Search

The effective monitoring of eutrophication in inland water bodies is crucial for environmental management and pollution prevention. This study conducts a comprehensive analysis of in situ hyperspectral reflectance data (400–900 nm) and the trophic state index (TSI) obtained from 365 points across ten lakes and reservoirs in Northern Vietnam to propose a trophic classification based on water reflectance spectra features and a TSI estimation model for diagnosis and assessment of lake trophic status. By analyzing the quantity of reflectance peaks and their heights, our study identifies three distinct water reflectance spectra classes corresponding to three trophic levels: mesotrophic to lightly eutrophic, highly eutrophic, and hypertrophic. This classification enables the quick identification of trophic levels directly at the in situ radiometric measurement sites. Our study demonstrates that a logarithmic function of the band ratio, ${{\mathbf{R}}_{\mathbf{rs}}}( {715} )/{{\mathbf{R}}_{\mathbf{rs}}}( {560} )$, is robust for estimating TSI (${{{\bm{R}}}^2}$ = 0.85 and 0.94; root-mean-square error = 5.0 and 3.7 in calibration and validation, respectively), particularly in algal-dominated waters. These findings represent a practical application of hyperspectral remote sensing for effective eutrophication management. They also highlight the potential use of multispectral optical imagery for monitoring eutrophication in tropical regions.

Fuzzing is a popular and effective software testing technique that automatically generates or modifies inputs to test the stability and vulnerabilities of a software system, which has been widely applied and improved by security researchers and experts. The goal of fuzzing is to uncover potential weaknesses in software by providing unexpected and invalid inputs to the target program to monitor its behavior and identify errors or unintended outcomes. Recently, researchers have also integrated promising machine learning algorithms, such as reinforcement learning, to enhance the fuzzing process. Reinforcement learning (RL) has been proven to be able to improve the effectiveness of fuzzing by selecting and prioritizing transformation actions with higher coverage, which reduces the required effort to uncover vulnerabilities. However, RL-based fuzzing models also encounter certain limitations, including an imbalance between exploitation and exploration. In this study, we propose a coverage-guided RL-based fuzzing model that enhances grey-box fuzzing, in which we leverage deep Q-learning to predict and select input variations to maximize code coverage and use code coverage as a reward signal. This model is complemented by simple input selection and scheduling algorithms that promote a more balanced approach to exploiting and exploring software. Furthermore, we introduce a multi-level input mutation model combined with RL to create a sequence of actions for comprehensive input variation. The proposed model is compared to other fuzzing tools in testing various real-world programs, where the results indicate a notable enhancement in terms of code coverage, discovered paths, and execution speed of our solution.

Hand pose estimation is vital for various applications, including virtual reality (VR), augmented reality (AR), gesture recognition, human-computer interaction (HCI), and robotics. Achieving accurate and real-time hand pose estimation is challenging due to factors such as the high degree of articulation in the human hand and the variability in hand shapes and sizes. While multimodal data offers advantages, developing a fast and resource-efficient hand pose estimation system remains challenging. Current state-of-the-art methods often require powerful graphics processing units (GPUs) for high performance, limiting deployment on edge platforms with limited computational resources. There is a critical need for higher efficiency without compromising accuracy, especially in real-world applications like mobile devices and embedded systems. Additionally, real-time performance is essential for practical applications, where systems must respond immediately to user interactions. Unfortunately, most current methods struggle to achieve real-time speeds, even on powerful GPUs, let alone on resource-constrained devices. To address these challenges, we propose an efficient hand pose estimation system that leverages both red-green-blue (RGB) and depth (RGBD) data through a unified fusion strategy. Our method combines appearance and geometric data early in the processing pipeline, significantly reducing computational complexity while maintaining real-time performance on resource-constrained devices. Experimental results show that the proposed model runs at over 110 fps on GPU, and 30 fps on the edge platform of NVidia Jetson NX Xavier, which is 4 to 5 times faster than existing methods, while achieving competitive accuracy.

There has been an increasing interest in physical layer security (PLS), which, compared with conventional cryptography, offers a unique approach to guaranteeing information confidentiality against eavesdroppers. In this paper, we study a joint design of adaptive M-ary pulse amplitude modulation (PAM) and precoding, which aims to optimize wiretap visible-light channels’ secrecy capacity and bit error rate (BER) performances. The proposed design is motivated by higher-order modulation, which results in better secrecy capacity at the expense of a higher BER. On the other hand, a proper precoding design, which can manipulate the received signal quality at the legitimate user and the eavesdropper, can also enhance secrecy performance and influence the BER. A reward function that considers the secrecy capacity and the BERs of the legitimate user’s (Bob) and the eavesdropper’s (Eve) channels is introduced and maximized. Due to the non-linearity and complexity of the reward function, it is challenging to solve the optical design using classical optimization techniques. Therefore, reinforcement learning-based designs using Q-learning and Deep Q-learning are proposed to maximize the reward function. Simulation results verify that compared with the baseline designs, the proposed joint designs achieve better reward values while maintaining the BER of Bob’s channel (Eve’s channel) well below (above) the pre-FEC (forward error correction) BER threshold.

Wearable devices have gained significant popularity for continuous Electrocardiogram (ECG) monitoring due to their compactness and convenience. Day-by-day and 24/7 monitoring without gaps is demanded to promptly detect unusual symptoms that are short-term warning signs of dangerous diseases. Cost-saving criteria are the highest priority to extend the use time of devices and maintain the system operation. However, not all measured signals are useful because the signal quality will be affected by many things such as motion artifacts and muscle noise. Demands on classifying and sending only the usable signal is a requirement, it can save the battery on wearable devices and the expense on the server side (i.e., cloud-based processing), such as storage and processing resources. Therefore, Signal Quality Indices (SQIs) have been developed and researched to determine signal quality to meet the above requirements. This study introduces a novel SQI approach to classify signals. The proposed method has three contributions: 1) exponentially Weighted Mean-Variance (EWMV), a lightweight equation, to identify peaks, followed by applying an adaptive threshold to define peaks that have the same shape as R-peaks; 2) outlier elimination process is proposed to enhance the accuracy; and 3) maximal Overlap Discrete Wavelet Transform (MODWT) is employed to categorize ECG signals into a new class, potentially containing signals relevant to pathological analysis. Experimental results demonstrate that our algorithm achieves the highest sensitivity for both noisy and noiseless data sets. Specifically, it achieved a sensitivity of 99.31% for clean signals and 97.69% for noisy signals. In the case of the PhysioNet challenge data set, while our sensitivity of 96.37% may not be the highest, our accuracy stands out at 95.10%, surpassing other methods recently reviewed. Additionally, our approach demonstrates the lowest trade-off between sensitivity and accuracy among the surveyed SQI techniques.

Groundwater spring plays a crucial role in human life, including water resource management and planning; therefore, developing accurate prediction models for groundwater spring potential mapping is essential. The objective of this research is to introduce and confirm a new modeling approach based on TensorFlow Deep Neural Networks (TF-DNN) and multisource geospatial data for spatial prediction of groundwater spring potential, with a case study in the tropical province in the central highland of Vietnam. For this task, the TF-DNN model structure with three hidden layers with 32 neurons each was established; therein, the Adaptive Moment Estimation (ADAM) algorithm was used as an optimizer, whereas the Rectified Linear Unit (ReLU) was used as the activation function, and the sigmoid was used as the transfer function. A geospatial database for the study area, consisting of 733 groundwater spring locations and 12 influencing factors, was prepared in ArcGIS Pro. Then, it was used to develop and verify the TF-DNN model. Decision Tree, Support Vector Machine, Logistic Regression, Random Forest, and Classification and Regression Trees were used as a benchmark for the model comparison. The results demonstrate that the proposed TF-DNN model (Accuracy = 80.5%, F-score = 0.797, and AUC = 0.864) achieves a high global prediction performance, outperforming the benchmark models. Thus, the TF-DNN represents a novel and effective tool for spatially predicting groundwater spring potential mapping. The groundwater spring potential map generated in this study has the potential to assist provincial authorities in formulating strategies concerning water management and socio-economic development.

