[go: up one dir, main page]
Next Issue
Volume 12, March-2
Previous Issue
Volume 12, February-2
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
5.3
2.6
Journals
Electronics
Volume 12
Issue 5
share announcement

Electronics, Volume 12, Issue 5 (March-1 2023) – 196 articles

Cover Story (view full-size image): Microwave applications in medicine are gaining interest with a significant trend in healthcare research and development. Artificial intelligence (AI)-assisted microwave applications in medicine are expected to disrupt several areas, namely microwave imaging, dielectric spectroscopy for tissue classification, molecular diagnostics, telemetry, biohazard waste management, diagnostic pathology, biomedical sensor design, drug delivery, ablation treatment, and radiometry. AI-enabled microwave systems can be developed to augment healthcare, including clinical decision making, guiding treatment, and increasing resource-efficient facilities. This contribution outlines a platform for AI-based microwave solutions for future advancements in both clinical and technical aspects to enhance patient care. View this paper
  • Issues are regarded as officially published after their release is announced to the table of contents alert mailing list.
  • You may sign up for e-mail alerts to receive table of contents of newly released issues.
  • PDF is the official format for papers published in both, html and pdf forms. To view the papers in pdf format, click on the "PDF Full-text" link, and use the free Adobe Reader to open them.
Order results
Result details
Section
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
19 pages, 2743 KiB  
Article
A New Hybrid Fault Diagnosis Method for Wind Energy Converters
by Jinping Liang and Ke Zhang
Electronics 2023, 12(5), 1263; https://doi.org/10.3390/electronics12051263 - 6 Mar 2023
Cited by 8 | Viewed by 1536
Abstract
Fault diagnostic techniques can reduce the requirements for the experience of maintenance crews, accelerate maintenance speed, reduce maintenance cost, and increase electric energy production profitability. In this paper, a new hybrid fault diagnosis method based on multivariate empirical mode decomposition (MEMD), fuzzy entropy [...] Read more.
Fault diagnostic techniques can reduce the requirements for the experience of maintenance crews, accelerate maintenance speed, reduce maintenance cost, and increase electric energy production profitability. In this paper, a new hybrid fault diagnosis method based on multivariate empirical mode decomposition (MEMD), fuzzy entropy (FE), and an artificial fish swarm algorithm (AFSA)-support vector machine (SVM) is proposed to identify the faults of a wind energy converter. Firstly, the measured three-phase output voltage signals are processed by MEMD to obtain three sets of intrinsic mode functions (IMFs). The multi-scale analysis tool MEMD is used to extract the common modes matching the timescale. It studies the multi-scale relationship between three-phase voltages, realizes their synchronous analysis, and ensures that the number and frequency of the modes match and align. Then, FE is calculated to describe the IMFs’ complexity, and the IMFs-FE information is taken as fault feature to increase the robustness to working conditions and noise. Finally, the AFSA algorithm is used to optimize SVM parameters, solving the difficulty in selecting the penalty factor and radial basis function kernel. The effectiveness of the proposed method is verified in a simulated wind energy system, and the results show that the diagnostic accuracy for 22 fault modes is 98.7% under different wind speeds, and the average accuracy of 30 running can be maintained above 84% for different noise levels. The maximum, minimum, average, and standard deviation are provided to prove the robust and stable performance. Compared with the other methods, the proposed hybrid method shows excellent performance in terms of high accuracy, strong robustness, and computational efficiency. Full article
(This article belongs to the Topic Condition Monitoring and Diagnostic Methods for Power Equipment in New Energy Power Systems)
►▼ Show Figures

Figure 1

Figure 1
<p>The topology of a wind turbine system.</p> Full article ">Figure 2
<p>The simulated voltage <span class="html-italic">U</span><sub>abcg</sub> (<span class="html-italic">U</span><sub>ab</sub>, <span class="html-italic">U</span><sub>bc</sub>, <span class="html-italic">U</span><sub>ca</sub>). (<b>a</b>) Normal state; (<b>b</b>) OC fault occurred in T1; (<b>c</b>) OC fault occurred in T1 and T2; (<b>d</b>) OC fault occurred in T1 and T3.</p> Full article ">Figure 3
<p>Simulation model of the converter.</p> Full article ">Figure 4
<p>The decomposition results of the three-phase voltages <span class="html-italic">U</span><sub>ab</sub>, <span class="html-italic">U</span><sub>bc</sub>, and <span class="html-italic">U</span><sub>ca</sub>. (<b>a</b>) MEMD; (<b>b</b>) EMD.</p> Full article ">Figure 4 Cont.
<p>The decomposition results of the three-phase voltages <span class="html-italic">U</span><sub>ab</sub>, <span class="html-italic">U</span><sub>bc</sub>, and <span class="html-italic">U</span><sub>ca</sub>. (<b>a</b>) MEMD; (<b>b</b>) EMD.</p> Full article ">Figure 5
<p>The fuzzy entropy of the MEMD-IMFs of three-phase voltages in different fault modes.</p> Full article ">Figure 6
<p>Diagnostic results of AFSA-SVM model.</p> Full article ">
22 pages, 7281 KiB  
Article
Statistical Study on the Time Characteristics of the Transient EMD Excitation Current from the Pantograph–Catenary Arcing Discharge
by Mengzhe Jin, Shaoqian Wang, Shanghe Liu, Qingyuan Fang and Weidong Liu
Electronics 2023, 12(5), 1262; https://doi.org/10.3390/electronics12051262 - 6 Mar 2023
Cited by 2 | Viewed by 1553
Abstract
Electromagnetic disturbances (EMDs) resulting from arcing discharge between the pantograph and catenary pose a serious threat to the electromagnetic safety of electrified trains. The time characteristic of EMD excitation current has a significant impact on the generation mechanism and characteristics of electromagnetic emission [...] Read more.
Electromagnetic disturbances (EMDs) resulting from arcing discharge between the pantograph and catenary pose a serious threat to the electromagnetic safety of electrified trains. The time characteristic of EMD excitation current has a significant impact on the generation mechanism and characteristics of electromagnetic emission from pantograph–catenary discharge, but there have been few studies on the topic. In this paper, a large sample of time-domain waveform parameters were collected through laboratory measurements considering the high randomness nature of the arcing discharge. The reference distributions of the waveform parameters were selected using the Kolmogorov–Smirnov test, and the probability density function parameters that vary with applied voltages and discharge gap spacings were examined. Then, a stochastic model for the derivation of the discharge current waveform was proposed based on statistical results using a modified double exponential function whose parameters can be derived from physical properties. Waveforms of the excitation currents representing different EMD severities were generated by adjusting the quantiles of the fitting distributions. The validity of the stochastic model was demonstrated by comparing the measured and simulated waveforms for both single pulses and pulse trains. The proposed method and generated waveforms can help recreate the electromagnetic environment of pantograph–catenary arcing. Full article
(This article belongs to the Topic EMC and Reliability of Power Networks)
►▼ Show Figures

Figure 1

Figure 1
<p>Overall layout of the experiment: (<b>a</b>) experiment photo; (<b>b</b>) simplified schematic.</p> Full article ">Figure 2
<p>Arcing discharge between the pantograph strip and catenary wire.</p> Full article ">Figure 3
<p>Current waveform of a single arcing discharge.</p> Full article ">Figure 4
<p>Current waveform of a repeating pulse train over 10 ms acquisition period.</p> Full article ">Figure 5
<p>Distribution histogram of measured waveform parameters and theoretical reference distributions: (<b>a</b>) pulse amplitude (<span class="html-italic">A</span><sub>+</sub>); (<b>b</b>) rise time (<span class="html-italic">t</span><sub>r</sub>); (<b>c</b>) pulse width (<span class="html-italic">t</span><sub>w</sub>); (<b>d</b>) repetition interval time (<span class="html-italic">t</span><sub>in</sub>).</p> Full article ">Figure 6
<p>Experimental and theoretical cumulative distribution of <span class="html-italic">A</span><sub>+</sub>.</p> Full article ">Figure 7
<p>Statistical distributions of pulse amplitudes under different conditions: (<b>a</b>) <span class="html-italic">U</span> = 20 kV, <span class="html-italic">d</span> = 5 mm; (<b>b</b>) <span class="html-italic">U</span> = 25 kV, <span class="html-italic">d</span> = 5 mm; (<b>c</b>) <span class="html-italic">U</span> = 30 kV, <span class="html-italic">d</span> = 5 mm; (<b>d</b>) <span class="html-italic">U</span> = 35 kV, <span class="html-italic">d</span> = 5 mm; (<b>e</b>) <span class="html-italic">U</span> = 35 kV, <span class="html-italic">d</span> = 10 mm; (<b>f</b>) <span class="html-italic">U</span> = 35 kV, <span class="html-italic">d</span> = 15 mm.</p> Full article ">Figure 8
<p><span class="html-italic">A</span><sub>+</sub> versus <span class="html-italic">A</span><sub>−</sub> from one single current pulse at different conditions: (<b>a</b>) <span class="html-italic">U</span> = 20 kV, <span class="html-italic">d</span> = 5 mm; (<b>b</b>) <span class="html-italic">U</span> = 25 kV, <span class="html-italic">d</span> = 10 mm; and (<b>c</b>) <span class="html-italic">U</span> = 35 kV, <span class="html-italic">d</span> = 15 mm.</p> Full article ">Figure 9
<p>Ratio relationship between the positive amplitude and negative amplitude at different voltages: (<b>a</b>) average values; (<b>b</b>) STDs.</p> Full article ">Figure 10
<p>Measured and fitted results of <span class="html-italic">K</span><sub>R</sub> and <span class="html-italic">U</span><sub>0</sub> with varying gap spacing: (<b>a</b>) <span class="html-italic">K</span><sub>R</sub>; (<b>b</b>) <span class="html-italic">U</span><sub>0</sub>. The solid lines are the least squares fitting lines.</p> Full article ">Figure 11
<p>Measured waveform and simulated waveform of discharge current pulse. Measured physical characteristics: <span class="html-italic">t</span><sub>r</sub> = 22.98, <span class="html-italic">t</span><sub>w</sub> = 45.84, <span class="html-italic">A</span><sub>+</sub> = 8.53, <span class="html-italic">A</span><sub>−</sub> = 3.88; Calculated mathematical parameters: <span class="html-italic">α</span> = 0.0242, <span class="html-italic">β</span> = 0.0289, <span class="html-italic">t</span><sub>p</sub> = 92.91, <span class="html-italic">t</span><sub>e</sub> = 185.81; Fitted function parameters: <span class="html-italic">k</span><sub>1</sub> = 8.064, <span class="html-italic">k</span><sub>2</sub> = 8.427.</p> Full article ">Figure 12
<p>Schematic diagram of the random current pulse train. Single pulses are stretched to show their shapes in a pulse train.</p> Full article ">Figure 13
<p>Comparison of the simulated and measured current pulse train: (<b>a</b>) measured pulse train with <math display="inline"><semantics> <mrow> <mover accent="true"> <mrow> <msub> <mi>t</mi> <mi>in</mi> </msub> </mrow> <mo stretchy="true">¯</mo> </mover> <mo>=</mo> <mn>0.185</mn> </mrow> </semantics></math> ms and <math display="inline"><semantics> <mrow> <msub> <mi>s</mi> <mi>in</mi> </msub> <mo>=</mo> <mn>0.064</mn> </mrow> </semantics></math> ms; (<b>b</b>) simulated pulse train with <math display="inline"><semantics> <mi>µ</mi> </semantics></math> = 0.214 and <math display="inline"><semantics> <mi>σ</mi> </semantics></math> = 0.050 for the Extreme Value distribution of <span class="html-italic">t</span><sub>i</sub>.</p> Full article ">Figure 14
<p>Flowchart of the calculation of the time domain waveform of discharge current.</p> Full article ">Figure 15
<p>Simulated excitation current for moderate and serious EMD situations: (<b>a</b>) moderate situation with parameters: <span class="html-italic">α</span> = 0.0217, <span class="html-italic">β</span> = 0.0413, <span class="html-italic">A</span><sub>+</sub> = 14.580, <span class="html-italic">R</span><sub>A</sub> = 2.2, <span class="html-italic">t</span><sub>r</sub> = 18.731 ns, <span class="html-italic">t</span><sub>w</sub> = 40.949 ns; (<b>b</b>) severe situation with parameters: <span class="html-italic">α</span> = 0.0196, <span class="html-italic">β</span> = 0.0489, <span class="html-italic">A</span><sub>+</sub> = 19.218, <span class="html-italic">R</span><sub>A</sub> = 2.2, <span class="html-italic">t</span><sub>r</sub> = 17.522 ns, <span class="html-italic">t</span><sub>w</sub> = 42.386 ns; (<b>c</b>) critical situation with parameters: <span class="html-italic">α</span> = 0.0182, <span class="html-italic">β</span> = 0.0588, <span class="html-italic">A</span><sub>+</sub> = 26.252, <span class="html-italic">R</span><sub>A</sub> = 2.2, <span class="html-italic">t</span><sub>r</sub> = 15.419 ns, <span class="html-italic">t</span><sub>w</sub> = 43.805.</p> Full article ">
16 pages, 2614 KiB  
Article
Two-Axis Optoelectronic Stabilized Platform Based on Active Disturbance Rejection Controller with LuGre Friction Model
by Xueyan Hu, Shunjie Han, Yangyang Liu and Heran Wang
Electronics 2023, 12(5), 1261; https://doi.org/10.3390/electronics12051261 - 6 Mar 2023
Cited by 6 | Viewed by 1768
Abstract
To realize the stable tracking control of the optoelectronic stabilized platform system under nonlinear friction and external disturbance, an active disturbance rejection controller (ADRC) with friction compensation is proposed to improve the target tracking ability and anti-disturbance performance. First, a nonlinear LuGre observer [...] Read more.
To realize the stable tracking control of the optoelectronic stabilized platform system under nonlinear friction and external disturbance, an active disturbance rejection controller (ADRC) with friction compensation is proposed to improve the target tracking ability and anti-disturbance performance. First, a nonlinear LuGre observer is designed to estimate friction behavior and preliminarily suppress the interference of friction torque on the system. Then, an ADRC is introduced to further suppress the residual disturbance after friction compensation, and the stability of the ADRC system is also proved. The effectiveness of this scheme is proved by simulation experiments, and this scheme is compared with conventional ADRC and LuGre friction feedforward compensation. The simulation results show that an ADRC with LuGre friction compensation is better with trajectory tracking performance, which suppresses the influence of disturbance and improves the stability of the optoelectronic stabilized platform system. Full article
(This article belongs to the Special Issue Advanced Control Techniques of Power Electronics)
►▼ Show Figures

