Computational Resources Research Articles

INOVAÇÕES TECNOLÓGICAS NA FORMAÇÃO DE PROFESSORES: DESAFIOS E OPORTUNIDADES

Teacher training to work with innovative practices is essential for educational inclusion to be successful. Technological advances bring a new methodological perspective, requiring teachers to reflect on their practice, methodology and the competent use of computer resources. It is a challenge for teachers to base their practice on the new educational paradigms proposed for the 21st century, understanding that continuing education must be part of their study routine. Continuing education is an urgent need in the daily lives of teachers who are aware of the need to be active and participatory in this new era. Technology is already part of students' daily lives, whether at school, at home or in the job market. Therefore, through bibliographic research, it was discovered that courses that contribute to streamlining teachers' practice in the teaching process will make significant differences, bringing contributions to all disciplines. Among the main results, it is noted that teachers must change their attitude, but there is a need to create public policies for continuing education in pedagogical coordination. The state has the duty to provide the conditions for teachers to improve their pedagogical practices.

Towards Scalable and Secure Blockchain in Internet of Things: A Preference-Driven Committee Member Auction Consensus Approach

Blockchain technology is acclaimed for eliminating the need for a central authority while ensuring stability, security, and immutability. However, its integration into Internet of Things (IoT) environments is hampered by the limited computational resources of IoT devices. Consensus algorithms, vital for blockchain safety and efficiency, often require substantial computational power and face challenges related to security, scalability, and resource demands. To address these critical issues, we propose a novel model that significantly enhances the security and performance of blockchain in IoT environments. Our model introduces three key innovations: (1) a bidirectional-linked blockchain system that strengthens security against long-range attacks by exploiting dual reference points for block validation; (2) the integration of user preferences into the Committee Member Auction (CMA) consensus algorithm, optimizing miner selection to balance resource efficiency with security; and (3) a comprehensive performance and frequency analysis that demonstrates the system’s resilience against double-spend, long-range, and eclipse attacks. The proposed model not only reduces block validation delays but also enhances overall system performance, as evidenced by simulations comparing its effectiveness with existing CMA algorithms. These advancements have the potential to significantly impact the deployment of blockchain in resource-constrained IoT environments, offering a more secure and efficient solution.

A Novel Approach for Solving the N-Queen Problem Using a Non-Sequential Conflict Resolution Algorithm

The N-Queens problem is a fundamental challenge in combinatorial optimization, commonly used as a benchmark for assessing the efficiency of algorithms. Traditional algorithms, such as Backtracking with Forward Checking (BFC), constraint satisfaction problem (CSP) techniques, Lookahead algorithms, and heuristic-based methods, often face challenges with exponential time complexity, making them less practical for large-scale instances. This paper introduces a novel algorithm, non-sequential conflict resolution (NSCR), which improves performance over traditional algorithms through dynamic conflict resolution. The NSCR algorithm iteratively resolves conflicts among queens by adjusting their positions, aiming to optimize both time complexity and memory usage. While NSCR also operates within exponential time bounds, it demonstrates improved scalability and efficiency compared to traditional methods. A significant strength of the NSCR algorithm lies in its space complexity, which is O(n), and a time complexity that, while typically lower than traditional methods, can reach O(n3) in the worst-case scenario. This linear space complexity is highly advantageous, particularly when dealing with large problem sizes, as it ensures efficient use of memory resources. Comparative analysis with the aforementioned algorithms shows that NSCR offers superior resource management, using up to 60% less memory and reducing runtime by approximately 50%, making it an efficient option for large-scale instances of the N-Queens problem. The algorithm’s performance, evaluated on problem sizes ranging from 8 to 1000 queens, highlights its ability to manage computational resources effectively, despite the inherent challenges of exponential time complexity.