This paper proposes a novel virtual vector space vector pulse width modulation (SVPWM) method called Average Zero common mode voltage (AZCMV) PWM to mitigate low-frequency CMV in four-level neutral point clamped (4L-NPC) inverters. Due to the absence of zero CMV vectors in the space vector diagram (SVD) of 4L inverter, the proposed SVPWM method introduces virtual average zero CMV vectors, which are the combination of active vectors, and have an average CMV of zero. The SVPWM scheme is simply implemented in the transformed two-level (2L) SVD by the aid of base vectors. Then, three types of 2L SVD are presented. Effective and hybrid 2L-SVPWM control strategy based on four nearest vectors SVPWM, symmetrical and asymmetrical SVPWM is proposed to optimize output harmonic and switching losses. With the proposed SVPWM method, the peak-to-peak value of CMV can be reduced to one ninth of the total DC link voltage, which is lower up to five times as compared to the sinusoidal PWM (SPWM). Furthermore, the average value of CMV is zero in a switching sequence. The effectiveness of the proposed method has been verified by simulation and experimental results.

In industry, for driving a permanent magnet synchronous motor (PMSM), it is more favorable to optimize control performances and reduce computational complexity and communication waste from a controller to actuators. For that reason, this paper employs an event-triggering mechanism to design a robust optimal control strategy for PMSM. Firstly, the PMSM model is presented as a strict-feedback nonlinear system with unknown internal dynamics, disturbances, and constrained inputs. Then, an event-triggered (ET) feedforward control strategy is introduced to convert the separated speed and current dynamics into an augmented system. Secondly, an ET-robust optimal feedback control strategy and an ET disturbance compensation strategy are designed using adaptive dynamic programming (ADP) and zero-sum game theory. All controller parameters are tuned online without identifying unknown dynamics or using a persistent excitation condition. It is shown that system stability and the exclusion of Zeno’s behavior are fulfilled. Finally, compared with the existing time-triggering control strategies in simulation and experiments with TMS320F28335 of Texas Instruments, the proposed strategy is more effective in reducing the burden of computation bandwidth and communication load.

Blockchain technology has brought a significant advancement in the development of academic certificate management systems by enhancing security, transparency, and decentralization. However, challenges such as certificate revocation, transaction costs, and latency still persist. This research proposes a novel mechanism combining smart contracts and Optimistic Rollups technique to address these issues. By leveraging the off-chain processing feature of Optimistic Rollups, the research has significantly reduced transaction latency and costs in certificate revocation. This integration not only optimizes performance but also maintains transparency and data integrity on the blockchain. Moreover, integrating machine learning for fraud detection not only reinforces the security of the certificate management system but also provides timely alerts before fraudulent transactions occur. The combination of blockchain to ensure decentralization and security, along with machine learning to detect and prevent fraud, creates a comprehensive and advanced certificate management system. The experimental outcomes validate the effectiveness of Optimistic Rollups in certificate revocation, showing a notable approximately 61.92% reduction in both transaction costs and latency. Moreover, the machine learning model displays impressive performance, achieving high accuracy in detecting fraudulent users, with an average F1-score of 99.42% and an AUC score nearing perfection. These results underscore the comprehensive and advanced nature of the certificate management system.

Digital Twin (DT) technology has recently emerged as a powerful tool with the potential to revolutionize wireless systems as it enables accurate simulations, better decision-making, and tangible operational improvements. Prior studies on DT within the context of next generation wireless technologies have primarily focused on identifying potential use cases, application scenarios, standardization challenges, and conceptual implementation steps. However, the existing research is limited in translating theoretical ideas into real-world applications. Our research, in this paper, contributes to the practical realization of DT technology in the context of 6G wireless networks, demonstrating its potential impact on network planning, performance, and user experience. In particular, we explore the construction, validation, and applications of DT utilizing an indoor over-the-air (OTA) 5G NR testbed powered by an in-house developed Next Generation Radio Access Network (NG-RAN) that is fully compliant with 3rd Generation Partnership Project (3GPP) and Open RAN standards. First, we explain the integration and implementation steps followed to integrate Qualcomm EdgewiseTM Suite and Service Management and Orchestration (SMO) tools into the NG-RAN architecture, that will eventually enable the applicability of DT for wireless network operations. We then describe our procedure to construct and validate a high-fidelity DT of our OTA testbed modeling both Radio Frequency (RF) environment and system components. We demonstrate two pre-deployment use cases by describing our extensive coverage estimation and network capacity planning tests in OTA. Lastly, we explore how DT enables practical machine learning solutions for post-deployment use cases and share our comprehensive OTA performance results, highlighting that our proposed mobility and positioning techniques outperform the classical approaches in terms of throughput, number of undesired handovers, and positioning accuracy.

Transformer-based models, such as Bidirectional Encoder Representations from Transformers (BERT), cannot process long sequences because their self-attention operation scales quadratically with the sequence length. To remedy this, we introduce the Look Near and Look Far BERT (LNLF-BERT) with a two-level self-attention mechanism at the sentence and document levels, which can handle document classifications with thousands of tokens. The self-attention mechanism of LNLF-BERT retains some of the benefits of full self-attention at each level while reducing the complexity of not using full self-attention on the whole document. Our theoretical analysis shows that the LNLF-BERT mechanism is an approximator of the full self-attention model. We pretrain the LNLF-BERT from scratch and fine-tune it on downstream tasks. The experiments were also conducted to demonstrate the feasibility of LNLF-BERT in long text processing. Moreover, LNLF-BERT effectively balances local and global attention, allowing for efficient document-level understanding. Compared to other long-sequence models like Longformer and BigBird, LNLF-BERT shows competitive performance in both accuracy and computational efficiency. The architecture is scalable to various downstream tasks, making it adaptable for different applications in natural language processing.

Most practical systems are influenced by disturbances and uncertainties within their models, increasing the need for robust control algorithms. Type-2 fuzzy theory and its variant, interval type-2 fuzzy, are well-regarded for their capability to effectively model and manage uncertainties. This paper presents a composite controller that integrates an interval type-2 Takagi-Sugeno fuzzy controller with a PID controller. The fuzzy controller ensures robust and stable system control, while the PID controller significantly enhances system performance. Simulation results demonstrate the superiority of the composite controller over a single fuzzy controller, particularly in the context of a cart-pole system with varying masses.

Today’s power system is transforming into an increasingly complex entity, consisting of numerous components, such as transmission lines, controllable loads, and especially different types of distributed renewable energy sources (DERs). With the growing integration of DERs into the grid, multiple operational challenges arise, including overvoltage, undervoltage, or increased energy losses. Resolving these issues demands the implementation of both advanced and effective control strategies. As the dynamic power system evolves by incorporating new technologies, these control strategies need to consider other different technical aspects, such as communication protocols or real-time considerations. Additionally, the rise of smart metering devices has transformed conventional power systems into cyber-physical power systems (CPPS), which can integrate the advanced control strategies into the cyber layer. Given the operating challenges and the integration of diverse technologies, it is proposed that a CPPS testbed platform constitutes an ideal solution for developing and validating technologies in future smart grids. For this purpose, this paper introduces a co-simulation framework for implementing a CPPS testbed, utilising the real-time simulator, Typhoon HIL, within a laboratory environment. Additionally, it presents a proposed optimal power flow (OPF) control strategy that emphasises two key objectives, minimisation of operating costs and power network loss. The investigation is illustrated by a modified version of IEEE 39-bus test system with the high integration of DERs. The findings indicate that adopting a CPPS testbed can be advantageous for implementing real-time research on monitoring and control in a wide area network.