Figure 1

Figure 1
<p>Schematic diagram of two-axis four-frame optoelectronic stabilized platform.</p> Full article ">Figure 2
<p>Block diagram of three closed-loop control system for single-gimbal.</p> Full article ">Figure 3
<p>Current loop control block diagram.</p> Full article ">Figure 4
<p>Second-order ADRC system.</p> Full article ">Figure 5
<p>Block diagram of system disturbance after friction compensation.</p> Full article ">Figure 6
<p>Block diagram of the ADRC structure with LuGre friction compensation.</p> Full article ">Figure 7
<p>Sinusoidal trajectory tracking results. (<b>a</b>) Diagram of the angle tracking curve; (<b>b</b>) Diagram of the angle error curve; (<b>c</b>) Diagram of the angle speed tracking curve; (<b>d</b>) Diagram of the angle speed error curve.</p> Full article ">Figure 8
<p>Multiple sinusoidal trajectory tracking results. (<b>a</b>) Diagram of the angle tracking curve; (<b>b</b>) Diagram of the angle error curve; (<b>c</b>) Diagram of the angle speed tracking curve; (<b>d</b>) Diagram of the angle speed error curve.</p> Full article ">
12 pages, 648 KiB  
Article
Towards Effective Feature Selection for IoT Botnet Attack Detection Using a Genetic Algorithm
by Xiangyu Liu and Yanhui Du
Electronics 2023, 12(5), 1260; https://doi.org/10.3390/electronics12051260 - 6 Mar 2023
Cited by 19 | Viewed by 2808
Abstract
With the large-scale use of the Internet of Things, security issues have become increasingly prominent. The accurate detection of network attacks in the IoT environment with limited resources is a key problem that urgently needs to be solved. The intrusion detection system based [...] Read more.
With the large-scale use of the Internet of Things, security issues have become increasingly prominent. The accurate detection of network attacks in the IoT environment with limited resources is a key problem that urgently needs to be solved. The intrusion detection system based on network traffic characteristics is one of the solutions for IoT security. However, the intrusion detection system has the problem of a large number of traffic features, which makes training and detection slow. Aiming at this problem, this work proposes a feature selection method based on a genetic algorithm. The experiments performed on the Bot-IoT botnet detection dataset show that this method successfully selects 6 features from the original 40 features, with a detection accuracy of 99.98% and an F1-score of 99.63%. Compared with other methods and without feature selection, this method has advantages in training time and detection accuracy. Full article
(This article belongs to the Special Issue Security Issues in the IoT)
►▼ Show Figures

Figure 1

Figure 1
<p>Confusion matrix of classification results for the decision tree.</p> Full article ">Figure 2
<p>Confusion matrix with six selected features using a decision tree.</p> Full article ">
12 pages, 5299 KiB  
Article
Detecting Human Falls in Poor Lighting: Object Detection and Tracking Approach for Indoor Safety
by Xing Zi, Kunal Chaturvedi, Ali Braytee, Jun Li and Mukesh Prasad
Electronics 2023, 12(5), 1259; https://doi.org/10.3390/electronics12051259 - 6 Mar 2023
Cited by 12 | Viewed by 3802
Abstract
Falls are one the leading causes of accidental death for all people, but the elderly are at particularly high risk. Falls are severe issue in the care of those elderly people who live alone and have limited access to health aides and skilled [...] Read more.
Falls are one the leading causes of accidental death for all people, but the elderly are at particularly high risk. Falls are severe issue in the care of those elderly people who live alone and have limited access to health aides and skilled nursing care. Conventional vision-based systems for fall detection are prone to failure in conditions with low illumination. Therefore, an automated system that detects falls in low-light conditions has become an urgent need for protecting vulnerable people. This paper proposes a novel vision-based fall detection system that uses object tracking and image enhancement techniques. The proposed approach is divided into two parts. First, the captured frames are optimized using a dual illumination estimation algorithm. Next, a deep-learning-based tracking framework that includes detection by YOLOv7 and tracking by the Deep SORT algorithm is proposed to perform fall detection. On the Le2i fall and UR fall detection (URFD) datasets, we evaluate the proposed method and demonstrate the effectiveness of fall detection in dark night environments with obstacles. Full article
(This article belongs to the Special Issue Deep Learning in Image Processing and Pattern Recognition)
►▼ Show Figures