Resource-optimized cnns for real-time rice disease detection with ARM cortex-M microprocessors

This study explores the application of Artificial Intelligence (AI), specifically Convolutional Neural Networks (CNNs), for detecting rice plant diseases using ARM Cortex-M microprocessors. Given the significant role of rice as a staple food, particularly in Malaysia where the rice self-sufficiency ratio dropped from 65.2% in 2021 to 62.6% in 2022, there is a pressing need for advanced disease detection methods to enhance agricultural productivity and sustainability. The research utilizes two extensive datasets for model training and validation: the first dataset includes 5932 images across four rice disease classes, and the second comprises 10,407 images across ten classes. These datasets facilitate comprehensive disease detection analysis, leveraging MobileNetV2 and FD-MobileNet models optimized for the ARM Cortex-M4 microprocessor. The performance of these models is rigorously evaluated in terms of accuracy and computational efficiency. MobileNetV2, for instance, demonstrates a high accuracy rate of 97.5%, significantly outperforming FD-MobileNet, especially in detecting complex disease patterns such as tungro with a 93% accuracy rate. Despite FD-MobileNet’s lower resource consumption, its accuracy is limited to 90% across varied testing conditions. Resource optimization strategies highlight that even slight adjustments, such as a 0.5% reduction in RAM usage and a 1.14% decrease in flash memory, can result in a notable 9% increase in validation accuracy. This underscores the critical balance between computational resource management and model performance, particularly in resource-constrained settings like those provided by microcontrollers. In summary, the deployment of CNNs on microcontrollers presents a viable solution for real-time, on-site plant disease detection, demonstrating potential improvements in detection accuracy and operational efficiency. This study advances the field of smart agriculture by integrating cutting-edge AI with practical agricultural needs, aiming to address the challenges of food security in vulnerable regions.

A Review of Hypergraph Neural Networks

In recent years, Graph Neural Networks (GNNs) have seen notable success in fields such as recommendation systems and natural language processing, largely due to the availability of vast amounts of data and powerful computational resources. GNNs are primarily designed to work with graph data that involve pairwise relationships. However, in many real-world networks, the relationships between entities are complex and go beyond simple pairwise connections, as seen in scientific collaboration networks, protein networks, and similar domains. If these complex relationships are directly represented as pairwise relationships using graph structures, it can lead to information loss. A hypergraph, as a special kind of graph-structured data, can represent higher-order relationships that cannot be fully captured by graphs, thereby addressing the limitations of graphs. In light of this, researchers have begun to focus on how to design neural networks on hypergraphs, leading to the proposal of hypergraph neural network (HGNN) models for downstream tasks. Therefore, this paper reviews the existing hypergraph neural network models. The review is conducted from two perspectives: spectral analysis methods and neural network methods on hypergraphs, discussing both unfolded and non-unfolded methods, and further subdividing them based on their algorithm characteristics and application scenarios. Subsequently, the design concepts of various algorithms are analyzed and compared, and the advantages and disadvantages of each type of algorithm are summarized based on experimental results. Finally, potential future research directions in hypergraph learning are discussed.

Neural-network approach to running high-precision atomic computations

Modern applications of atomic physics, including the determination of frequency standards and the analysis of astrophysical spectra, require prediction of atomic properties with exquisite accuracy. For complex atomic systems, high-precision calculations are a major challenge due to the exponential scaling of the involved electronic configuration sets. This exacerbates the problem of required computational resources for these computations and makes indispensable the development of approaches to select the most important configurations out of otherwise intractably huge sets. We have developed a neural-network (NN) tool for running high-precision atomic configuration interaction (CI) computations with iterative selection of the most important configurations. Integrated with the established atomic codes, our approach results in computations with significantly reduced computational requirements in comparison with those without NN support. We showcase a number of NN-supported computations for the energy levels of Fe16+ and Ni12+ and demonstrate that our approach can be reliably used and automated for solving specific computational problems for a wide variety of systems. Published by the American Physical Society 2024

ShiftTransUNet: An Efficient Deep Learning Model for Medical Image Segmentation Using ShiftViT Framework