Log anomaly detection is a crucial task in monitoring IT systems along with metrics and traces. An anomaly could be detected by either one of two types of logs: individual logs or log sequences. While an individual log indicates an independent system status, combining multiple logs describes the execution paths of systems. Once the patterns of log sequences deviate much from normal execution behaviors, that might indicate system anomalies. For log anomaly detection using log sequences, supervised learning methods are preferable due to their high performance. However, these methods require labeled data to train models. As systems evolve, the number of logs increases significantly, which makes labeling data labor-intensive and impractical. Therefore, other learning techniques, such as semi-supervised or unsupervised, are better alternatives for detection. In practice, detecting log anomalies is quite challenging because of several problems, such as unstable logs, new types of logs, and unexplored log semantics. To address these problems and enhance detection performance, we propose a lightweight semi-supervised multi-task learning method named MultiLog in this paper. The key components of the proposed method are pre-trained language model BERT, dimension reduction, attention mechanism from Transformer, and multi-task learning. Similar to previous studies, we conduct comprehensive experiments on three widely used datasets: HDFS, BGL, and Thunderbird. In terms of efficiency, our proposed model is 50 times smaller while the F1-Scores are maintained compared to the original model. In terms of effectiveness, the proposed model outperforms baseline methods and achieves performance comparable to supervised learning models.

In the large-scale deployment of federated learning (FL) systems, the heterogeneity of clients, such as mobile phones and Internet of Things (IoT) devices with different configurations, constitutes a significant problem regarding fairness, training performance, and accuracy. Such system heterogeneity leads to an inevitable trade-off between model complexity and data accessibility as a bottleneck. To avoid this situation and to achieve resource-adaptive FL, we introduce CrossHeteroFL to deal with heterogeneous clients equipped with different computational and communication capabilities. Our solution enables the training of heterogeneous local models with additional computational complexity and still generates a single global inference model. We demonstrate several CrossHeteroFL training scenarios and conduct extensive empirical evaluation, covering four levels of the computational complexity of three-model architectures on two datasets. The proposed mechanism provides the system with non-elementary access to a scattered fit among clients. However, the proposed method generalizes soft handover-based solutions by adjusting the model width according to clients’ capabilities and a tiered balance of data-source overviews to assess clients’ interests accurately. The evaluation results indicate our method solves the challenges in previous studies and produces greater top-1 accuracy and consistent performance under heterogeneous client conditions.

Tripolar fuzzy sets offer a robust framework that integrates the aspects of intuitionistic fuzzy sets and bipolar fuzzy sets. Complex decision-making problems often involve multiple attributes with specific properties that reflect both negative and positive aspects. In such cases, intuitionistic fuzzy sets (IFS) and bipolar fuzzy sets (BFS) may be insufficient to handle these complexities that makes tripolar fuzzy sets an essential tool for identifying optimal solutions. In our study, we introduce a Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) method using connection number of set pair analysis under the tripolar fuzzy environment. Initially, we express tripolar fuzzy number as a connection number, which is a pivotal concept in set pair analysis. We then extend the TOPSIS method based on the connection number of set pair analysis within the realm of tripolar fuzzy sets. By applying the proposed methodology, we provide the computational framework for the problem related to multi-criteria decision making. Moreover, we observe that our proposed technique also satisfies the existing test criterions. In this context, we conduct the comparative study of the newly established technique with the other ones existing in intuitionistic fuzzy and bipolar fuzzy environments. Finally, we show that our proposed technique is superior than the others.

Recent advancements in Artificial Intelligence (AI) have greatly impacted cybersecurity, particularly in detecting phishing websites. Traditional methods struggle to address evolving vulnerabilities, but research shows that Machine Learning (ML), Ensemble Learning (EL), and Deep Learning (DL) are effective in developing defenses. However, these methods face challenges with adversarial examples (AEs). The multimodal model (MM) is a promising solution, yet there is a significant lack of research using multimodal techniques specifically for phishing website detection (PWD) against adversarial websites. To tackle this challenge, this paper assesses 15 learning-based models, particularly multimodal ones, for phishing and adversarial detection, aiming to enhance their defense capabilities. Due to the scarcity of adversarial websites, training and testing models are limited. Therefore, this study proposes an innovative attack framework, AWG - Adversarial Website Generation that employs Generative Adversarial Networks (GAN) and transfer-based black box attacks to create AEs. This framework closely mirrors real-world attack scenarios, ensuring high effectiveness and realism. Finally, we present defense strategies with straightforward implementation and high effectiveness to enhance the resistance of models. The models underwent training and testing on a dataset collected from reputable sources such as OpenPhish, PhishTank, Phishing Database, and Alexa. This approach was chosen to ensure the dataset’s diversity and relevance to reflect real-world conditions. Experimental results highlight that the Generator’s effectiveness is demonstrated by a domain structure generation rate exceeding 90%. Moreover, AEs generated by this Generator effectively bypass most state-of-the-art ML, DL, and EL models with an evasion rate of up to 88%. Notably, the Support Vector Machine (SVM) model is the most vulnerable, with a detection rate of only 10.02%. On the other hand, the MM Shark-Eyes demonstrates outstanding resistance against AEs, with a detection rate of up to 99%. Upon applying our defense strategy, the resistance of models is significantly boosted, with all detection rates surpassing 90%. These findings underscore the robustness of our methods and pave the way for further exploration into advanced attack and defense strategies in the context of phishing and adversarial website detection.

Accurate estimation of dual walking trajectories remains a challenge in human gait tracking systems due to limitations in sensor precision and data integration methods. To address these issues, this paper presents a novel human gait tracking system that integrates a downward-looking waist-mounted red-green-blue-depth (RGB-D) camera with two inertial measurement units (IMUs) mounted on each foot. Our approach utilizes a fully convolutional network (FCN) for precise foot detection from RGB-D images. The positions of both feet are then computed using the detected foot and the camera’s rotation matrix relative to the floor plane. These position estimates are incorporated into a Kalman filter, with a quadratic optimization-based smoothing method applied to improve accuracy. Experimental results demonstrate a significant improvement in dual trajectory estimation, achieving a root mean square error (RMSE) of 3.3 cm in stride length estimation. This system enhances the accuracy and reliability of gait analysis, effectively addressing the limitations of existing methods.

Accidents resulting from fires caused by electrical devices are frequent occurrences, inflicting substantial damage to both human lives and infrastructure in the Republic of Korea. To ascertain whether these fires stem from external or internal infrastructure factors, investigators such as the police, The National Institute of Scientific Investigation, and the National Fire Research Institute conduct fire-causing inspections. However, obtaining conclusive results in this regard is an intricate process, exacerbated by the dearth of adequate digital forensics tools and related programs. Among electrical devices, multi-socket outlets also contribute to fire incidents. This study explores the feasibility of employing CNN-based deep learning object detection models for fire-causing inspection systems targeting multi-socket outlets. Specifically, we introduce a novel image dataset comprising 6009 images of post-fire multi-socket outlets remaining, categorized into two classes: “burnt-in” and “burnt-out.” This dataset is utilized for training various models, including the YOLO-series (v5, v6, and v8), Faster-RCNN, RetinaNet, and SSD. Results from our experiments show the feasibility of six CNN models in detecting the cause of fire in post-fire sockets. Particularly, YOLOv5s surpasses other models with an accuracy of 89.1% mAP@0.5, a model size of 14.4MB, and an inference time of 44.5ms (equivalent to 22 fps) on RTX 3050. Subsequently, the trained models are implemented in an operational application for trial testing during an executive period.