Figure 1

Figure 1
<p>Some example video frames from the publicly available datasets (<b>a</b>) Le2i and (<b>b</b>) URFD dataset.</p> Full article ">Figure 2
<p>The detection results of YOLOv5 and Manual post-correction. (<b>a</b>) YOLOv5 pre-trained model. (<b>b</b>) Manual post-correction.</p> Full article ">Figure 3
<p>Schematic representation of the proposed framework.</p> Full article ">Figure 4
<p>Examples of images from Le2i Fall Detection dataset. (<b>a</b>) The original images. (<b>b</b>) The images after processing by the DUAL illumination estimation.</p> Full article ">Figure 5
<p>Visual results of object tracking on Le2i Fall Detection dataset. (<b>a</b>) YOLOv7 + Deep SORT method. (<b>b</b>) The proposed method.</p> Full article ">Figure 6
<p>Visual results of object tracking on UR Fall Detection dataset (<b>a</b>) YOLOv7 + Deep SORT method (<b>b</b>) the proposed method.</p> Full article ">
22 pages, 8167 KiB  
Article
A Novel Hybrid Deep Learning Model for Detecting and Classifying Non-Functional Requirements of Mobile Apps Issues
by Abdulsamad E. Yahya, Atef Gharbi, Wael M. S. Yafooz and Arafat Al-Dhaqm
Electronics 2023, 12(5), 1258; https://doi.org/10.3390/electronics12051258 - 6 Mar 2023
Cited by 18 | Viewed by 2532
Abstract
As a result of the speed and availability of the Internet, mobile devices and apps are in widespread usage throughout the world. Thus, they can be seen in the hands of nearly every person, helping us in our daily activities to accomplish many [...] Read more.
As a result of the speed and availability of the Internet, mobile devices and apps are in widespread usage throughout the world. Thus, they can be seen in the hands of nearly every person, helping us in our daily activities to accomplish many tasks with less effort and without wasting time. However, many issues occur while using mobile apps, which can be considered as issues of functional or non-functional requirements (NFRs). Users can add their comments as a review on the mobile app stores that provide for technical feedback, which can be used to improve the software quality and features of the mobile apps. Minimum attention has been given to such comments by scholars in addressing, detecting, and classifying issues related to NFRs, which are still considered challenging. The purpose of this paper is to propose a hybrid deep learning model to detect and classify NFRs (according to usability, reliability, performance, and supportability) of mobile apps using natural language processing methods. The hybrid model combines three deep learning (DL) architectures: a recurrent neural network (RNN) and two long short-term memory (LSTM) models. It starts with a dataset construction extracted from the user textual reviews that contain significant information in the Arabic language. Several experiments were conducted using machine learning classifiers (MCLs) and DL, such as ANN, LSTM, and bidirectional LSTM architecture to measure the performance of the proposed hybrid deep learning model. The experimental results show that the performance of the proposed hybrid deep learning model outperforms all other models in terms of the F1 score measure, which reached 96%. This model helps mobile developers improve the quality of their apps to meet user satisfaction and expectations by detecting and classifying issues relating to NFRs. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Research methods.</p> Full article ">Figure 2
<p>Feed-forward neural network.</p> Full article ">Figure 3
<p>LSTM model architecture.</p> Full article ">Figure 4
<p>Proposed hybrid deep learning model.</p> Full article ">Figure 5
<p>F1 score improvement: (<b>a</b>) dataset 1; (<b>b</b>) dataset 2.</p> Full article ">Figure 6
<p>Confusion matrix for ML classifiers using dataset 1 without data augmentation: (<b>a</b>) LR, (<b>b</b>) SVC, (<b>c</b>) KNN.</p> Full article ">Figure 7
<p>Confusion matrix for ML classifiers suing dataset 1 with data augmentation: (<b>a</b>) RF, (<b>b</b>) SVC, (<b>c</b>) DT.</p> Full article ">Figure 7 Cont.
<p>Confusion matrix for ML classifiers suing dataset 1 with data augmentation: (<b>a</b>) RF, (<b>b</b>) SVC, (<b>c</b>) DT.</p> Full article ">Figure 8
<p>Confusion matrix for ML classifiers suing dataset 2 without data augmentation: (<b>a</b>) SVC, (<b>b</b>) SGD, (<b>c</b>) KNN.</p> Full article ">Figure 9
<p>Confusion matrix for ML classifiers using dataset 2 with data augmentation: (<b>a</b>) RF, (<b>b</b>) SVC, (<b>c</b>) DT.</p> Full article ">Figure 10
<p>ANN model validation and testing accuracy and loss without data augmentation (dataset 1): (<b>a</b>) model accuracy (WODA); (<b>b</b>) model accuracy (WDA).</p> Full article ">Figure 11
<p>ANN model validation and testing accuracy and loss without data augmentation (dataset 2): (<b>a</b>) model accuracy (WODA); (<b>b</b>) model accuracy (WDA).</p> Full article ">Figure 12
<p>LSTM model validation and testing accuracy and loss dataset 1: (<b>a</b>) model accuracy (WODA); (<b>b</b>) model accuracy (WDA).</p> Full article ">Figure 13
<p>LSTM model validation and testing accuracy and lose (dataset 2): (<b>a</b>) model accuracy (WODA); (<b>b</b>) model accuracy (WDA).</p> Full article ">Figure 14
<p>Bi-LSTM model validation and testing accuracy and loss (dataset 1): (<b>a</b>) model accuracy (WODA); (<b>b</b>) model accuracy (WDA).</p> Full article ">Figure 15
<p>Bi-LSTM model validation and testing accuracy and loss using dataset 2: (<b>a</b>) model accuracy (WODA); (<b>b</b>) model accuracy (WDA).</p> Full article ">Figure 16
<p>Proposed model validation and testing accuracy and loss using dataset 1: (<b>a</b>) model accuracy using dataset 1—WODA; (<b>b</b>) model accuracy using dataset 1—WDA; (<b>c</b>) confusion matrix (WDA); (<b>d</b>) training and validation loss.</p> Full article ">Figure 17
<p>Proposed model validation and testing accuracy and loss using dataset 2: (<b>a</b>) model accuracy using dataset 2—WODA; (<b>b</b>) model accuracy using dataset 2—WDA; (<b>c</b>) confusion matrix (WDA); (<b>d</b>) training and validation loss.</p> Full article ">
16 pages, 4090 KiB  
Article
Object Detection for Hazardous Material Vehicles Based on Improved YOLOv5 Algorithm
by Pengcheng Zhu, Bolun Chen, Bushi Liu, Zifan Qi, Shanshan Wang and Ling Wang
Electronics 2023, 12(5), 1257; https://doi.org/10.3390/electronics12051257 - 6 Mar 2023
Cited by 7 | Viewed by 2616
Abstract
Hazardous material vehicles are a non-negligible mobile source of danger in transport and pose a significant safety risk. At present, the current detection technology is well developed, but it also faces a series of challenges such as a significant amount of computational effort [...] Read more.
Hazardous material vehicles are a non-negligible mobile source of danger in transport and pose a significant safety risk. At present, the current detection technology is well developed, but it also faces a series of challenges such as a significant amount of computational effort and unsatisfactory accuracy. To address these issues, this paper proposes a method based on YOLOv5 to improve the detection accuracy of hazardous material vehicles. The method introduces an attention module in the YOLOv5 backbone network as well as the neck network to achieve the purpose of extracting better features by assigning different weights to different parts of the feature map to suppress non-critical information. In order to enhance the fusion capability of the model under different sized feature maps, the SPPF (Spatial Pyramid Pooling-Fast) layer in the network is replaced by the SPPCSPC (Spatial Pyramid Pooling Cross Stage Partial Conv) layer. In addition, the bounding box loss function was replaced with the SIoU loss function in order to effectively speed up the bounding box regression and enhance the localization accuracy of the model. Experiments on the dataset show that the improved model has effectively improved the detection accuracy of hazardous chemical vehicles compared with the original model. Our model is of great significance for achieving traffic accident monitoring and effective emergency rescue. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Illustration of the architecture of the Mosaic data enhancement.</p> Full article ">Figure 2
<p>Illustration of the architecture of the Focus module.</p> Full article ">Figure 3
<p>Illustration of the architecture of the C3 module.</p> Full article ">Figure 4
<p>Illustration of the architecture of the attention module.</p> Full article ">Figure 5
<p>Illustration of the architecture of the channel attention module.</p> Full article ">Figure 6
<p>Illustration of the architecture of the spatial attention module.</p> Full article ">Figure 7
<p>(<b>a</b>) Illustration of the architecture of the SPPF module. (<b>b</b>) Illustration of the architecture of the SPPCSPC module.</p> Full article ">Figure 8
<p>Illustration of Path aggregation network. (<b>a</b>) FPN Backbone. (<b>b</b>) Bottom-up path augmentation. (<b>c</b>) Each building block.</p> Full article ">Figure 9
<p>Illustration of the architecture of the designed vehicle detection model.</p> Full article ">Figure 10
<p>The hazardous material vehicle object detection dataset. (<b>a</b>) Hazardous material vehicle images samples. (<b>b</b>) Ground-truth with bounding boxes.</p> Full article ">Figure 11
<p>A subset of single hazardous material vehicle detection results.</p> Full article ">Figure 12
<p>A subset of multi-hazardous material vehicle detection results.</p> Full article ">Figure 13
<p>A subset of multiple categories vehicle detection results.</p> Full article ">Figure 14
<p>Comparison of the improved model with the original. (<b>a</b>) Original YOLOv5 experimental result. (<b>b</b>) Improved YOLOv5 experimental results.</p> Full article ">
16 pages, 21467 KiB  
Article
SDRC-YOLO: A Novel Foreign Object Intrusion Detection Algorithm in Railway Scenarios
by Caixia Meng, Zhaonan Wang, Lei Shi, Yufei Gao, Yongcai Tao and Lin Wei
Electronics 2023, 12(5), 1256; https://doi.org/10.3390/electronics12051256 - 6 Mar 2023
Cited by 17 | Viewed by 3023
Abstract
Foreign object intrusion detection is vital to ensure the safety of railway transportation. Recently, object detection algorithms based on deep learning have been applied in a wide range of fields. However, in complex and volatile railway environments, high false detection, missed detection, and [...] Read more.
Foreign object intrusion detection is vital to ensure the safety of railway transportation. Recently, object detection algorithms based on deep learning have been applied in a wide range of fields. However, in complex and volatile railway environments, high false detection, missed detection, and poor timeliness still exist in traditional object detection methods. To address these problems, an efficient railway foreign object intrusion detection approach SDRC-YOLO is proposed. First, a hybrid attention mechanism that fuses local representation ability is proposed to improve the identification accuracy of small targets. Second, DW-Decoupled Head is proposed to construct a mixed feature channel to improve localization and classification ability. Third, a large convolution kernel is applied to build a larger receptive field and improve the feature extraction capability of the network. In addition, the lightweight universal upsampling operator CARAFE is employed to sample the size and proportion of the intruding foreign body features in order to accelerate the convergence speed of the network. Experimental results show that, compared with the baseline YOLOv5s algorithm, SDRC-YOLO improved the mean average precision (mAP) by 2.8% and 1.8% on datasets RS and Pascal VOC 2012, respectively. Full article
(This article belongs to the Section Computer Science & Engineering)
►▼ Show Figures

Figure 1

Figure 1
<p>SDRC-YOLO network structure.</p> Full article ">Figure 2
<p>SSA hybrid attention module.</p> Full article ">Figure 3
<p>Schematic diagram of the coupled head and the DW-Decoupled Head.</p> Full article ">Figure 4
<p>Schematic diagram of the RepLKNet architecture.</p> Full article ">Figure 5
<p>Schematic representation of the CARAFE upsampling operator structure.</p> Full article ">Figure 6
<p>Dataset examples.</p> Full article ">Figure 7
<p>Heat map comparison of various attention mechanisms.</p> Full article ">Figure 8
<p>Comparison between the CARAFE upsample and the nearest upsample.</p> Full article ">Figure 9
<p>Detection effect comparison of SDRC-YOLO and the original YOLOv5s.</p> Full article ">
14 pages, 1760 KiB  
Article
A Network Intrusion Detection Method Based on Domain Confusion
by Yanze Qu, Hailong Ma, Yiming Jiang and Youjun Bu
Electronics 2023, 12(5), 1255; https://doi.org/10.3390/electronics12051255 - 6 Mar 2023
Cited by 4 | Viewed by 1458
Abstract
Network intrusion detection models based on deep learning encounter problems in the migration application. The performance is not as good as expected. In this paper, a network intrusion detection method based on domain confusion is proposed to improve the migration performance of the [...] Read more.
Network intrusion detection models based on deep learning encounter problems in the migration application. The performance is not as good as expected. In this paper, a network intrusion detection method based on domain confusion is proposed to improve the migration performance of the model. A domain confusion network is designed for feature transformation based on the idea of domain adaptation, mapping the traffic data in different network environments to the same feature space. Meanwhile, a regularizer is proposed to control the information loss in the mapping process to ensure that the transformed feature obtains enough information for intrusion detection. The experiment results show that the detection performance of the model in this paper is similar to or even better than the traditional models, and the migration performance in different network environments is better than the traditional models. Full article
(This article belongs to the Section Networks)
►▼ Show Figures

Figure 1

Figure 1
<p>The Method to Train a Network Domain Confusion Model.</p> Full article ">Figure 2
<p>The Training Process. Traffic samples are mapped to the points in the feature space as (<b>a</b>), in which the orange ones represent these samples of source domain and the blue ones represent those of target domain. The discriminator is trained based on the dataset as (<b>b</b>). The parameters in domain confusion network are adjusted to make the points closer as (<b>c</b>). Then, the discriminator needs to be updated as (<b>d</b>). After enough iterations, as (<b>e</b>) shows, the discriminator cannot complete its task smoothly, and the probability that one point comes from source domain is about 0.5.</p> Full article ">Figure 3
<p>The Structure of Domain Confusion Network.</p> Full article ">Figure 4
<p>The Loss During Training (<span class="html-italic">λ</span> = 0.1).</p> Full article ">Figure 5
<p>The Loss at The End of Training (<span class="html-italic">λ</span> = 0.1).</p> Full article ">Figure 6
<p>The Loss During Training (<span class="html-italic">λ</span> = 0.5).</p> Full article ">Figure 7
<p>The Loss at The End of Training (<span class="html-italic">λ</span> = 0.5).</p> Full article ">Figure 8
<p>The Loss During Training (<span class="html-italic">λ</span> = 1).</p> Full article ">Figure 9
<p>The Loss at The End of Training (<span class="html-italic">λ</span> = 1).</p> Full article ">
18 pages, 890 KiB  
Article
Intelligent Computation Offloading Mechanism with Content Cache in Mobile Edge Computing
by Feixiang Li, Chao Fang, Mingzhe Liu, Ning Li and Tian Sun
Electronics 2023, 12(5), 1254; https://doi.org/10.3390/electronics12051254 - 6 Mar 2023
Cited by 4 | Viewed by 1760
Abstract
Edge computing is a promising technology to enable user equipment to share computing resources for task offloading. Due to the characteristics of the computing resource, how to design an efficient computation incentive mechanism with the appropriate task offloading and resource allocation strategies is [...] Read more.
Edge computing is a promising technology to enable user equipment to share computing resources for task offloading. Due to the characteristics of the computing resource, how to design an efficient computation incentive mechanism with the appropriate task offloading and resource allocation strategies is an essential issue. In this manuscript, we proposed an intelligent computation offloading mechanism with content cache in mobile edge computing. First, we provide the network framework for computation offloading with content cache in mobile edge computing. Then, by deriving necessary and sufficient conditions, an optimal contract is designed to obtain the joint task offloading, resource allocation, and a computation strategy with an intelligent mechanism. Simulation results demonstrate the efficiency of our proposed approach. Full article
(This article belongs to the Special Issue Resource Allocation in Cloud–Edge–End Cooperation Networks)
►▼ Show Figures