Deep learning has significantly advanced the field of medical image segmentation. However, the complexity of network structures often leads to high computational demands, limiting their practical efficiency. To enhance the efficiency of image segmentation, this paper introduces an innovative, concise, and lightweight deep learning network. First, to reduce model complexity, we replaced the attention mechanism in the traditional vision transformer (ViT) structure with a shift operation, creating the ShiftViT architecture. This substitution significantly decreased computation and the number of parameters while preserving model performance. Second, to retain and enhance fine-grained features and facilitate more precise information transfer across different layers, we employed a full-scale progressive skip connection strategy. This approach effectively integrates multi-scale feature information, further enhancing model performance. Additionally, to further reduce network complexity, inspired by the independence of probabilities, we opted for depth-wise separable convolution over traditional convolution. This enhances the relative independence between layers. Together, these modifications achieved superior segmentation results on both the Synapse and Automated Cardiac Diagnostic Challenge (ACDC) datasets compared to mainstream models, with substantial advantages in terms of computational efficiency and parameter count. The proposed approach represents an effective solution for medical image applications with limited computational resources and holds great promise for clinical practice.

Physics informed data-driven near-wall modelling for lattice Boltzmann simulation of high Reynolds number turbulent flows

Data-driven approaches offer novel opportunities for improving the performance of turbulent flow simulations, which are critical to wide-ranging applications from wind farms and aerodynamic designs to weather and climate forecasting. However, current methods for these simulations often require large amounts of data and computational resources. While data-driven methods have been extensively applied to the continuum Navier-Stokes equations, limited work has been done to integrate these methods with the highly scalable lattice Boltzmann method. Here, we present a physics-informed neural network framework for improving lattice Boltzmann-based simulations of near-wall turbulent flow. Using a small amount of data and integrating physical constraints, our model accurately predicts flow behaviour at a wide range of friction Reynolds numbers up to 1.0 × 106. In contradistinction with other models that use direct numerical simulation datasets, this approach reduces data requirements by three orders of magnitude and allows for sparse grid configurations. Our work broadens the scope of lattice Boltzmann applications, enabling efficient large-scale simulations of turbulent flow in diverse contexts.

Comparison of deep and conventional machine learning models for prediction of one supply chain management distribution cost

Strategic supply chain management (SCM) is essential for organizations striving to optimize performance and attain their goals. Prediction of supply chain management distribution cost (SCMDC) is one branch of SCM and it’s essential for organizations striving to optimize performance and attain their goals. For this purpose, four machine learning algorithms, including random forest (RF), support vector machine (SVM), multilayer perceptron (MLP) and decision tree (DT), along with deep learning using convolutional neural network (CNN), was used to predict and analyze SCMDC. A comprehensive dataset consisting of 180,519 open-source data points was used for analyze and make the structure of each algorithm. Evaluation based on Root Mean Square Error (RMSE) and Correlation coefficient (R2) show the CNN model has high accuracy in SCMDC prediction than other models. The CNN algorithm demonstrated exceptional accuracy on the test dataset, with an RMSE of RMSE of 0.528 and an R2 value of 0.953. Notable advantages of CNNs include automatic learning of hierarchical features, proficiency in capturing spatial and temporal patterns, computational efficiency, robustness to data variations, minimal preprocessing requirements, end-to-end training capability, scalability, and widespread adoption supported by extensive research. These attributes position the CNN algorithm as the preferred choice for precise and reliable SCMDC predictions, especially in scenarios requiring rapid responses and limited computational resources.

Accurate optimal quantum error correction thresholds from coherent information

Quantum error correcting (QEC) codes protect quantum information from decoherence as long as error rates fall below critical error thresholds. In general, obtaining thresholds implies simulating the QEC procedure using, in general, suboptimal decoding strategies. In a few cases and for sufficiently simple noise models, optimal decoding of QEC codes can be framed as a phase transition in disordered classical spin models. In both situations, accurate estimation of thresholds demands intensive computational resources. Here we use the coherent information of the mixed state of noisy QEC codes to accurately estimate the associated optimal QEC thresholds already from small-distance codes at moderate computational cost. We show the effectiveness and versatility of our method by applying it first to the topological surface and color code under bit-flip and depolarizing noise. We then extend the coherent information based methodology to phenomenological and quantum circuit level noise settings. For all examples considered, we obtain highly accurate estimates of optimal error thresholds from small, low-distance instances of the codes, in close agreement with threshold values reported in the literature. Our findings establish the coherent information as a reliable competitive tool for the calculation of optimal thresholds of state-of-the-art QEC codes under realistic noise models. Published by the American Physical Society 2024