The most significant challenge the maximum power point tracking (MPPT) algorithms face is reaching the global MPPT under partial shading conditions (PSC). In this case, the power-voltage (P-V) and current-voltage (I-V) characteristic curves of the photovoltaic (PV) system produce many maximum power points, which is the reason that the algorithms often trapped in local maximum power point (LMPP). The Perturbation and Observation (P&O) algorithm is simple and effective for finding MPP on a mono-extreme graph. Under PSC with many extremes, it is necessary to limit the range of potential GMPP so that this algorithm can avoid the LMPP trap. This paper introduces a global MPPT (GMPPT) method based on the combination of modified P&O and simulates the behavior of the I-V curve under PSC. In particular, the starting parameters of the algorithm are adjusted in the potential GMPP region to exploit simplicity and efficiency to increase performance and convergence speed. The proposed method determines the area where the GMPP is potentially found more accurately, and this contribution is crucial for increasing the efficiency of the strings of PV arrays under partial shading conditions. By comparing results with other algorithms, such as Modified P&O, Adaptive Jaya (Ajaya), and Jaya, the proposed algorithm has verified its outstanding advantages in MPPT performance and search time under the same test conditions.

Medical image segmentation plays a crucial role in modern clinical practice, enabling accurate diagnosis and personalized treatment plans. Advancements in machine learning, particularly deep learning techniques, have significantly driven this progress. While Convolutional Neural Networks (CNNs) dominate the field, transformer-based models are emerging as powerful alternatives for computer vision tasks. However, most existing CNN-Transformer models underutilize the full potential of Transformers, often relegating them to assistant modules. To address this issue, we propose a novel and efficient UNet-Transformer (seUNet-Trans) model for medical image segmentation. The seUNet-Trans framework leverages a UNet architecture for feature extraction, generating rich representations from input images. These features are then passed through a bridge layer that connects the UNet to a transformer module. To improve efficiency, we employ a novel pixel-wise embedding method that eliminates the need for position embedding vectors. We utilize spatially reduced attention within the transformer to reduce computational complexity. By combining the strengths of UNet’s localization capabilities and the transformer’s ability to capture long-range dependencies, seUNet-Trans effectively captures both local and global information within medical images. This holistic understanding enables the model to achieve superior segmentation performance. The efficacy of our model is demonstrated through extensive experimentation on seven medical image segmentation datasets. The seUNet-Trans model outperforms several state-of-the-art segmentation models, achieving impressive mean Dice Coefficient (mDC) and mean Intersection over Union (mIoU) scores. On the CVC-ClinicDB dataset, it achieves scores of 0.945 and 0.895, respectively; on the GlaS dataset, it scores 0.899 and 0.823, respectively; on the ISIC 2018 dataset, it achieves 0.922 and 0.854, respectively; and on the Data Science Bowl dataset, it scores 0.928 and 0.867, respectively. The code is available on seUnet-Trans.

The increasing integration of renewable energy sources (RESs) into the power system reduces system inertia, which presents substantial challenges for maintaining frequency stability. Existing studies have mainly focused on using the load frequency control method (LFC) to stabilize area frequency and tie-line power by optimizing the parameters of supplementary control for synchronous generators (SGs). However, these works primarily used simplified first-order transfer functions for wind turbines (WT), high-voltage direct current (HVDC), and battery energy storage systems (BESS). They neglected dynamic models and, importantly, the impact of control parameters in supplementary frequency control strategies on the stability of the studied system. Therefore, this study proposes a novel load frequency control strategy that coordinates the operation of WT, HVDC, and BESS in conjunction with traditional SGs to participate in frequency regulation. The proposed strategy develops a comprehensive state-space model that incorporates accurate models of WT, HVDC, BESS, and SGs, along with their supplementary control strategies. Then, the developed state-space model was utilized to investigate the impact of control parameters in the supplementary frequency control strategies of BESS, HVDC, and WT on system stability. As a result, this work yields a specific range of coefficients for supplementary strategies to improve stability in the studied system. Lastly, a two-area interconnected benchmark system is adopted to validate the effectiveness of the proposed strategy. Based on the suggested parameters, the results demonstrate that the proposed strategy ensures frequency stability with a high penetration of RESs into the power grid. Furthermore, the results indicate that incorporating diverse sources, such as HVDC, WT, and BESS, to support SGs in frequency regulation reduces the impacts of low system inertia. This approach leads to an improved frequency response not only in areas where disturbances occur but also in other regions, provided that tie-line power interchanges are maintained in balance.

Monocular Depth Estimation (MDE) and Object Pose Estimation (OPE) are important tasks in visual scene understanding. Traditionally, these challenges have been addressed independently, with separate deep neural networks designed for each task. However, we contend that MDE, which provides information about object distances from the camera, and OPE, which focuses on determining precise object position and orientation, are inherently connected. Combining these tasks in a unified approach facilitates the integration of spatial context, offering a more comprehensive understanding of object distribution in three-dimensional space. Consequently, this work addresses both challenges simultaneously, treating them as a multi-task learning problem. Our proposed solution is encapsulated in a Unified Framework for Depth-Assisted Monocular Object Pose Estimation. Leveraging Red-Green-Blue (RGB) images as input, our framework estimates pose of multiple object instances alongside an instance-level depth map. During training, we utilize both depth and color images, but during inference, the model relies exclusively on color images. To enhance the depth-aware features crucial for robust object pose estimation, we introduce a depth estimation branch supervised by depth images. These features undergo further refinement through a cross-task attention module, contributing to the innovation of our method in significantly improving feature discriminability and robustness in object pose estimation. Through extensive experiments, our approach demonstrates competitive performance compared to state-of-the-art methods in object pose estimation. Moreover, our method operates in real-time, underscoring its efficiency and practical applicability in various scenarios. This unified framework not only advances the state of the art in monocular depth estimation and object pose estimation but also underscores the potential of multi-task learning for enhancing the understanding of complex visual scenes.

Quantum computing offers transformative processing capabilities, particularly in data analysis. However, its adoption in developing countries like Vietnam faces significant barriers, such as financial constraints, infrastructural limitations, and skill shortages. This study aims to determine and analyze these barriers’ impact weights and evaluate and rank potential solutions for overcoming the most critical barriers. This study proposes a multi-faceted approach integrating Multi-Criteria Decision Making (MCDM) methods, including the Delphi Method (DELPHI), Decision Making Trial And Evaluation Laboratory (DEMATEL), Combined Compromise Solution (COCOSO), Technique for Order Performance by Similarity to Ideal Solution (TOPSIS), and the Multiplicative form of Multi-objective Optimization based on Ratio Analysis method (MULTIMOORA), alongside the Quantum Spherical Fuzzy (QSF) set analysis. This combination, especially integrating QSF DELPHI and QSF COCOSO models, represents a significant methodological contribution. Results indicate that Vietnam’s primary barriers to quantum computing are limited infrastructure, high costs, and a lack of skilled workforce. Correspondingly, the most highly rated solution strategies are cost reduction initiatives, investment in infrastructure, and training and education programs, which should be prioritized and considered to remove potential barriers. These findings highlight the need to prioritize addressing issues in facilities, human resources, and financial resources. This research bridges theoretical insights with actionable recommendations, providing a comprehensive roadmap for mitigating barriers and promoting the integration of quantum computing into Vietnam’s data analysis landscape.