Figure 1

Figure 1
<p>Architecture of mobile edge computing.</p> Full article ">Figure 2
<p>Computation offloading with caching in ultra-dense networks.</p> Full article ">Figure 3
<p>Deep reinforcement learning for computation offloading with content cache.</p> Full article ">Figure 4
<p>Architecture of servers in MEC.</p> Full article ">Figure 5
<p>Description of different times.</p> Full article ">Figure 6
<p>Description of State Space.</p> Full article ">Figure 7
<p>Convolution Neural Network.</p> Full article ">Figure 8
<p>Simulation scenario.</p> Full article ">Figure 9
<p>Reward value changes versus iterations.</p> Full article ">Figure 10
<p>Learning rate changes versus iterations.</p> Full article ">Figure 11
<p>Reward value changes with different learning rate.</p> Full article ">Figure 12
<p>Reward value changes with different batch size.</p> Full article ">Figure 13
<p>Reward value changes with different interval.</p> Full article ">Figure 14
<p>Comparison with other algorithms.</p> Full article ">
22 pages, 3274 KiB  
Article
A Coverless Audio Steganography Based on Generative Adversarial Networks
by Jing Li, Kaixi Wang and Xiaozhu Jia
Electronics 2023, 12(5), 1253; https://doi.org/10.3390/electronics12051253 - 5 Mar 2023
Cited by 4 | Viewed by 3084
Abstract
Traditional audio steganography by cover modification causes changes to the cover features during the embedding of a secret, which is easy to detect with emerging neural-network steganalysis tools. To address the problem, this paper proposes a coverless audio-steganography model to conceal a secret [...] Read more.
Traditional audio steganography by cover modification causes changes to the cover features during the embedding of a secret, which is easy to detect with emerging neural-network steganalysis tools. To address the problem, this paper proposes a coverless audio-steganography model to conceal a secret audio. In this method, the stego-audio is directly synthesized by our model, which is based on the WaveGAN framework. An extractor is meticulously designed to reconstruct the secret audio, and it contains resolution blocks to learn the different resolution features. The method does not perform any modification to an existing or generated cover, and as far as we know, this is the first directly generated stego-audio. The experimental results also show that it is difficult for the current steganalysis methods to detect the existence of a secret in the stego-audio generated by our method because there is no cover audio. The MOS metric indicates that the generated stego-audio has high audio quality. The steganography capacity can be measured from two perspectives, one is that it can reach 50% of the stego-audio from the simple size perspective, the other is that 22–37 bits can be hidden in a two-second stego-audio from the semantic. In addition, we prove using spectrum diagrams in different forms that the extractor can reconstruct the secret audio successfully on hearing, which guarantees complete semantic transmission. Finally, the experiment of noise impacts on the stego-audio transmission shows that the extractor can still completely reconstruct the semantics of the secret audios, which indicates that the proposed method has good robustness. Full article
(This article belongs to the Special Issue AI-Driven Network Security and Privacy)
►▼ Show Figures

Figure 1

Figure 1
<p>The model of the proposed steganography.</p> Full article ">Figure 2
<p>The generator and discriminator structure.</p> Full article ">Figure 3
<p>The extractor structure.</p> Full article ">Figure 4
<p>The changes of the loss indicators for the generator and discriminator.</p> Full article ">Figure 5
<p>The changes of the loss indicators for the extractor.</p> Full article ">Figure 6
<p>The sample value distribution comparison. On the left is the distribution curve and histogram of the sample values in the real audios, where the <span class="html-italic">x</span>-axis represents random variables, the left <span class="html-italic">y</span>-axis represents the frequency, and the right <span class="html-italic">y</span>-axis represents the density which is calculated as multiplying the frequency by the group distance. The same is true on the right for the stego-audios.</p> Full article ">Figure 7
<p>The distributions of the stego-audios and real audio samples.</p> Full article ">Figure 8
<p>Illustration of Step (1).</p> Full article ">Figure 9
<p>Illustration of Step (2).</p> Full article ">Figure 10
<p>The Euclidean distances between the MFCCs of each stego-audio and the nearest neighbor in the training dataset as well as its nearest neighbor in the test dataset.</p> Full article ">Figure 11
<p>The time-domain waveform. The left column is the time-domain waveform of the secret audios, and the right column is the time-domain waveform of the reconstructed secret audios.</p> Full article ">Figure 12
<p>The frequency-domain waveform. The left column is the frequency-domain waveform of the secret audios, and the right column is the frequency-domain waveform of the reconstructed secret audios.</p> Full article ">Figure 13
<p>The log spectrogram. The left column is the log spectrogram of the secret audios, and the right column is log spectrogram of the reconstructed secret audios.</p> Full article ">
17 pages, 2372 KiB  
Article
RESTful API Analysis, Recommendation, and Client Code Retrieval
by Shang-Pin Ma, Ming-Jen Hsu, Hsiao-Jung Chen and Chuan-Jie Lin
Electronics 2023, 12(5), 1252; https://doi.org/10.3390/electronics12051252 - 5 Mar 2023
Viewed by 3148
Abstract
Numerous companies create innovative software systems using Web APIs (Application Programming Interfaces). API search engines and API directory services, such as ProgrammableWeb, Rapid API Hub, APIs.guru, and API Harmony, have been developed to facilitate the utilization of various APIs. Unfortunately, most API systems [...] Read more.
Numerous companies create innovative software systems using Web APIs (Application Programming Interfaces). API search engines and API directory services, such as ProgrammableWeb, Rapid API Hub, APIs.guru, and API Harmony, have been developed to facilitate the utilization of various APIs. Unfortunately, most API systems provide only superficial support, with no assistance in obtaining relevant APIs or examples of code usage. To better realize the “FAIR” (Findability, Accessibility, Interoperability, and Reusability) features for the usage of Web APIs, in this study, we developed an API inspection system (referred to as API Prober) to provide a new API directory service with multiple supplemental functionalities. To facilitate the findability and accessibility of APIs, API Prober transforms OAS (OpenAPI Specifications) into a graph structure and automatically annotates the semantic concepts using LDA (Latent Dirichlet Allocation) and WordNet. To enhance interoperability, API Prober also classifies APIs by clustering OAS documents and recommends alternative services to be substituted or merged with the target service. Finally, to support reusability, API Prober makes it possible to retrieve examples of API utilization code in Java by parsing source code in GitHub. The experimental results demonstrate the effectiveness of the API Prober in recommending relevant services and providing usage examples based on real-world client code. This research contributes to providing viable methods to appropriately analyze and cluster Web APIs, and recommend APIs and client code examples. Full article
(This article belongs to the Special Issue Feature Papers in Computer Science & Engineering)
►▼ Show Figures

Figure 1

Figure 1
<p>API Prober: operational concepts.</p> Full article ">Figure 2
<p>API Prober: system architecture.</p> Full article ">Figure 3
<p>API page in API Prober: an example of a “Language Tool” API.</p> Full article ">Figure 4
<p>The process used to differentiate substitutable services from mergeable services.</p> Full article ">Figure 5
<p>URL syntax diagram for extraction of code examples.</p> Full article ">Figure 6
<p>Number of services in all service clusters.</p> Full article ">Figure 7
<p>Top-K SRP for three types of service recommendation.</p> Full article ">Figure 8
<p>EP Top-3 evaluation results.</p> Full article ">
19 pages, 7357 KiB  
Article
Multi-Inverter Resonance Modal Analysis Based on Decomposed Conductance Model
by Lin Chen, Yonghai Xu, Shun Tao, Tianze Wang and Shuguang Sun
Electronics 2023, 12(5), 1251; https://doi.org/10.3390/electronics12051251 - 5 Mar 2023
Cited by 2 | Viewed by 1517
Abstract
The Norton equivalent model based on the transfer function and the frequency domain analysis method for inverter resonance analysis lacks a comprehensive analysis of the resonant characteristics, and more information about the resonant key components and the degree of participation cannot be obtained. [...] Read more.
The Norton equivalent model based on the transfer function and the frequency domain analysis method for inverter resonance analysis lacks a comprehensive analysis of the resonant characteristics, and more information about the resonant key components and the degree of participation cannot be obtained. In this paper, a decomposed conductance model is proposed to characterize the resonance characteristics of the multi-inverter grid-connected system and the effect of the equivalent control link of the inverter on the resonance in more detail by combining the modal analysis method and the sensitivity analysis method. Firstly, based on αβ coordinates, the conductance division is carried out for the dual-loop inverter control link with the voltage external loop and current internal loop using capacitor-current feedback damping, and the inverter model based on the decomposition conductance is derived. The mathematical model of the multi-inverter grid-connected system is then established. Secondly, the resonance characteristics of the system are analyzed by combining the modal and frequency domain analysis methods when the number of inverters, inverter parameters, and grid-side impedance are changed. Thirdly, the degree of involvement of the system components, especially the equivalent control link of the inverter in resonance conditions, is determined in combination with the proposed model and the sensitivity analysis method, which is the basis for proposing an effective suppression strategy. Finally, a simulation model is built to verify the proposed method and the analysis results. Full article
(This article belongs to the Special Issue Application of Power Electronics Technology in Energy System)
►▼ Show Figures

Figure 1

Figure 1
<p>A structure diagram of an <span class="html-italic">LCL</span> inverter.</p> Full article ">Figure 2
<p>Mathematical model of capacitive current feedback active damping control method.</p> Full article ">Figure 3
<p>The Bode diagram of the open-loop transfer function using damping control and conventional undamped control.</p> Full article ">Figure 4
<p>The establishment of the Norton Equivalence Model: (<b>a</b>) Equivalent transformation of the mathematical model; (<b>b</b>) The Norton two-port equivalent model.</p> Full article ">Figure 5
<p>The establishment of the double-decomposed conductance model: (<b>a</b>) the double-decomposed conductance small signal model; (<b>b</b>) the two-port equivalent model under the double-decomposed conductance.</p> Full article ">Figure 6
<p>The establishment of the triple decomposed conductance model: (<b>a</b>) The triple decomposition conductance small signal model; (<b>b</b>) The two-port equivalent model under triple decomposition conductance.</p> Full article ">Figure 7
<p>The equivalent model of a multi-inverter grid-connected system.</p> Full article ">Figure 8
<p>The specific process of resonant mode analysis.</p> Full article ">Figure 9
<p>The modal impedance curve when <span class="html-italic">n</span> is changed: (<b>a</b>) <span class="html-italic">n</span> = 1; (<b>b</b>) <span class="html-italic">n</span> = 2; (<b>c</b>) <span class="html-italic">n</span> = 3; (<b>d</b>) <span class="html-italic">n</span> = 4.</p> Full article ">Figure 10
<p>The characteristics of the amplitude-frequency response of each command of the inverter’s grid-connected current.</p> Full article ">Figure 11
<p>The resonance characteristic curve changes when <span class="html-italic">L</span><sub>g</sub> is changed.</p> Full article ">Figure 12
<p>The resonance characteristics curve when the inverter controller parameters are changed: (<b>a</b>) <span class="html-italic">k</span><sub>p</sub>; (<b>b</b>) <span class="html-italic">k</span><sub>r</sub>.</p> Full article ">Figure 13
<p>The resonance characteristics curve when the <span class="html-italic">LCL</span> parameters are changed: (<b>a</b>) <span class="html-italic">L</span><sub>1</sub>; (<b>b</b>) <span class="html-italic">L</span><sub>2</sub>; (<b>c</b>) <span class="html-italic">C</span>.</p> Full article ">Figure 14
<p>The results of the sensitivity analysis comparison are: (<b>a</b>) the sensitivity without the decomposed conductance model; (<b>b</b>) the sensitivity using the decomposed conductance model.</p> Full article ">Figure 15
<p>The node voltage spectrum at <span class="html-italic">n</span> = 2: (<b>a</b>) <span class="html-italic">U</span><sub>PCC</sub>; (<b>b</b>) <span class="html-italic">U</span><sub>1</sub>.</p> Full article ">Figure 16
<p>The node voltage spectrum when <span class="html-italic">L</span><sub>g</sub> is varied: (<b>a</b>) <span class="html-italic">L</span><sub>g</sub> = 0.1 mH; (<b>b</b>) <span class="html-italic">L</span><sub>g</sub> = 1 mH.</p> Full article ">Figure 17
<p>The node voltage spectrum when <span class="html-italic">L</span><sub>1</sub> is changed: (<b>a</b>) <span class="html-italic">L</span><sub>1</sub> = 0.8 mH; (<b>b</b>) <span class="html-italic">L</span><sub>1</sub> = 1.6 mH.</p> Full article ">Figure 18
<p>The node voltage spectrum when <span class="html-italic">L</span><sub>1</sub> is changed: (<b>a</b>) <span class="html-italic">k</span><sub>p</sub> = 3; (<b>b</b>) <span class="html-italic">k</span><sub>p</sub> = 3.8.</p> Full article ">Figure 19
<p>The resonance suppression strategy with PCC voltage feedforward.</p> Full article ">Figure 20
<p>The comparison of the harmonic voltage of Inverter 1 before and after adopting the resonance suppression strategy.</p> Full article ">
36 pages, 3268 KiB  
Article
Federated Learning-Based Lightweight Two-Factor Authentication Framework with Privacy Preservation for Mobile Sink in the Social IoMT
by B. D. Deebak and Seong Oun Hwang
Electronics 2023, 12(5), 1250; https://doi.org/10.3390/electronics12051250 - 5 Mar 2023
Cited by 5 | Viewed by 3213
Abstract
The social Internet of Medical Things (S-IoMT) highly demands dependable and non-invasive device identification and authentication and makes data services more prevalent in a reliable learning system. In real time, healthcare systems consistently acquire, analyze, and transform a few operational intelligence into actionable [...] Read more.
The social Internet of Medical Things (S-IoMT) highly demands dependable and non-invasive device identification and authentication and makes data services more prevalent in a reliable learning system. In real time, healthcare systems consistently acquire, analyze, and transform a few operational intelligence into actionable forms through digitization to capture the sensitive information of the patient. Since the S-IoMT tries to distribute health-related services using IoT devices and wireless technologies, protecting the privacy of data and security of the device is so crucial in any eHealth system. To fulfill the design objectives of eHealth, smart sensing technologies use built-in features of social networking services. Despite being more convenient in its potential use, a significant concern is a security preventing potential threats and infringement. Thus, this paper presents a lightweight two-factor authentication framework (L2FAK) with privacy-preserving functionality, which uses a mobile sink for smart eHealth. Formal and informal analyses prove that the proposed L2FAK can resist cyberattacks such as session stealing, message modification, and denial of service, guaranteeing device protection and data integrity. The learning analysis verifies the features of the physical layer using federated learning layered authentication (FLLA) to learn the data characteristics by exploring the learning framework of neural networks. In the evaluation, the core scenario is implemented on the TensorFlow Federated framework to examine FLLA and other relevant mechanisms on two correlated datasets, namely, MNIST and FashionMNIST. The analytical results show that the proposed FLLA can analyze the protection of privacy features effectively in order to guarantee an accuracy ≈89.83% to 93.41% better than other mechanisms. Lastly, a real-time testbed demonstrates the significance of the proposed L2FAK in achieving better quality metrics, such as transmission efficiency and overhead ratio than other state-of-the-art approaches. Full article
(This article belongs to the Special Issue Advanced Future Communication Techniques and Security Solutions for 6G and Internet of Things)
►▼ Show Figures