AMVFNet: Attentive Multi-View Fusion Network for 3D Object Detection

Pillar-based method is significant in the field of LiDAR-based 3D object detection which could directly make use of efficient 2D backbone and save computational resources during reference. Existing methods usually sequentially project the original point clouds into the cylindrical view or the bird-eye view for feature extraction. However, the former suffers from obscured problems and the scales of instances vary greatly with distance, while the latter leads to considerable confusion problems due to the loss of semantic information caused by the sparsity of the projected point cloud. In this paper, we present a novel and efficient two-stage point-pillar hybrid architecture named Attentive Multi-View Fusion Network (AMVFNet), in which we abstract features from all cylindrical view, bird-eye view, and raw point clouds. Rather than designing more complex modules to solve the problems inherent in the single-view approach, our multi-view fusion architecture effectively combines the strengths of multiple perspectives to improve performance at a more fundamental level. Besides, to compensate for quantization distortion caused by projection operations, we propose attentive feature enhancement layers to further improve the capability of contextual information capturing. Extensive experiments on the KITTI detection benchmark illustrate that our proposed AMVFNet achieves competitive performance compared with other SOTA 3D object detectors.

Compact Ultra-Wideband MIMO Antenna for 5G Communication Systems

Ultra-wideband (UWB) technology is highly respected in modern communication systems because it can enable high-speed, high-data-rate communication over short distances. It is still difficult to achieve optimal multiple-input multiple-output (MIMO) performance in real-world multipath settings, even with breakthroughs in traditional antenna designs. UWB-MIMO antenna solutions have been created and simulated in order to address this problem and fulfill present demands. Nevertheless, maintaining high antenna efficiency and the requirement for antenna designs tailored to a particular bandwidth are two of the difficulties UWB systems must overcome. Recently, a variety of optimization strategies have been employed to obtain high gain and UWB performance by fine-tuning antenna characteristics. The major obstacles are resolved by the optimization process, including the time and computational resource requirements. The difficulty is further compounded by the intricacy of accurately modifying antenna settings to maximize gain. A multi-objective advancement for the design and improvement of the reduced UWB-MIMO test radio wire for 5G was offered as a solution to these problems. In order to attain the UWB feature, we first aim to use a single elliptical-square form antenna that operates from 3.4 to 70 GHz with isolation larger than 29.54 dB. Furthermore, we optimize the impedance bandwidth by tuning the parameters of the structural dimensions using the improved proportional topology optimization (IPTO) technique. Additionally, we present the orthogonal correction using the binarized convolutional neural network (BCNN).

Structural topology optimization based on deep learning

In the mechanical design of structures, traditional topology optimization methods involve numerous finite element iterative analyses, leading to a significant expenditure of computational resources. Therefore, the improved multi-scale gradient generative adversarial networks topology optimization technique is proposed. The topology optimization condition parameters are compressed into a low-dimensional latent space feature representation using the encoder, allowing the model to better extract features from these parameters. To speed up model training, the generator and discriminator networks use lightweight residual convolutional blocks. The hybrid attention mechanism extracts prominent region features from the topology optimization structure map. The model training process is guided by a multi-dimensional fusion loss function to enhance the quality of generated model samples. Finally, transferring the parameters of the low-resolution topology optimization model to the high-resolution model enables complete training on a limited amount of high-resolution topology optimization datasets. The experimental data on the low- and high-resolution topology optimization datasets demonstrate that, when compared to alternative methods, this method produces better-quality topology optimization structure maps. Additionally, it can generate high-resolution topology optimization structure maps in minimal time, enabling real-time topology optimization.