In recent years, deep learning methodologies have been increasingly applied to the intricate challenges of visual-inertial odometry (VIO), especially in scenarios with rapid movements and scenes lacking clear structure. This paper introduces a novel hybrid approach that leverages the inherent strengths of traditional VIO techniques, while harnessing the potential of advanced machine learning technologies. By seamlessly integrating an iterated extended Kalman filter with deep learning techniques, our approach systematically takes into account uncertainties, thereby enhancing the overall reliability and robustness of the system. The proposed algorithm has been rigorously evaluated on the KITTI and EuroC datasets, outperforming other deep learning VIO methods. It achieved a translation error of 2.28% and a rotation error of 0.226 degrees per 100 meters on the KITTI odometry dataset.

Today, unmanned aerial vehicles (UAVs) are extensively employed in wireless communication systems due to their mobility and exceptional maneuverability. With the advantage of flying high to increase the line of sight (LOS) probability with ground user equipment (UE), many studies have demonstrated that using UAVs to replace ground base stations (BS) can reduce transmission power loss. However, most research results are still limited to simulations with implicit formulas. This paper provides a closed formula for calculating the communication power gain of a cell based on environmental parameters. To achieve this, we propose an approximately linear mathematical model that allows for determining the formula for the optimal elevation angle, thereby calculating the coverage area of the UAV-BS and communication power gain of the coverage area. Simulations show that the proposed formula can be applied over a wide range of environmental parameters, and the quantitative results enable rapid cell planning for communication designers.

In recent years, with the advancement of information technology and digital transformation, the number of IoT devices has been rapidly increasing. Despite efforts to raise awareness and encourage users to enhance security for these IoT devices, a large number of them remain connected to the Internet with poor security, using simple passwords or default credentials. These insecure IoT devices unintentionally become useful tools for malicious actors. By utilizing Worm-type Malware, a type of malware capable of rapid infection, attackers can infiltrate and control numerous devices in a very short period. Once compromised, these IoT devices are used in network attacks such as DDoS or malware propagation. Representing modern Worm-type Malware, two Malware have continuously attracted attention since their appearance: Mirai and Hajime. Specifically, Mirai is a highly infectious malware that has caused numerous DDoS attacks, resulting in significant financial and reputational damage to several corporations and businesses. Similarly, Hajime is also a powerful malware with a propagation principle similar to Mirai. Similar to past investigations into malware such as Code-red, SQL Slammer and Conficker, previous studies focused on these two malware were based on statistical analysis of the number of infected IP addresses within fixed time frames or geographical locations. However, relying on this simple statistical method, previous studies didn’t assess the persistence, intermittency, or variations in the infection process of each IoT device. This led to incomplete reflections of the infection process and accurate changes in the numbers of Mirai and Hajime. Intending to collect comprehensive information and make accurate assessments reflecting the infection status, variations, and characteristics of Mirai and Hajime during the infection process, this study applies a new investigation method and evaluation index using data collected from the Network Telescope in the Darknet space. The investigative results of this study have revealed the changes and assessed the infection scale of Mirai and Hajime since their appearance. Additionally, our research has identified the update time of Hajime, estimated the current number of infected devices, and quantitatively evaluated the impact of Hajime on the infection activities of Mirai.

This paper proposes a new space vector pulse-width modulation technique for reducing the common-mode voltage (SVPWM2) in a three-level T-type neutral-point-clamped voltage source inverter working under switch-open-circuit fault on a neutral-point-connected switch. Under this circumstance, the space vector diagram of the inverter becomes asymmetrical so that if the conventional reduced common-mode voltage space vector PWM (SVPWM1) patterns are still applied to IGBT drivers, the output waveforms are heavily deformed, thus leading to deteriorated harmonic distortion. Regarding SOC fault tolerant control, there are several traditional methods available. However, they are either costly due to additional hardware expenses or suffer from poor PMW performances using only two voltage level (2L) switching techniques. To address these issues, this paper introduces a cost-effective strategy: a reduced CMV space vector PWM strategy for fault tolerant control. This approach utilizes a hybrid two- and three-level switching principle to enhance PWM performances, resulting in reduced output harmonic distortion, lower voltage stress, and lower switching loss compared to the economical 2L PWM method. The SVPWM2 algorithm is firstly analyzed and synthesized in the $\alpha $ - $\beta $ coordinates and then implemented in MATLAB/Simulink and experimental hardware with TMS320F28377 microcontroller. Simulation and experiment waveforms are further evaluated in terms of total harmonic distortion and weighted total harmonic distortion to validate the effectiveness of the proposed algorithm.

The novel methods for determining the optimal moving trajectories and formation control of the Multiple Mobile Robot System (MMRS) are studied in this paper. The Ant Colony Optimization (ACO) algorithm combined with the A $^{\ast }$ algorithm is employed to solve the Inner Travel Salesmen Problem (ITSP) and Inner Multiple Travel Salesman Problem (IMTSP) to obtain the optimal path for MMRS. The maps of the working environment are built by merging three scanned single maps obtained by three mobile robots to save time. The multiple mobile robots are located by using the Adaptive Monte Carlo Localization (AMCL) algorithm and the Dynamic Window Approach (DWA) algorithm is designed to avoid both static and dynamic obstacles. The leader-follower is employed to control the moving of the MMRS under the specific formations. Finally, the simulation results in Gezabo and Matlab, and the experiments are provided to prove the success of the proposed methods.

Medical image segmentation is crucial for deep learning (DL) applications in clinical settings. Ensuring accurate segmentation is challenging due to diverse image sources and significant data sharing and privacy concerns in centralized learning setups. To address these challenges, we introduce a novel federated learning (FL) framework tailored for breast cancer. First, we use random regions of interest (ROIs) and bilinear interpolation to determine pixel color intensity based on neighboring pixels, addressing data inconsistencies from heterogeneous distribution parameters and increasing dataset size. We then employ the UNet model with a deep convolutional backbone (Visual Geometry Group [VGG]) to train the augmented data, enhancing recognition during training and testing. Second, we apply the Gaussian Mixture Model (GMM) to improve segmentation quality. This approach effectively manages distinct data distributions across hospitals and highlights images with a higher likelihood of tumor presence. Compared to other segmentation algorithms, GMM enhances the salience of valuable images, improving tumor detection. Finally, extensive experiments in two scenarios, federated averaging (FedAvg) and federated batch normalization (FedBN), demonstrate that our method outperforms several state-of-the-art segmentation methods on five public breast cancer datasets. These findings validate the effectiveness of our proposed framework, promising significant benefits for the community and society.

Terrestrial networks form the fundamental infrastructure of modern communication systems, serving more than 4 billion users globally. However, terrestrial networks are facing a wide range of challenges, from coverage and reliability to interference and congestion. As the demands of the 6G era are expected to be much higher, it is crucial to address these challenges to ensure a robust and efficient communication infrastructure for the future. To address these problems, Non-terrestrial Network (NTN) has emerged to be a promising solution. NTNs are communication networks that leverage airborne (e.g., unmanned aerial vehicles) and spaceborne vehicles (e.g., satellites) to facilitate ultra-reliable communications and connectivity with high data rates and low latency over expansive regions. This article aims to provide a comprehensive survey on the utilization of network slicing, Artificial Intelligence/Machine Learning (AI/ML), and Open Radio Access Network (ORAN) to address diverse challenges of NTNs from the perspectives of both academia and industry. Particularly, we first provide an in-depth tutorial on NTN and the key enabling technologies including network slicing, AI/ML, and ORAN. Then, we provide a comprehensive survey on how network slicing and AI/ML have been leveraged to overcome the challenges that NTNs are facing. Moreover, we present how ORAN can be utilized for NTNs. Finally, we highlight important challenges, open issues, and future research directions of NTN in the 6G era.