Figure 1

Figure 1
<p>A Data Cloud-Centric Architecture of eHealth System Using Federated Learning Approach.</p> Full article ">Figure 2
<p>A Smart eHealth System Model with Authentic Gateway and Layered Authentication.</p> Full article ">Figure 3
<p>Phase 4—Login and Authentication.</p> Full article ">Figure 4
<p>Test Accuracy Rate <math display="inline"><semantics> <mrow> <mrow> <mo>(</mo> <mo>%</mo> <mo>)</mo> </mrow> </mrow> </semantics></math> on MNIST.</p> Full article ">Figure 5
<p>Test Accuracy Rate <math display="inline"><semantics> <mrow> <mrow> <mo>(</mo> <mo>%</mo> <mo>)</mo> </mrow> </mrow> </semantics></math> on FashionMNIST.</p> Full article ">Figure 6
<p>Data Transmission Ratio <math display="inline"><semantics> <mrow> <mrow> <mo>(</mo> <mo>%</mo> <mo>)</mo> </mrow> </mrow> </semantics></math> vs. the Number of User Identities <math display="inline"><semantics> <mrow> <mo>(</mo> <mo>#</mo> <mo>)</mo> </mrow> </semantics></math>.</p> Full article ">Figure 7
<p>Overhead Ratio <math display="inline"><semantics> <mrow> <mrow> <mo>(</mo> <mo>%</mo> <mo>)</mo> </mrow> </mrow> </semantics></math> vs. the Number of User Identities <math display="inline"><semantics> <mrow> <mo>(</mo> <mo>#</mo> <mo>)</mo> </mrow> </semantics></math>.</p> Full article ">
16 pages, 2805 KiB  
Article
Robust Subspace Clustering with Block Diagonal Representation for Noisy Image Datasets
by Qiang Li, Ziqi Xie and Lihong Wang
Electronics 2023, 12(5), 1249; https://doi.org/10.3390/electronics12051249 - 5 Mar 2023
Cited by 2 | Viewed by 1496
Abstract
As a relatively advanced method, the subspace clustering algorithm by block diagonal representation (BDR) will be competent in performing subspace clustering on a dataset if the dataset is assumed to be noise-free and drawn from the union of independent linear subspaces. Unfortunately, this [...] Read more.
As a relatively advanced method, the subspace clustering algorithm by block diagonal representation (BDR) will be competent in performing subspace clustering on a dataset if the dataset is assumed to be noise-free and drawn from the union of independent linear subspaces. Unfortunately, this assumption is far from reality, since the real data are usually corrupted by various noises and the subspaces of data overlap with each other, the performance of linear subspace clustering algorithms, including BDR, degrades on the real complex data. To solve this problem, we design a new objective function based on BDR, in which l2,1 norm of the reconstruction error is introduced to model the noises and improve the robustness of the algorithm. After optimizing the objective function, we present the corresponding subspace clustering algorithm to pursue a self-expressive coefficient matrix with a block diagonal structure for a noisy dataset. An affinity matrix is constructed based on the coefficient matrix, and then fed to the spectral clustering algorithm to obtain the final clustering results. Experiments on several artificial noisy image datasets show that the proposed algorithm has robustness and better clustering performance than the compared algorithms. Full article
(This article belongs to the Special Issue Advances in Spatiotemporal Data Management and Analytics)
►▼ Show Figures

Figure 1

Figure 1
<p>The framework of OBDR.</p> Full article ">Figure 2
<p>MNIST confused with a small number of ORL images.</p> Full article ">Figure 3
<p>Masking photos from ORL.</p> Full article ">Figure 4
<p>YaleB dataset.</p> Full article ">Figure 5
<p>The computational time of algorithms on Dataset1~Dataset4.</p> Full article ">Figure 6
<p>The effect of <span class="html-italic">ρ</span>.</p> Full article ">Figure 7
<p>The ACCs of OBDR with different λ and γ.</p> Full article ">
13 pages, 2044 KiB  
Article
Energy Efficient Data Dissemination for Large-Scale Smart Farming Using Reinforcement Learning
by Muhammad Yasir Ali, Abdullah Alsaeedi, Syed Atif Ali Shah, Wael M. S. Yafooz and Asad Waqar Malik
Electronics 2023, 12(5), 1248; https://doi.org/10.3390/electronics12051248 - 5 Mar 2023
Cited by 4 | Viewed by 1675
Abstract
Smart farming is essential to increasing crop production, and there is a need to consider the technological advancements of this era; modern technology has helped us to gain more accuracy in fertilizing, watering, and adding pesticides to the crops, as well as monitoring [...] Read more.
Smart farming is essential to increasing crop production, and there is a need to consider the technological advancements of this era; modern technology has helped us to gain more accuracy in fertilizing, watering, and adding pesticides to the crops, as well as monitoring the conditions of the environment. Nowadays, more and more sophisticated sensors are being developed, but on a larger scale, agricultural networks and the efficient management of them is very crucial in order to obtain proper benefits from technology. Our idea is to achieve sustainability in large-scale farms by improving communication between wireless sensor nodes and base stations. We want to increase communication efficiency by introducing machine learning algorithms. Reinforcement learning is the area of machine learning which is concerned with how involved agents are supposed to take action in specified environments to maximize reward and achieve a common goal. In our network, a large number of sensors are being deployed on large-scale fields; reinforcement learning is used to find the optimal set of paths towards the base station. After a number of successful paths have been developed, they are then used to transmit the sensed data from the fields. The simulation results have shown that in larger scales, our proposed model had less transmission delay than the shortest path transmission model and broadcasting techniques that were tested against the data transmission paths developed by reinforcement learning. Full article
(This article belongs to the Special Issue Real-Time Digital Control Technologies and Applications)
►▼ Show Figures

Figure 1

Figure 1
<p>Large-scale farming scenario, base node is deployed at the west region.</p> Full article ">Figure 2
<p>Geographic layout of the farm fields marked with numbers. The base station is located at the west side, and collects all the data for the decision.</p> Full article ">Figure 3
<p>Average transmission delay per message for different sensor sets.</p> Full article ">Figure 4
<p>Average energy consumption for different sensor sets.</p> Full article ">Figure 5
<p>Average residual energy after sending messages for different sensor sets.</p> Full article ">Figure 6
<p>Packet delivery rate for S1 to S6 deployment.</p> Full article ">
25 pages, 453 KiB  
Article
Exploring Personal Data Processing in Video Conferencing Apps
by Georgios Achilleos, Konstantinos Limniotis and Nicholas Kolokotronis
Electronics 2023, 12(5), 1247; https://doi.org/10.3390/electronics12051247 - 5 Mar 2023
Cited by 3 | Viewed by 2189
Abstract
The use of video conferencing applications has increased tremendously in recent years, particularly due to the COVID-19 pandemic and the associated restrictions on movements. As a result, the corresponding smart apps have also seen increased usage, leading to a surge in downloads of [...] Read more.
The use of video conferencing applications has increased tremendously in recent years, particularly due to the COVID-19 pandemic and the associated restrictions on movements. As a result, the corresponding smart apps have also seen increased usage, leading to a surge in downloads of video conferencing apps. However, this trend has generated several data protection and privacy challenges inherent in the smart mobile ecosystem. This paper aims to study data protection issues in video conferencing apps by statistically and dynamically analyzing the most common such issues in real-time operation on Android platforms. The goal is to determine what these applications do in real time and verify whether they provide users with sufficient information regarding the underlying personal data processes. Our results illustrate that there is still room for improvement in several aspects, mainly because the relevant privacy policies do not always provide users with sufficient information about the underlying personal data processes (especially with respect to data leaks to third parties), which, in turn, raises concerns about compliance with data protection by design and default principles. Specifically, users are often not informed about which personal data are being processed, for what purposes, and whether these processes are necessary (and, if yes, why) or based on their consent. Furthermore, the permissions required by the apps during runtime are not always justified. Full article
(This article belongs to the Special Issue Next Generation Networks and Systems Security)
►▼ Show Figures

Figure 1

Figure 1
<p>Data flow to ATS from Viber.</p> Full article ">
13 pages, 1069 KiB  
Article
Non-Linear Adapted Spatio-Temporal Filter for Single-Trial Identification of Movement-Related Cortical Potential
by Luca Mesin, Usman Ghani and Imran Khan Niazi
Electronics 2023, 12(5), 1246; https://doi.org/10.3390/electronics12051246 - 5 Mar 2023
Cited by 4 | Viewed by 1461
Abstract
The execution or imagination of a movement is reflected by a cortical potential that can be recorded by electroencephalography (EEG) as Movement-Related Cortical Potentials (MRCPs). The identification of MRCP from a single trial is a challenging possibility to get a natural control of [...] Read more.
The execution or imagination of a movement is reflected by a cortical potential that can be recorded by electroencephalography (EEG) as Movement-Related Cortical Potentials (MRCPs). The identification of MRCP from a single trial is a challenging possibility to get a natural control of a Brain–Computer Interface (BCI). We propose a novel method for MRCP detection based on optimal non-linear filters, processing different channels of EEG including delayed samples (getting a spatio-temporal filter). Different outputs can be obtained by changing the order of the temporal filter and of the non-linear processing of the input data. The classification performances of these filters are assessed by cross-validation on a training set, selecting the best ones (adapted to the user) and performing a majority voting from the best three to get an output using test data. The method is compared to another state-of-the-art filter recently introduced by our group when applied to EEG data recorded from 16 healthy subjects either executing or imagining 50 self-paced upper-limb palmar grasps. The new approach has a median accuracy on the overall dataset of 80%, which is significantly better than that of the previous filter (i.e., 63%). It is feasible for online BCI system design with asynchronous, self-paced applications. Full article
(This article belongs to the Special Issue Development and Validation of Innovative Low Cost Brain-Computer-Interfaces)
►▼ Show Figures