ANEL: a novel efficient and lightweight authentication scheme for enhancing security in Vehicular Ad Hoc Networks (VANETs) using elliptic curve cryptography

Vehicular Ad-hoc Networks (VANETs) offer immense potential for revolutionizing modern transportation systems through real-time communication and interaction between vehicles and roadside infrastructure. These networks can significantly enhance traffic management, improve safety, and support various vehicular services. However, ensuring secure communication in VANETs remains a critical challenge due to the highly sensitive nature of the data exchanged, such as location and personal information. Current authentication schemes, including the ANEL scheme, exhibit vulnerabilities like data interception and tracking risks. In this paper, we propose a novel, efficient, and lightweight authentication scheme that leverages elliptic curve cryptography (ECC) and message mapping techniques to address these security issues. The proposed solution strengthens encryption, enhances privacy protection, and mitigates risks such as unauthorized access and man-in-the-middle attacks. Furthermore, the scheme offers scalability and distinguishes between regular and priority users, making it adaptable to various VANET applications. This approach significantly improves the overall security and privacy framework of VANET systems while ensuring the efficient use of computational resources in vehicular environments.

On the Modelling of Asymptotic Wavefronts in Long Ducts with Chambers

A novel method of determining the possible shapes of pressure wavefronts in ducts after they have travelled sufficient distances to evolve to asymptotic states is introduced. Although it is possible in principle to achieve the same outcome by simulating complete flow histories from the time of the creation of the wavefronts, this can be impracticable. It is especially unsuitable to use such methods when extremely small grid lengths are needed to represent the final outcome adequately. The new method does not simulate the propagation phase at all. Instead, it explores what final end states are possible, but gives no information about the initiating disturbance or the wavefront evolution towards the assessed asymptotic state. Accordingly, the two methods do not overlap, but instead are complementary to each other. A typical case in which the new capability has high potential is described and used to illustrate the purpose and use of the methodology. However, the primary focus is on the presentation and assessment of the method, not on any particular phenomenon. It is shown that the required computational resources are far smaller than those needed for conventional unsteady flow simulations of propagating wavefronts. The potential numerical limitations of the method are highlighted and, with one exception, are shown to be either of no consequence or easily reduced to acceptable levels. Special attention is paid to the one exception because it cannot be proven to be unimportant and, indeed, it would be unsafe to use it in general analyses of wave propagation. However, strong evidence is presented of its acceptability for the study of asymptotic wavefronts.

Dynamic Response Prediction Model for Jack-Up Platform Pile Legs Based on Random Forest Algorithm

Jack-up offshore platforms are widely used in many fields, and it is of great importance to quickly and accurately predict the dynamic response of platform pile leg structures in real time. The current analytical techniques are founded upon numerical modelling of the platform structure. Although these methods can be used to accurately analyze the dynamic response of the platform, they require a large quantity of computational resources and cannot meet the requirements of real-time prediction. A predictive model for the dynamic response of the pile leg of a jack-up platform based on the random forest algorithm is proposed. Firstly, a pile leg dynamic response database is established based on high-fidelity numerical model simulation calculations. The data are subjected to cleaning and dimensional reduction in order to facilitate the training of the random forest model. Cross-checking and Bayesian optimization algorithms are used for the selection of random forest parameters. The results show that the prediction model is capable of outputting response results for new environmental load inputs within a few milliseconds, and the prediction results remain highly accurate and perform well at extreme values.

An Efficient Detection Mechanism of Network Intrusions in IoT Environments Using Autoencoder and Data Partitioning

In recent years, with the development of the Internet of Things and distributed computing, the “server-edge device” architecture has been widely deployed. This study focuses on leveraging autoencoder technology to address the binary classification problem in network intrusion detection, aiming to develop a lightweight model suitable for edge devices. Traditional intrusion detection models face two main challenges when directly ported to edge devices: inadequate computational resources to support large-scale models and the need to improve the accuracy of simpler models. To tackle these issues, this research utilizes the Extreme Learning Machine for its efficient training speed and compact model size to implement autoencoders. Two improvements over the latest related work are proposed: First, to improve data purity and ultimately enhance detection performance, the data are partitioned into multiple regions based on the prediction results of these autoencoders. Second, autoencoder characteristics are leveraged to further investigate the data within each region. We used the public dataset NSL-KDD to test the behavior of the proposed mechanism. The experimental results show that when dealing with multi-class attacks, the model’s performance was significantly improved, and the accuracy and F1-Score were improved by 3.5% and 2.9%, respectively, maintaining its lightweight nature.