Massive jellyfish outbreaks have put human lives and marine ecosystems in great danger. As a result, the jellyfish detection methods have drawn a lot of attention, following two directions optical and sonar imaging. This work focuses on using optical imagery and CNN-based deep-learning object detection models to detect jellyfish. While labeled data of jellyfish play an important part in training deep learning models, there are a few open and available labeled datasets. Hence, we create our dataset to train these models using our model-assisted labeling method with over 11 thousand images of underwater jellyfish and corresponding annotation files in PASCAL VOC format. Our model-assisted labeling method saves the work of classical manual labeling by 70 percent, which is developed into application with YOLOv5. However, the YOLOv5 baseline suffers from the trade-off between real-time performance and low accuracy. Hence, an improved YOLOv5-nano is introduced based on adding GAM and replacing conventional Conv with CoordCov modules into the backbone of the conventional structure. The experiment results show that our improved model increases the accuracy of the conventional one by 1.3% and outperforms others including RetinaNet, SSD, Faster R-CNN, YOLOv6, and YOLOv8 at 89.1% mAP@0.5. On generalization performance, we verify the effectiveness of our work by conducting a test set of 15 different types of jellyfish with various shapes, colors, resolutions, and backgrounds. To conclude, our work establishes a comprehensive system from labeling the data, improving object detectors, and developing a feasible real-time jellyfish detector.

Many real-life problems present multiple options necessitating modern researchers to identify the most suitable choice or decision. In this context, fuzzy sets have proven to be valuable tools for addressing such problems. Picture fuzzy sets (PFSs), due to their increased flexibility and broader domain as compared to other extensions of fuzzy sets (FSs) have been extensively used in solving various decision-making problems. To achieve optimal decisions, various fuzzy graphs and fuzzy hypergraph structures are employed. Picture fuzzy directed hypergraphs (PFDHGs) offer greater utility in addressing problems with multiple options or uncertainties. In this study, we introduce novel concepts of PFDHGs, along with their applications in decision-making theory and hazardous chemicals management. Initially, we present the foundational concepts of PFDHGs, including essential terminologies such as hyper-edge and height. Subsequently, we explore various types of PFDHGs, such as simple PFDHGs and support simple PFDHGs. Additionally, we introduce the concept of $({\widetilde {\alpha }}, {\widetilde {\beta }}, {\widetilde {\gamma }})$ -level hyperarcs and elucidate the fundamental sequence in PFDHGs based on these arcs. The study further examines elementary PFDHGs, partial PFDHGs, and ordered PFDHGs, investigating the relationships among them. Structural properties of PFDHGs are analyzed through different types of homomorphisms of PFDHGs. We also define the order and size of PFDHGs. Finally, we present the applications of PFDHGs in decision-making and managing hazardous chemicals, emphasizing the innovative contributions and practical implications of our work.

This paper presents a Blockchain-based framework for providing Blockchain services for purposes of stability in terms of consensus protocol infrastructure and governance mechanisms and accessible auxiliary services suitable for the vast majority of current business needs, including fundamental factors such as digital identity with autonomous identity, building solutions to ensure transaction privacy with zero-knowledge proofs, and other services related to digital assets. The proposed framework helps promote digital transformation for businesses, especially small and medium enterprises with limited resources and costs, to apply Blockchain technology to their business models, increasing competitive advantages and assisting the companies in focusing on business logic while still using Blockchain technology in their functions.

While computer vision and computing technology advances have facilitated advanced driver assistance applications, systems with multi-task design remain highly demanding to operate at high speed on resource-constrained devices. Our study addresses this challenge by proposing a real-time driver assistance solution specifically developed for a single Jetson AGX Xavier embedded device. It simultaneously performs lane detection, traffic object detection and recognition, and rule-based scene analysis. To achieve high throughput (up to 43 frames per second) without reliance on additional hardware or cloud server, the system exploits Jetson device’s specialized AI accelerator and employs various optimization techniques: multithreaded programming, the TensorRT framework, and post-training quantization. The modular design integrates state-of-the-art task-specific methods and ensures adaptation to diverse traffic scenarios across countries as well as future hardware and solutions. Experimental results using a Korean dashcam traffic dataset validated the system’s effectiveness and practicality.

Substitution boxes (S-Boxes) are essential elements of modern block ciphers, serving as non-linear characteristics that enhance the resistance of these ciphers against cryptanalysis. This research presents a new approach for constructing lightweight S-Boxes that possess strong cryptographic characteristics by combining an enhanced logistic map and an enhanced sine map. The proposed novel algorithm has optimized multiple parameters according to the security threshold set. This study has conducted the most comprehensive evaluation of criteria for S-Boxes to date. The analysis results of the generated $4\times 4$ and $5\times 5$ S-Boxes have achieved optimal criteria in terms of the Strict Avalanche Criterion (SAC) and the Bit Independence Criterion (BIC) that no previous S-Boxes have achieved. Furthermore, parameters related to side-channel attack resistance have been optimized in the design stage of S-Boxes. The constructed S-Boxes with high security can be applied to replace S-Boxes of the same size in existing algorithms or to develop lightweight block cipher algorithms. This research has made a significant contribution to the construction of S-Boxes.

In modern power systems, the increasing penetration of renewable energy resources has reduced frequency stability. Consequently, wind turbines (WTs) are expected to participate in frequency regulation. By integrating the electrostatic energy stored in the DC capacitor with the kinetic energy stored in the rotating mass of the WT, an advanced control method is implemented to facilitate frequency regulation. However, previous studies have mainly concentrated on frequency control using WTs but ignored the rotor speed recovery of WT. Therefore, this work proposes a novel control strategy that separates the operating mode of a WT into two distinct states: the frequency support and the recovery of rotation speed stages. The proposed method improves the recovery process for rotor speeds after a WT participates in frequency support, returning the WT quickly to its normal operation to reduce wind energy loss. Furthermore, a new adaptive DC virtual inertia (DVIC) coefficient that utilizes real-time DC voltage to adjust the DVIC’s output smoothly is proposed. The proposed strategy is compared with other mature control strategies and implemented in PSCAD/EMTDC to verify their effectiveness and robustness. The simulation results demonstrate that the proposed method outperforms other mature methods because it reduces wind energy loss by at least 15 % when WTs participate in the frequency support process.

A novel strategy is proposed to tackle an optimal dispatch of a microgrid in response to dynamic conditions, utilizing a water wave optimization (WWO) algorithm and considering a day-ahead load forecasting. Amongst meta-heuristic algorithms, the WWO algorithm stands out in terms of population size, parameter tuning, exploitation and exploration, convergence speed, as well as optimization mechanism. It leverages its ability to efficiently explore solution spaces and adapt to changing conditions. It is applied to the dynamic optimal dispatch of a microgrid with the uncertainty of load power considered and solved by day-ahead load forecasting. It dynamically adjusts the microgrid operation in response to these inputs, ensuring optimal decision-making in the face of varying load scenarios. With the competition of various day-ahead load forecasting techniques in the microgrid, a multi-variate linear regression (MLR) model shows its advantage features, being more transparent, more effective, and more robust than other techniques, especially transparent explainability, as well as simple and fast in model training. These are requirements to achieve the result of day-ahead load forecasting. Thus, the MLR model is proposed to forecast day-ahead load in the microgrid in this paper. The simulation results show that the percentage error (PE) between the MLR model-based forecasted and actual load powers is always less than 4.42%, the mean absolute percentage error (MAPE) of the forecasting result is 3.33%, and the execution time is 49 (s). These achievements meet the accurate and fast requirements. They are completely competitive with the results of using other techniques such as convolutional neural networks (CNN) and long short-term memory (LSTM), especially in the execution time. This has contributed to improving the efficiency of the dynamic optimal dispatch in the microgrid. Then, the diesel generation, battery energy storage, and total microgrid generation costs are 68.76 ( ${\$}$ ), 5.09 ( ${\$}$ ), and 73.85 ( ${\$}$ ) respectively by using the WWO algorithm which are better than those by using a genetic algorithm (GA), a non-dominated sorting genetic algorithm-II (NSGA-II), a particle swarm optimization (PSO) algorithm, and a transient search optimization (TSO) algorithm in the microgrid. The findings offer valuable insights for microgrid operators, energy planners, and policymakers seeking sustainable and cost-effective solutions for distributed energy resource management.