Figure 1

Figure 1
<p>(<b>A</b>) Visual cue provided to the participants. The template was displayed during the whole trial. A moving cursor was shown to help the subjects to follow the template. The output of the force transducer was used as moving cursor during motor execution; in case of motor imagination task, the cursor moved over the template to cue the subjects. (<b>B</b>) Sketchy representation of the location of the subject during the experiment. (<b>C</b>) Indication of the considered EEG electrodes.</p> Full article ">Figure 2
<p>Example of signals (second subject during motor execution). (<b>A</b>) Raw data (red color) and EEG after compensating for jumps (black). (<b>B</b>) EEG data (with jumps compensated) after band-pass filtering (red color) and artifacts removal (black).</p> Full article ">Figure 3
<p>Example of processing (same data as in <a href="#electronics-12-01246-f002" class="html-fig">Figure 2</a> are used). (<b>A</b>) Output of the Non-Linear Optimized Spatial Filter (NL-SF) on test data, after template matching (reference prototype indicating movements onset in red). (<b>B</b>) Outputs of the three best Non-Linear Spatio-Temporal Filters (NLSTF) on test data, after template matching. (<b>C</b>) Templates of the filters (NL-SF and the 3 best NLSTFs), obtained by averaging the outputs of the filters on MRCPs of the training set.</p> Full article ">Figure 4
<p>Histograms of orders of temporal filters (<b>A</b>,<b>B</b>) and order of polynomial non-linearity (<b>C</b>,<b>D</b>) of the filters selected to be applied on test data (majority voting on the outputs of three filters) for EEGs recorded during either motor execution (ME, panels <b>A</b>,<b>C</b>) or imagination (MI, <b>B</b>,<b>D</b>). NL-SF indicates the non-linear spatial filter, sometimes preferred over a polynomial non-linearity.</p> Full article ">Figure 5
<p>Performances on the testing set of the filters: non-linear spatial filter (NL-SF, gray color) and adapted non-linear spatio-temporal filter (NLSTF, black color). Accuracy, true positive rate (TPR) and false positive rate (FPR) are shown, considering either motor execution (ME) or imagination (MI). The default condition (i.e., nine EEG channels and 70% of training data) is compared with the reduction of either the number of channels (six instead of nine channels) or the size of the training set (40% of the MRCPs instead of the 70%). Box and whiskers plots are shown, indicating median, quartiles, range and outliers (using +markers). Statistical differences in paired comparisons are shown with marker * (<span class="html-italic">p</span> &lt; 0.05) or ** (<span class="html-italic">p</span> &lt; 0.01) and a segment joining the two tested distributions.</p> Full article ">
10 pages, 8297 KiB  
Communication
One-Step, In Situ Hydrothermal Fabrication of Cobalt-Doped ZnO/CdS Nanosheets for Optoelectronic Applications
by Lakshmiprasad Maddi, Khidhirbrahmendra Vinukonda, Thirumala Rao Gurugubelli and Ravindranadh Koutavarapu
Electronics 2023, 12(5), 1245; https://doi.org/10.3390/electronics12051245 - 5 Mar 2023
Cited by 4 | Viewed by 1531
Abstract
An in-situ hydrothermal process was used to create Co-doped ZnO/CdS nanosheets in order to examine the effects of the divalent impurity (Co) ions on the structural, morphological, optical, and magnetic characteristics of the test material. For both ZnO and CdS, XRD verified the [...] Read more.
An in-situ hydrothermal process was used to create Co-doped ZnO/CdS nanosheets in order to examine the effects of the divalent impurity (Co) ions on the structural, morphological, optical, and magnetic characteristics of the test material. For both ZnO and CdS, XRD verified the development of a hexagonal wurtzite structure. SEM, TEM, and HR-TEM studies produced sheet-like morphology. Elemental mapping and XPS examination verified the presence of essential elements (S, Cd, O, Co, and Zn). Co-doping dramatically increased the nanosheets’ ability to absorb light in the visible area. Comparing the bandgap energy to pure ZnO and ZnO/CdS nanocomposites, the bandgap energy (2.59 eV) was well-regulated. The PL spectrum at 577 nm showed a prominent yellow emission band that was attributed to the 4A2g(F) → 4T1g(F) transition. Improvement in the room temperature ferromagnetic properties was observed due to doping of Co2+ ions. Warm white light harvesting was confirmed by the estimated CCT value (3540 K). The test material appears to be suitable for the creation of next-generation optoelectronic devices. Full article
(This article belongs to the Special Issue Advanced Electronics and Optoelectronics Based on 2D Materials and vdW Heterostructures)
►▼ Show Figures

Figure 1

Figure 1
<p>X-ray diffraction pattern of Co-doped ZnO/CdS nanosheets.</p> Full article ">Figure 2
<p>(<b>a</b>) SEM, (<b>b</b>) TEM, (<b>c</b>) HRTEM, and (<b>d</b>) lattice fringe pattern of Co-doped ZnO/CdS nanosheets.</p> Full article ">Figure 3
<p>Elemental mapping of Co-doped ZnO/CdS nanosheets. (<b>a</b>) Mapping region, (<b>b</b>) Zn, (<b>c</b>) O, (<b>d</b>) Cd, (<b>e</b>) S, and (<b>f</b>) Co.</p> Full article ">Figure 4
<p>XPS spectra of Co-doped ZnO/CdS nanosheets. (<b>a</b>) Survey, (<b>b</b>) S 2p, (<b>c</b>) Cd 3d, (<b>d</b>) O 1s, (<b>e</b>) Co 2p, and (<b>f</b>) Zn 2p.</p> Full article ">Figure 5
<p>(<b>a</b>) DRS spectrum and (<b>b</b>) Tauc plot of Co-doped ZnO/CdS nanosheets.</p> Full article ">Figure 6
<p>Photoluminescence spectrum of Co-doped ZnO/CdS nanosheets.</p> Full article ">Figure 7
<p>Magnetic hysteresis curve of Co-doped ZnO/CdS nanosheets.</p> Full article ">Figure 8
<p>1931 CIE diagram of Co-doped ZnO/CdS nanosheets.</p> Full article ">
16 pages, 8756 KiB  
Communication
Research on Improved Wavelet Threshold Denoising Method for Non-Contact Force and Magnetic Signals
by Xiaoxiao Li, Kexi Liao, Guoxi He and Jianhua Zhao
Electronics 2023, 12(5), 1244; https://doi.org/10.3390/electronics12051244 - 4 Mar 2023
Cited by 5 | Viewed by 2481
Abstract
In order to solve the problem of noise interference in the collected magneto mechanical signals, a new wavelet shrinkage threshold based on adaptive estimation is proposed. Based on the shortcomings of the traditional threshold function, an improved threshold function is proposed, and the [...] Read more.
In order to solve the problem of noise interference in the collected magneto mechanical signals, a new wavelet shrinkage threshold based on adaptive estimation is proposed. Based on the shortcomings of the traditional threshold function, an improved threshold function is proposed, and the parameters of the threshold function are solved by the improved genetic algorithm, and the optimal denoising effect is finally obtained. The new threshold function can not only make up the defects of each threshold function, ensure the continuity of threshold function, but also flexibly adjust the threshold to adapt to different noise conditions, and solve the deviation caused by inherent threshold function, and protect the useful information with noise signals. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Signal reconstruction and decomposition process.</p> Full article ">Figure 2
<p>Non-contact 3D magnetic testing system. (<b>a</b>) The whole testing instrument; (<b>b</b>) Field operation; (<b>c</b>) data collector; (<b>d</b>) data analyzer.</p> Full article ">Figure 3
<p>Magnetic signal data of sensor in a 1 z direction.</p> Full article ">Figure 4
<p>SNR values under different wavelet basis functions.</p> Full article ">Figure 5
<p>RMSE under different wavelet basis functions.</p> Full article ">Figure 6
<p>Scaling function.</p> Full article ">Figure 7
<p>Wavelet function.</p> Full article ">Figure 8
<p>SNR under different decomposition levels.</p> Full article ">Figure 9
<p>RMSE with different decomposition levels.</p> Full article ">Figure 10
<p>Denoised signals with different decomposition levels. (<b>a</b>) 5 levels; (<b>b</b>) 6 levels; (<b>c</b>) 7 levels; (<b>d</b>) 8 levels; (<b>e</b>) 9 levels.</p> Full article ">Figure 11
<p>Detailed coefficient under different decomposition layers.</p> Full article ">Figure 12
<p>Approximate coefficients under different decomposition levels.</p> Full article ">Figure 13
<p>Constructed function curve.</p> Full article ">Figure 14
<p>Comparison of hard threshold, soft threshold, and improved threshold function.</p> Full article ">Figure 15
<p>SNR and RMSE curves under different parameter values (<b>a</b>) SNR values under different <math display="inline"><semantics> <mi>α</mi> </semantics></math>; (<b>b</b>) RMSE values under different <math display="inline"><semantics> <mi>α</mi> </semantics></math>; (<b>c</b>) SNR values under different <math display="inline"><semantics> <mi>β</mi> </semantics></math>; (<b>d</b>) RMSE values under different <math display="inline"><semantics> <mi>β</mi> </semantics></math>.</p> Full article ">Figure 16
<p>New wavelet threshold denoising process.</p> Full article ">Figure 17
<p>Comparison of signal processing results (<b>a</b>) hard threshold function; (<b>b</b>) soft threshold function.</p> Full article ">Figure 17 Cont.
<p>Comparison of signal processing results (<b>a</b>) hard threshold function; (<b>b</b>) soft threshold function.</p> Full article ">
15 pages, 3336 KiB  
Article
High-Performance Embedded System for Offline Signature Verification Problem Using Machine Learning
by Umair Tariq, Zonghai Hu, Rokham Tariq, Muhammad Shahid Iqbal and Muhammad Sadiq
Electronics 2023, 12(5), 1243; https://doi.org/10.3390/electronics12051243 - 4 Mar 2023
Cited by 1 | Viewed by 2615
Abstract
This paper proposes a high-performance embedded system for offline Urdu handwritten signature verification. Though many signature datasets are publicly available in languages such as English, Latin, Chinese, Persian, Arabic, Hindi, and Bengali, no Urdu handwritten datasets were available in the literature. So, in [...] Read more.
This paper proposes a high-performance embedded system for offline Urdu handwritten signature verification. Though many signature datasets are publicly available in languages such as English, Latin, Chinese, Persian, Arabic, Hindi, and Bengali, no Urdu handwritten datasets were available in the literature. So, in this work, an Urdu handwritten signature dataset is created. The proposed embedded system is then used to distinguish genuine and forged signatures based on various features, such as length, pattern, and edges. The system consists of five steps: data acquisition, pre-processing, feature extraction, signature registration, and signature verification. A majority voting (MV) algorithm is used for improved performance and accuracy of the proposed embedded system. In feature extraction, an improved sinusoidal signal multiplied by a Gaussian function at a specific frequency and orientation is used as a 2D Gabor filter. The proposed framework is tested and compared with existing handwritten signature verification methods. Our test results show accuracies of 66.8% for ensemble, 86.34% for k-nearest neighbor (KNN), 93.31% for support vector machine (SVM), and 95.05% for convolutional neural network (CNN). After applying the majority voting algorithm, the overall accuracy can be improved to 95.13%, with a false acceptance rate (FAR) of 0.2% and a false rejection rate (FRR) of 41.29% on private dataset. To test the generalization ability of the proposed model, we also test it on a public dataset of English handwritten signatures and achieve an overall accuracy of 97.46%. Full article
(This article belongs to the Special Issue High-Performance Embedded Computing)
►▼ Show Figures