Retrieval Augmented Docking Using Hierarchical Navigable Small Worlds.

Make-on-demand chemical libraries have drastically increased the reach of molecular docking, with the enumerated ready-to-dock ZINC-22 library approaching 6.4 billion molecules (July 2024). While ever-growing libraries result in better-scoring molecules, the computational resources required to dock all of ZINC-22 make this endeavor infeasible for most. Here, we organize and traverse chemical space with hierarchical navigable small-world graphs, a method we term retrieval augmented docking (RAD). RAD recovers most virtual actives, despite docking only a fraction of the library. Furthermore, RAD is protein-agnostic, supporting additional docking campaigns without additional computational overhead. In depth, we assess RAD on published large-scale docking campaigns against D4 and AmpC spanning 99.5 million and 138 million molecules, respectively. RAD recovers 95% of DOCK virtual actives for both targets after evaluating only 10% of the libraries. In breadth, RAD shows widespread applicability against 43 DUDE-Z proteins, evaluating 50.3 million associations. On average, RAD recovers 87% of virtual actives while docking 10% of the library without sacrificing chemical diversity.

SODD-YOLOv8: an insulator defect detection algorithm based on feature enhancement and variable row convolution

Abstract To address the detection challenges in defective images of insulators in transmission lines, including tiny object size, significant scale variations, a wide variety of defects, and complex background interference. In this study, an improved insulator defect detection algorithm is proposed, based on the YOLOv8s framework and combining feature enhancement and deformable convolution techniques. Firstly, to address the image feature distortion problem caused by aerial photography, a deformable convolutional feature extraction module (DCFEM) is introduced, which is designed to enhance the model’s ability to adapt to the local geometric deformation, so as to effectively recover the distorted feature information in the image. Moreover, to enhance the detection ability of the model for small objects, a small object feature enhancement module is designed, which adopts an efficient multi-scale attention mechanism, and aims to enhance the feature extraction ability of small objects, improve the sensitivity to small-size defects, and improve the detection accuracy. Eventually, to optimize the computational efficiency of the model, the average pooling-sparse convolution-batch normalization (BN) module is proposed. This module combines average pooling, sparse convolution and BN techniques to achieve a lightweight model while maintaining a high level of feature extraction capability. Experimental results on the China power line insulator dataset show that the improved model achieves a 4.3 percentage point improvement in the mAP metric compared to YOLOv8s, and the number of parameters in the model is reduced by 10%. The proposed scheme not only improves the accuracy and efficiency of defect detection, but also reduces the demand for computational resources, thus providing a more reliable and efficient solution for insulator defect detection in practical applications.

Hybrid covariance super-resolution data assimilation

AbstractThe super-resolution data assimilation (SRDA) enhances a low-resolution (LR) model with a Neural Network (NN) that has learned the differences between high and low-resolution models offline and performs data assimilation in high-resolution (HR). The method enhances the accuracy of the EnKF-LR system for a minor computational overhead. However, performance quickly saturates when the ensemble size is increased due to the error introduced by the NN. We therefore combine the SRDA with the mixed-resolution data assimilation method (MRDA) into a method called “Hybrid covariance super-resolution data assimilation” (Hybrid SRDA). The forecast step runs an ensemble at two resolutions (high and low). The assimilation is done in the HR space by performing super-resolution on the LR members with the NN. The assimilation uses the hybrid covariance that combines the emulated and dynamical HR members. The scheme is extensively tested with a quasi-geostrophic model in twin experiments, with the LR grid being twice coarser than the HR. The Hybrid SRDA outperforms the SRDA, the MRDA, and the EnKF-HR at a given computational cost. The benefit is the largest compared to the EnKF-HR for small ensembles. However, even with larger computational resources, using a mix of high and low-resolution members is worth it. Besides, the Hybrid SRDA, the EnKF-HR, and the SRDA, unlike the MRDA, prevent the smoothing of dynamical structures of the background error covariance matrix. The Hybrid SRDA method is also attractive because it is customizable to available resources.