Nowadays, video streaming has become a popular form of multimedia for communication and entertainment. The rapid traffic explosion spurred by the development of emerging video-centric services such as e-science, virtual reality, and video conference has caused network congestion and the degradation in quality of experience (QoE), especially in heterogeneous mobile networks. To cope with that, the development of QoE-efficient video streaming solutions, i.e., HTTP adaptive video streaming, is critical. In this work, we investigate HTTP adaptive video streaming solutions that are capable of improving QoE for heterogeneous mobile networks in which the network conditions including bandwidth are significantly varied. We propose an effective QoE-aware adaptive bitrate video streaming scheme that integrates a bandwidth prediction based on Gated Recurrent Unit (GRU) neural networks with an adaptive bitrate selection strategy to wisely determine the suitable quality level for each video chunk. Thanks to the accurate bandwidth estimation and the adaptation of each video chunk bitrate to the network conditions, QoE metrics have been enhanced. Numerical experiments have been deployed to verify the performance of our proposed solution in comparison with that of the notable conventional methods. The attained simulation results demonstrate that the developed solution is significantly more effective than the conventional methods. The proposed method obtains a performance increment, in terms of QoE, of up to 19.4% compared to the conventional ones.

In this paper, we present a novel time-efficient tag identification protocol for active radio frequency identification (RFID) systems. Our protocol design is based on a conventional M-ary Detecting Tree (MDT) that divides contention tags into M subgroups and performs the identification in two phases: scanning and collecting. While the MDT protocol effectively eliminates redundant empty slots (i.e., slots without any tag responses), it requires two success slots (i.e., slots with exactly one tag’s response) to identify one tag, raising concerns about potential identification delays. We introduce an innovative anti-collision protocol called Enhanced M-ary Detecting Tree (EMDT) to address this issue. The proposed EMDT switches from the scanning phase to the collecting phase when the estimated number of responding tags becomes smaller than a predefined threshold, which is found via theoretical analysis and has an optimal value of three. This adjustment allows the reader to typically require just one success slot for identifying each tag, thus significantly reducing the total time needed to identify all tags. The proposed phase-switching mechanism is backed by Zero-Estimation (ZE) for tag cardinality estimation and Manchester encoding for data transmission, both widely employed in RFID standards. Theoretical analysis and computer simulations are performed in different system settings, including ideal and non-ideal transmission channels, to validate the efficacy of the proposed protocol in comparison with conventional alternatives.

The efficiency of polynomial multiplication execution majorly impacts the performance of lattice-based post-quantum cryptosystems. In this research, we propose a high-speed hardware architecture to accelerate polynomial multiplication based on the Number Theoretic Transform (NTT) in CRYSTAL-Kyber and CRYSTAL-Dilithium. We design a Digital Signal Processing (DSP) architecture for modular multiplication in butterfly and Point-Wise Multiplication (PWM) operations. Our method reduces the critical path delay of an $n$ -bit multiplier to that of a ( $2n$ -2)-bit adder, optimizing both area and speed. These dedicated DSPs are employed in butterfly and PWM operations, completely eliminating the pre-process and post-process of NTT transforms. Furthermore, we introduce a novel unified pipelined architecture for the NTT and Inverse NTT (INTT) transformations of Kyber and Dilithium, with corresponding high-speed (Radix-2) and ultra-high-speed (Radix-4) versions. Lastly, we construct a complete hardware accelerator for polynomial matrix-vector multiplication in Kyber. The Field-Programmable Gate Array (FPGA) implementation results have proven that our designs have significantly improved execution time by $3.4\times $ – $9.6\times $ for the NTT transforms in Dilithium and $1.36\times $ – $34.16\times $ for Kyber polynomial multiplication, compared to previous studies reported to date. Additionally, the hardware footprint results indicate that our proposed architectures exhibit superior hardware performance in Area-Time-Product (ATP), corresponding to a 44%–96% improvement. The proposed architectures are efficient and well-suited for high-performance lattice-based cryptography systems.

The presence of sparsity in both input features and weights within convolutional neural networks offers a valuable opportunity to significantly reduce the number of computations required during inference. Moreover, the practice of compressing input data serves to diminish storage requirements and lower data transfer costs, ultimately enhancing overall power efficiency. However, the compression of randomly sparse inputs introduces challenges in the input matching process, often resulting in substantial hardware overhead and increased power consumption. These challenges arise due to the irregular nature of sparse inputs and the variability in convolutional strides. In response to these challenges, this research introduces an innovative data compression method, named Stride-Aware Sparsity Compression (StarSPA), designed to effectively locate valid input values and expedite the multiplication process. To fully capitalize on this proposed compression method, a weight-stationary approach is employed for efficient convolution. Comprehensive simulations demonstrate that the proposed accelerator achieves speedup factors of $1.17\times $ , $1.05\times $ , $1.09\times $ , $1.23\times $ , and $1.12\times $ when compared to the recent accelerator named SparTen for AlexNet, VGG16, GoogLeNet, ResNet34, and EfficientNetV2, respectively. Furthermore, FPGA implementation of the core reveals a noteworthy $2.55\times $ reduction in hardware size and a $5\times $ enhancement in energy efficiency when compared to SparTen.

In this study, we present some results on the preparation of a conical-shaped W-Cu composite by using the combination of the spark plasma sintering and subsequent spinning process. The experimental results demonstrated that the full density, fine grain and high hardness composite could be obtained. The relative density of the W-Cu composite after the spinning process increased up to 99.8% which is higher than that of the composite prepared by SPS processing (98.8%). Similarly, the microhardness of the composite after the spinning process is improved compared to the sample obtained after SPS processing. This was attributed to the grain size refinement and the increase of the dislocation density generated during the spinning process. Besides, the dependence of the microstructure and microhardness of the conical-shaped W-Cu composite on the positions of the cone was investigated in detail. As a result, the microstructure and microhardness of the cone were altered at various locations from the inside to the outside walls. This could be attributed to the difference in the interface bond strength between W and Cu matrix at different positions. The current work promises to be an innovative and encouraging approach to producing high-performance W-Cu composites for possible industrial applications and specially in shaped charge liners.

This paper proposes constructing a smart parking service model based on the Internet of Things and Fog computing architecture. The weaknesses of previous parking service delivery systems are that they did not offer a complete set of end-to-end parking services, such as inaccurate information about parking in real-time, poor system storage and expansion capacity, and knowledge to assist the driver in successfully reaching the parking space at a minimum cost. Therefore, our system offers a complete set of parking assistance services, from when the driver searches and makes a reservation to navigation services that help the driver successfully park at the desired parking spot. To address the problems of storage and system expansion, we also implemented services based on the Internet of Things and Fog computing platforms. The simulation and experimental implementation results show that the services we designed significantly improve the performance of the existing parking system compared to the other systems. Real-time vehicles are handled quickly, significantly reducing driver costs and increasing profits for managers.

This article proposes the design of a global-scale free-space optics/quantum key distribution (FSO/QKD) network based on a geosynchronous (GEO) satellite as the secret key source and low-Earth orbit (LEO) satellites as relay nodes for multiple legitimate users on the ground. The continuous variable QKD (CV-QKD) protocol with dual-threshold/direct detection (DT/DD) receivers is employed. The system performance is analyzed by considering the spreading loss, atmospheric attenuation, and turbulence. Based on the design criteria for the proposed system, we investigate the feasibility of a case study for the Japan QKD network considering the unauthorized receiver attack (URA) and beam-splitting attack (BSA). In addition, we analyze the secret-key rate performance of the proposed system and perform Monte Carlo simulations to verify analytical results.