Figure 1

Figure 1
<p>Embedded System for UHSV.</p> Full article ">Figure 2
<p>(<b>a</b>,<b>c</b>) after and (<b>b</b>,<b>d</b>) before pre-processing the image.</p> Full article ">Figure 3
<p>Accuracy of k-nearest neighbor (KNN) in different models.</p> Full article ">Figure 4
<p>Accuracy of support vector machine (SVM) classifier in different models.</p> Full article ">Figure 5
<p>Bagging Loss of Training.</p> Full article ">Figure 6
<p>Overall accuracy of the ensemble classifier.</p> Full article ">Figure 7
<p>CNN Model for Offline Signature Verification.</p> Full article ">Figure 8
<p>Training and Testing of CNN Model.</p> Full article ">Figure 9
<p>Comparison of the proposed method with existing published methods.</p> Full article ">Figure 10
<p>Confusion Matrix of KNN, CNN, and SVM Classifiers.</p> Full article ">Figure 11
<p>Comparison of FAR and FRR.</p> Full article ">
21 pages, 3292 KiB  
Article
A New Social Media-Driven Cyber Threat Intelligence
by Fahim Sufi
Electronics 2023, 12(5), 1242; https://doi.org/10.3390/electronics12051242 - 4 Mar 2023
Cited by 8 | Viewed by 5159
Abstract
Cyber threats are projected to cause USD 10.5 trillion in damage to the global economy in 2025. Comprehending the level of threat is core to adjusting cyber posture at the personal, organizational, and national levels. However, representing the threat level with a single [...] Read more.
Cyber threats are projected to cause USD 10.5 trillion in damage to the global economy in 2025. Comprehending the level of threat is core to adjusting cyber posture at the personal, organizational, and national levels. However, representing the threat level with a single score is a daunting task if the scores are generated from big and complex data sources such as social media. This paper harnesses the modern technological advancements in artificial intelligence (AI) and natural language processing (NLP) to comprehend the contextual information of social media posts related to cyber-attacks and electronic warfare. Then, using keyword-based index generation techniques, a single index is generated at the country level. Utilizing a convolutional neural network (CNN), the innovative process automatically detects any anomalies within the countrywide threat index and explains the root causes. The entire process was validated with live Twitter feeds from 14 October 2022 to 27 December 2022. During these 75 days, AI-based language detection, translation, and sentiment analysis comprehended 15,983 tweets in 47 different languages (while most of the existing works only work in one language). Finally, 75 daily cyber threat indexes with anomalies were generated for China, Australia, Russia, Ukraine, Iran, and India. Using this intelligence, strategic decision makers can adjust their cyber preparedness for mitigating the detrimental damages afflicted by cyber criminals. Full article
(This article belongs to the Special Issue Machine Learning (ML) and Software Engineering)
►▼ Show Figures

Figure 1

Figure 1
<p>Overview of the design for a new cyber index system by a systematic literature review, finding challenges, developing requirements, and finally, designing the solution.</p> Full article ">Figure 2
<p>Conceptual diagram of deep learning-based, countrywide cyber index implementation.</p> Full article ">Figure 3
<p>CNN-based anomaly detection executed on cyber threat indexes within a Windows environment.</p> Full article ">Figure 4
<p>Social media-based cyber threat dashboard executed on a Samsung Galaxy Note 10 Lite Mobile (Android Device).</p> Full article ">Figure 5
<p>Social media-based cyber threat dashboard executed on an Apple iPad 9th Generation (iOS Device).</p> Full article ">
13 pages, 774 KiB  
Article
A Next POI Recommendation Based on Graph Convolutional Network by Adaptive Time Patterns
by Jiang Wu, Shaojie Jiang and Lei Shi
Electronics 2023, 12(5), 1241; https://doi.org/10.3390/electronics12051241 - 4 Mar 2023
Cited by 3 | Viewed by 2039
Abstract
Users’ activities in location-based social networks (LBSNs) can be naturally transformed into graph structural data, and more advanced graph representation learning techniques can be adopted for analyzing user preferences, which benefits a variety of real-world applications. This paper focuses on the next point-of-interest [...] Read more.
Users’ activities in location-based social networks (LBSNs) can be naturally transformed into graph structural data, and more advanced graph representation learning techniques can be adopted for analyzing user preferences, which benefits a variety of real-world applications. This paper focuses on the next point-of-interest (POI) recommendation task in LBSNs. We argue that existing graph-based POI recommendation methods only consider user preferences from several individual contextual factors, ignoring the influence of interactions between different contextual information. This practice leads to the suboptimal learning of user preferences. To address this problem, we propose a novel method called hierarchical attention-based graph convolutional network (HAGCN) for the next POI recommendation, a technique which leverages graph convolutional networks to extract the representations of POIs from predefined graphs via different time patterns and develops a hierarchical attention mechanism to adaptively learn user preferences from the interactions between different contextual data. Moreover, HAGCN uses a dynamic preference estimation to precisely learn user preferences. We conduct extensive experiments on real-world datasets to evaluate the performance of HAGCN against representative baseline models in the field of next POI recommendation. The experimental results demonstrate the superiority of our proposed method on the next POI recommendation task. Full article
(This article belongs to the Special Issue Mechanism and Modeling of Graph Convolutional Networks)
►▼ Show Figures

Figure 1

Figure 1
<p>The overall of HAGCN.</p> Full article ">Figure 2
<p>The influence of GCN layers on NYC dataset.</p> Full article ">Figure 3
<p>The influence of hidden dimensions on NYC dataset.</p> Full article ">
21 pages, 1957 KiB  
Article
Software-Defined Radio Implementation and Performance Evaluation of Frequency-Modulated Antipodal Chaos Shift Keying Communication System
by Arturs Aboltins and Nikolajs Tihomorskis
Electronics 2023, 12(5), 1240; https://doi.org/10.3390/electronics12051240 - 4 Mar 2023
Cited by 4 | Viewed by 3697
Abstract
This paper is devoted to software-defined radio (SDR) implementation of frequency modulated antipodal chaos shift keying (FM-ACSK) transceiver and presents results of prototype testing in real conditions. This novel and perspective class of spread-spectrum communication systems employs chaotic synchronization for the acquisition and [...] Read more.
This paper is devoted to software-defined radio (SDR) implementation of frequency modulated antipodal chaos shift keying (FM-ACSK) transceiver and presents results of prototype testing in real conditions. This novel and perspective class of spread-spectrum communication systems employs chaotic synchronization for the acquisition and tracking of the analog chaotic spreading code and does not need resource-demanding cross-correlation. The main motivation of the given work is to assess the performance of FM-ACSK in real conditions and demonstrate that chaotic synchronization can be considered an efficient spread-spectrum demodulation method. The work focuses on the real-time implementation aspects of the modulation-demodulation algorithms, forward error correction (FEC) and symbol timing synchronization approach in MATLAB Simulink. The performance of the presented prototype is assessed via extensive testing, which includes measurement of bit error ratio (BER) in single-user and multi-user scenarios, estimation of carrier frequency offset (CFO) impact and image transmission over-the-air between two independent sites and comparison with classical frequency hopping spread spectrum (FHSS). The paper shows that the presented class of the spread spectrum communication systems demonstrates good performance in low signal-to-noise ratio (SNR) conditions and in terms of BER significantly outperforms the classic spread-spectrum modulation schemes which employ correlation-based detection. Full article
(This article belongs to the Special Issue Electronic Systems with Dynamic Chaos: Design and Applications)
►▼ Show Figures

Graphical abstract

Graphical abstract
Full article ">Figure 1
<p>Master chaos generator’s output signal <math display="inline"><semantics> <msub> <mi>R</mi> <mrow> <mi>o</mi> <mi>u</mi> <mi>t</mi> </mrow> </msub> </semantics></math>.</p> Full article ">Figure 2
<p>Block diagram of master and slave chaos generators.</p> Full article ">Figure 3
<p>Block diagram of SDR-based FM-ACSK transceiver.</p> Full article ">Figure 4
<p>Spreading of the data bits using a continuous chaotic waveform. Transmitted data bits are drawn in the upper graph, switched chaotic waveform-lower graph. Each data bit is spread with approximately 9 oscillations of the chaotic waveform.</p> Full article ">Figure 5
<p>Baseband spectrum of the ACSK waveform (spread data bits).</p> Full article ">Figure 6
<p>Baseband spectrum of the FM-ACSK at 10 KHZ FM deviation.</p> Full article ">Figure 7
<p>Implementation of chaotic generator in the ACSK modulator.</p> Full article ">Figure 8
<p>Implementation of the chaotic generator and synchronization circuit in the ACSK demodulator.</p> Full article ">Figure 9
<p>Signal at the input of decision-making unit.</p> Full article ">Figure 10
<p>Implementation of the symbol timing (ST) unit.</p> Full article ">Figure 11
<p>BER versus SNR for data transmission with (1023, 1013) and (7, 4) Hamming codes in FM-ACSK system. SD—strobe-based detector; ED—energy-based detector. FM deviation is 10 <math display="inline"><semantics> <mrow> <mi mathvariant="normal">k</mi> <mi>Hz</mi> </mrow> </semantics></math>.</p> Full article ">Figure 12
<p>BER versus transmit power in <math display="inline"><semantics> <mi>dBm</mi> </semantics></math> with different bit detectors in a single model scenario. SD—strobe-based detector; ED—energy-based detector; DEV—FM deviation. Receiver gain is 10 dB.</p> Full article ">Figure 13
<p>Measured power spectra at the output of Adalm Pluto. DEV—FM deviation. The transmitted power of Adalm Pluto is −26 <math display="inline"><semantics> <mi>dBm</mi> </semantics></math>, and the receiver gain is 10 dB.</p> Full article ">Figure 14
<p>BER versus relative CFO with the strobe-based detector in the single model scenario. DEV—FM deviation. The receiver gain is 10 dB, and the transmitter gain is −26 <math display="inline"><semantics> <mi>dBm</mi> </semantics></math>.</p> Full article ">Figure 15
<p>BER versus transmit power of the malicious transmitter. DEV—FM deviation. The receiver gain is 10 dB and the first transmitter power is –36 <math display="inline"><semantics> <mi>dBm</mi> </semantics></math>. The receiver uses a strobe-based detector.</p> Full article ">Figure 16
<p>Received images after transmission between two computers. (<b>a</b>) Transmitted image. (<b>b</b>) Received image using FM-ACSK without a Hamming code. (<b>c</b>) Received image using FM-ACSK with a Hamming code.</p> Full article ">Figure 17
<p>Auto-correlation of used spreading sequence.</p> Full article ">Figure 18
<p>BER versus SNR for data transmission with FM-ACSK and FHSS. SD—strobe-based detector; ED—energy-based detector. FM deviation is 10 <math display="inline"><semantics> <mrow> <mi mathvariant="normal">k</mi> <mi>Hz</mi> </mrow> </semantics></math>.</p> Full article ">Figure A1
<p>Creation of data bits from the image where <span class="html-italic">n</span> is image pixel row length, <span class="html-italic">m</span> is image pixel column length, and <span class="html-italic">i</span> is frame or image column number. (<b>a</b>) Image matrix with decimal values. (<b>b</b>) Image matrix reshaped into column vector with decimal values. (<b>c</b>) Matrix of each image pixel (matrix row) and its binary value in matrix columns. (<b>d</b>) Data row containing the image in bits where black lines show a start of an image pixel column.</p> Full article ">
22 pages, 1343 KiB  
Article
Unmanned-Aircraft-System-Assisted Early Wildfire Detection with Air Quality Sensors
by Doaa Rjoub, Ahmad Alsharoa and Ala’eddin Masadeh
Electronics 2023, 12(5), 1239; https://doi.org/10.3390/electronics12051239 - 4 Mar 2023
Cited by 5 | Viewed by 2415
Abstract
Numerous hectares of land are destroyed by wildfires every year, causing harm to the environment, the economy, and the ecology. More than fifty million acres have burned in several states as a result of recent forest fires in the Western United States and [...] Read more.
Numerous hectares of land are destroyed by wildfires every year, causing harm to the environment, the economy, and the ecology. More than fifty million acres have burned in several states as a result of recent forest fires in the Western United States and Australia. According to scientific predictions, as the climate warms and dries, wildfires will become more intense and frequent, as well as more dangerous. These unavoidable catastrophes emphasize how important early wildfire detection and prevention are. The energy management system described in this paper uses an unmanned aircraft system (UAS) with air quality sensors (AQSs) to monitor spot fires before they spread. The goal was to develop an efficient autonomous patrolling system that detects early wildfires while maximizing the battery life of the UAS to cover broad areas. The UAS will send real-time data (sensor readings, thermal imaging, etc.) to a nearby base station (BS) when a wildfire is discovered. An optimization model was developed to minimize the total amount of energy used by the UAS while maintaining the required levels of data quality. Finally, the simulations showed the performance of the proposed solution under different stability conditions and for different minimum data rate types. Full article
(This article belongs to the Special Issue Advanced Technologies in Unmanned Aerial Vehicle (UAV)-Based Communications)
►▼ Show Figures

