed by OWASP's Dependency Track, an open-source tool for monitoring software components and their dependencies for vulnerabilities. To enhance flexibility in the security-availability trade-off, we introduce a grace period mechanism, enabling organizations to defer enforcement of newly identified vulnerabilities based on workload criticality, thus supporting availability for non-critical workloads without compromising long-term security. Finally, we evaluate the performance impact of this approach on a SPIRE environment, showing that the added resource usage reliably remains within the recommended 16 GiB of RAM and 4 vCPUs to run Dependency Track in production. We also show that the plugin adds less than 6 seconds of latency to the attestation process, which is insignificant given its default frequency of twice per hour. Moreover, the results confirm that the approach successfully prevents vulnerability exploitation by prioritizing security, while enabling controlled flexibility in less critical contexts.