The translation quality of machine translation systems depends on the parallel corpus used for training, particularly on the quantity and quality of the corpus. However, building a high-quality and large-scale parallel corpus is complex and expensive, particularly for a specific domain-parallel corpus. Therefore, data augmentation techniques are widely used in machine translation. The input of the back-translation method is monolingual text, which is available from many sources, and therefore, this method can be easily and effectively implemented to generate synthetic parallel data. In practice, monolingual texts can be collected from different sources, in which sources from websites often contain errors in grammar and spelling, sentence mismatch, or freestyle. Therefore, the quality of the output translation is reduced, leading to a low-quality parallel corpus generated by back-translation. In this study, we proposed a method for improving the quality of monolingual texts for back-translation. Moreover, we supplemented the data by pruning the translation table. We experimented with an English-Vietnamese neural machine translation using the IWSLT2015 dataset for training and testing in the legal domain. The results showed that the proposed method can effectively augment parallel data for machine translation, thereby improving translation quality. In our experimental cases, the BLEU score increased by 16.37 points compared with the baseline system.

With the rising data evolution, the demand for secured communications over networks is rising immensely. Elliptic Curve Cryptography (ECC) provides an attractive solution to fulfill the requirements of modern network applications. Many proposals published over the year over different variants of ECC satisfied some of the issues. Nevertheless, modern network applications such as Internet-of-Thing (IoT) and Software-Defined Networking (SDN) put the requirements on various aspects and can only be solved by different ECC algorithms. Looking at this point of view, an efficient architecture that could combine multiple ECC algorithms becomes an urgent request. In addition, even though many investigations of ECC on Field-Programmable Gate Arrays (FPGA), an efficient architecture that could be well-deployed on Application-Specific Integrated Circuit (ASIC) needs to get more study. Therefore, this paper proposes an area-efficient ECC hardware design that could integrate multiple ECC algorithms. The proposed design is deployed on both ASIC and FPGA platforms. Four well-known ECC-based Digital Signature Algorithms (DSAs), which are the Edwards-curve Digital Signature Algorithm (EdDSA) with Curve25519, Elliptic Curve Digital Signature Algorithm (ECDSA) with National Institute of Standards and Technology (NIST) Curve P-256, P-384, and P-521, are implemented. Furthermore, the design supports all DSA schemes: public-key generation, signature generation, and verification. We also provide optimized calculation flows for modular multiplication, modular inversion, point addition, point doubling, and Elliptic Curve Point Multiplication (ECPM) for two different elliptic curves: the NIST curve and Edward curve, on a unique architecture. The calculation processes are designed in projective coordinates and optimized in time and space to achieve a high level of parallelism. The proposed ECC processor could run up to 102-MHz on ASIC 180-nm and 109.7-MHz on Xilinx Virtex-7. In terms of area, The processor occupies 377,471 gates with 4.87- $mm^{2}$ on the ASIC platform and 11,401 slices on the FPGA platform. The experimental results show that our combinational design achieves area-efficient even when compared with other single-functional architecture on both ASIC and FPGA.

Attribute reduction, often referred to as feature selection, is a vital step in data preprocessing aimed at eliminating unnecessary attributes and enhancing the efficiency of classification models. Intuitionistic fuzzy sets are widely acknowledged for their highly effective approach to the attribute reduction problem in decision tables, especially in the context of noisy decision tables with low classification accuracy. Nonetheless, the computational complexity of this approach increases significantly due to the additional incorporation of a non-membership component when calculating the significance measures of each attribute. Besides, some intuitionistic fuzzy equivalence classes may include objects with relatively low similarity or high diversity degrees generated from noisy data. Performing calculations on these objects can be time-consuming and lead to inefficiencies in the attribute reduction process. To address the aforementioned issue, we first define an $\alpha,\beta $ -level set. This allows us to eliminate the mentioned noisy objects from the intuitionistic fuzzy equivalence classes and transform them into $\alpha,\beta $ -level intuitionistic fuzzy equivalence classes. Subsequently, we construct a distance measure between two $\alpha, \beta $ -level intuitionistic fuzzy partitions and define a new reduct to preserve the distance between two $\alpha, \beta $ -level intuitionistic fuzzy partitions generated by the condition attribute set and the decision attribute set. Finally, we propose a novel and efficient heuristic attribute reduction algorithm to find the new reduct, in which we also use a significance measure based on the $\alpha, \beta $ -level intuitionistic fuzzy partition distance to determine the vital attribute at each step of the algorithm. Obviously, our algorithm is also applied to various benchmark datasets for comparative analysis against existing algorithms. The experimental results demonstrate that the proposed algorithm not only improves the accuracy of the classification model but also significantly reduces execution time when compared to intuitionistic fuzzy rough set-based algorithms applied to noisy datasets with high dimensionality.

With the growing demand for high-speed connectivity and global coverage in future 6G networks, free-space optics (FSO)-based aerospace integrated networks, incorporating low Earth orbit (LEO) satellites, high-altitude platforms (HAP), and unmanned aerial vehicles (UAV), have recently attracted research efforts worldwide. Nevertheless, critical challenges on FSO links include weather conditions, atmospheric turbulence, and pointing misalignment. This paper addresses the design of error-control protocols for reliable FSO-based aerospace backhaul networks, when multiple UAVs are deployed as flying base stations (BSs). Specifically, we introduce a design proposal of a cooperative hybrid automatic repeat request (C-HARQ)-based frame allocation mechanism (FAM)/rate adaptation. The design proposal guarantees the latency fairness constraints among multiple UAVs experiencing varying turbulence channel conditions. An analytical channel model for HAP-aided relaying LEO satellite to the emerging UAV-mounted BS FSO links is provided. Moreover, we develop a comprehensive analytical framework taking into account the imperfect channel state information (CSI) to assess system performance metrics, including throughput, average frame delay, and energy efficiency. Numerical results confirm the effectiveness of our design proposal by comparing it with the conventional approach without FAM for various turbulence channel conditions and quality of service (QoS) requirements. Additionally, we offer a design guideline for the proper selection of parameters that can be helpful for the practical design of reliable FSO-based aerospace backhaul networks. Finally, the theoretical results are verified by Monte-Carlo simulations, along with some in-depth discussions.

ECHONET Lite stands as a leading protocol for smart home appliances in Japan, and the publication of data models plays a critical role in fostering collaboration with other ecosystems, not just for ECHONET Lite but for any protocol or standard. Typically, data models are meticulously crafted by experts through the arduous task of condensing and summarizing extensive specification documents. As an illustration, generating a data model solely for the ECHONET Lite protocol (without incorporating other protocols) can demand thousands of working hours. This paper presents an AI-driven solution aimed at alleviating the burden of laborious, repetitive tasks prone to errors during the creation of data models from ECHONET Lite specifications, automating these processes to save human effort. The proposed solution employs Natural Language Processing techniques to extract key vocabularies from natural language descriptions of the specification, mirroring the approach of experts. Consequently, this solution can generate several data models for the ECHONET Lite protocol in a matter of seconds, all using a standard laptop. The findings indicate that machines are capable of emulating experts in extracting vocabularies, ensuring both syntactic error-free outcomes and consistency in the generated data models. Furthermore, the machine offers rapid, dependable results and enhances the reusability of exported data models across various platforms. The generated data models meet the same requirements as those created by humans. This solution is integrated into an official workflow for generating data models for the ECHONET Lite web API and others.