Figure 1

Figure 1
<p>System model.</p> Full article ">Figure 2
<p>A Gaussian dispersion model for estimating the levels of air contaminants.</p> Full article ">Figure 3
<p>A graphical representation of the problem under consideration. The UAS patrol algorithm should be optimized to find the optimal flying path for detecting wildfires.</p> Full article ">Figure 4
<p>The profile PM concentrations are displayed along the y axis in the Gaussian dispersion model at various points (x) downstream of the emission site at altitudes (z) of (<b>a</b>) 50 m and (<b>b</b>) 100 m.</p> Full article ">Figure 5
<p>The 2D Gaussian pollutant concentration dispersion model in a plume for the slightly stable situation. The gray bar represents the effective height of the emission point.</p> Full article ">Figure 6
<p>The 2D Gaussian pollutant concentration dispersion model in a plume for the very/extremely unstable situation. The gray bar represents the effective height of the emission point.</p> Full article ">Figure 7
<p>The 2D Gaussian pollutant concentration dispersion model in a plume for the neutral stability situation. The gray bar represents the effective height of the emission point.</p> Full article ">Figure 8
<p>The horizontal gap for various stability situations as a function of wind speed.</p> Full article ">Figure 9
<p>The horizontal gap versus the emission rate factor for various stability situations.</p> Full article ">Figure 10
<p>The impact of the PM2.5 emission rate on the horizontal gap while keeping the CO emission rate constant. The black arrow indicates the chosen optimal value of the horizontal gap when the PM2.5 pollutant’s rate factor is changed while the CO pollutant’s rate remains constant.</p> Full article ">Figure 11
<p>The impact of the CO emission rate on the horizontal gap while keeping the PM2.5 emission rate constant. The black arrow indicates the chosen optimal value of the horizontal gap when the CO pollutant’s rate factor is changed while the PM2.5 pollutant’s rate remains constant.</p> Full article ">Figure 12
<p>The area covered by the UAS as a function of the horizontal gap for various UAS battery limitations.</p> Full article ">Figure 13
<p>Average data rate attained in relation to UAS transmit power budget.</p> Full article ">Figure 14
<p>Loss tolerance as a function of UAS transmit power budget.</p> Full article ">
34 pages, 1856 KiB  
Article
Security Quantification of Container-Technology-Driven E-Government Systems
by Subrota Kumar Mondal, Tian Tan, Sadia Khanam, Keshav Kumar, Hussain Mohammed Dipu Kabir and Kan Ni
Electronics 2023, 12(5), 1238; https://doi.org/10.3390/electronics12051238 - 4 Mar 2023
Cited by 5 | Viewed by 2120
Abstract
With the rapidly increasing demands of e-government systems in smart cities, a myriad of challenges and issues are required to be addressed. Among them, security is one of the prime concerns. To this end, we analyze different e-government systems and find that an [...] Read more.
With the rapidly increasing demands of e-government systems in smart cities, a myriad of challenges and issues are required to be addressed. Among them, security is one of the prime concerns. To this end, we analyze different e-government systems and find that an e-government system built with container-based technology is endowed with many features. In addition, overhauling the architecture of container-technology-driven e-government systems, we observe that securing an e-government system demands quantifying security issues (vulnerabilities, threats, attacks, and risks) and the related countermeasures. Notably, we find that the Attack Tree and Attack–Defense Tree methods are state-of-the-art approaches in these aspects. Consequently, in this paper, we work on quantifying the security attributes, measures, and metrics of an e-government system using Attack Trees and Attack–Defense Trees—in this context, we build a working prototype of an e-government system aligned with the United Kingdom (UK) government portal, which is in line with our research scope. In particular, we propose a novel measure to quantify the probability of attack success using a risk matrix and normal distribution. The probabilistic analysis distinguishes the attack and defense levels more intuitively in e-government systems. Moreover, it infers the importance of enhancing security in e-government systems. In particular, the analysis shows that an e-government system is fairly unsafe with a 99% probability of being subject to attacks, and even with a defense mechanism, the probability of attack lies around 97%, which directs us to pay close attention to e-government security. In sum, our implications can serve as a benchmark for evaluation for governments to determine the next steps in consolidating e-government system security. Full article
(This article belongs to the Special Issue Big Data and Cloud Computing: Innovations and Challenges)
►▼ Show Figures

Figure 1

Figure 1
<p>Attack–Defense Tree structure.</p> Full article ">Figure 2
<p>Architecture of a container-technology-driven e-government system.</p> Full article ">Figure 3
<p>Kubernetes cluster architecture [<a href="#B78-electronics-12-01238" class="html-bibr">78</a>].</p> Full article ">Figure 4
<p>Normal distribution [<a href="#B85-electronics-12-01238" class="html-bibr">85</a>,<a href="#B86-electronics-12-01238" class="html-bibr">86</a>].</p> Full article ">Figure 5
<p>Attack Tree for e-government infrastructure with cloud computing.</p> Full article ">Figure 6
<p>Probabilities in an Attack Tree for e-government infrastructure with cloud computing.</p> Full article ">Figure 7
<p>Attack–Defense Tree for e-government infrastructure with cloud computing.</p> Full article ">Figure 8
<p>Probabilities in an Attack–Defense Tree for e-government infrastructure with cloud computing.</p> Full article ">Figure 9
<p>Ideal probabilities in the ADTree for e-government introducing cloud computing.</p> Full article ">Figure 10
<p>Attack Tree for attacking the KubeAPI server.</p> Full article ">Figure 11
<p>Probabilities in the Attack Tree for attacking the KubeAPI server.</p> Full article ">Figure 12
<p>Attack–Defense Tree for attacking the KubeAPI server.</p> Full article ">Figure 13
<p>Probabilities in the Attack–Defense Tree for attacking the KubeAPI server.</p> Full article ">Figure 14
<p>Attack Tree for e-government portals.</p> Full article ">Figure 15
<p>Probabilities in the Attack Tree for e-government portals.</p> Full article ">Figure 16
<p>Attack–Defense Tree for e-government portals.</p> Full article ">Figure 17
<p>Probabilities in the Attack–Defense Tree for e-government portals.</p> Full article ">Figure 18
<p>ADTree for the access layer.</p> Full article ">
21 pages, 30012 KiB  
Article
SODAS: Smart Open Data as a Service for Improving Interconnectivity and Data Usability
by Heesun Won, Jiwoo Han, Myeong-Seon Gil and Yang-Sae Moon
Electronics 2023, 12(5), 1237; https://doi.org/10.3390/electronics12051237 - 4 Mar 2023
Cited by 2 | Viewed by 1823
Abstract
In this study, we proposed Smart Open Data as a Service (SODAS) as a new open data platform based on the international standards Data Catalog Vocabulary (DCAT) and Comprehensive Knowledge Archive Network (CKAN) to facilitate the release and sharing of data. We first [...] Read more.
In this study, we proposed Smart Open Data as a Service (SODAS) as a new open data platform based on the international standards Data Catalog Vocabulary (DCAT) and Comprehensive Knowledge Archive Network (CKAN) to facilitate the release and sharing of data. We first analyze the five problems in the legacy CKAN and then draw up corresponding solutions through three core strategies: CKAN expansion, DCATv2 support, and extendable DataMap. We then define four components and nine function blocks of SODAS for each core strategy. As a result, SODAS drives Open Data Portal, Open Data Reference Model, DataMap Publisher, and Analytics and Development Environment (ADE) Provisioning for connecting the defined function blocks. We confirm that each function works correctly through the SODAS Web portal, and then we apply SODAS to actual data distribution sites to prove its efficiency and practical use. SODAS is the first open data platform that provides secure interoperability between heterogeneous platforms based on international standards, and it enables domain-free data management with flexible metadata. Full article
(This article belongs to the Special Issue Multi-Service Cloud-Based IoT Platforms)
►▼ Show Figures

Figure 1

Figure 1
<p>Basic operations of CKAN.</p> Full article ">Figure 2
<p>Structure comparison of DCATv1 and DCATv2.</p> Full article ">Figure 3
<p>An example of DCATv2-based profile management.</p> Full article ">Figure 4
<p>The overall framework of SODAS.</p> Full article ">Figure 5
<p>Relationship among SODAS major components and function blocks.</p> Full article ">Figure 6
<p>Menu layout of Open Data Portal.</p> Full article ">Figure 7
<p>Main operations of Open Data Reference Model.</p> Full article ">Figure 8
<p>Harvesting operation of DataMap Publisher.</p> Full article ">Figure 9
<p>Operation of metadata conversion tool.</p> Full article ">Figure 10
<p>Operational structure of ADE Provisioning.</p> Full article ">Figure 11
<p>Main screen of SODAS data portal.</p> Full article ">Figure 12
<p>Functions of the Category menu.</p> Full article ">Figure 13
<p>Example screens of the Analytics and Service menus.</p> Full article ">Figure 14
<p>Example screen of catalog mapping using DataMap Publisher.</p> Full article ">Figure 15
<p>Example screen of ADE Provisioning main menu.</p> Full article ">Figure 16
<p>SODAS-based actual operation services.</p> Full article ">Figure 17
<p>Working scenarios of PartnerHub applications.</p> Full article ">
17 pages, 22335 KiB  
Review
A Synthesis of Algorithms Determining a Safe Trajectory in a Group of Autonomous Vehicles Using a Sequential Game and Neural Network
by Józef Lisowski
Electronics 2023, 12(5), 1236; https://doi.org/10.3390/electronics12051236 - 4 Mar 2023
Cited by 4 | Viewed by 1480
Abstract
This paper presents a solution to the problem of providing an autonomous vehicle with a safe control task when moving around many other autonomous vehicles. This is achieved by developing an appropriate computer control algorithm that takes into account the possible risk of [...] Read more.
This paper presents a solution to the problem of providing an autonomous vehicle with a safe control task when moving around many other autonomous vehicles. This is achieved by developing an appropriate computer control algorithm that takes into account the possible risk of a collision resulting from both the impact of environmental disturbances and the imperfection of the rules of maneuvering in situations where many vehicles pass each other, giving the control process a decisive character. For this purpose, three types of algorithms were synthesized: kinematic and dynamic optimization with neural domains, as well as sequential game control of an autonomous vehicle. The control algorithms determine a safe trajectory, which is implemented by the actuators of the autonomous vehicle. Computer simulations of the control algorithms in the Matlab/Simulink software allow for their comparative analysis in terms of meeting the criteria for the optimality and safety of an autonomous vehicle when passing a larger number of other autonomous vehicles. For this purpose, scenarios of multidirectional and one-way traffic of autonomous vehicles were used. Full article
(This article belongs to the Special Issue Feature Papers in Electrical and Autonomous Vehicles)
►▼ Show Figures

Figure 1

Figure 1
<p>The structure of a group of autonomous vehicles in a hierarchical control system.</p> Full article ">Figure 2
<p>Imaging of the traffic situation of a group of autonomous vehicles: σ<sub>0</sub>, χ<sub>0</sub>—speed and course of our autonomous vehicle; σ<sub>n</sub>, χ<sub>n</sub>—speed and course of the n-th other autonomous vehicle, where n = 1,…, N; φ<sub>n</sub>, δ<sub>n</sub>—bearing and distance to the n-th other autonomous vehicle; X<sub>0</sub>, Y<sub>0</sub>—position coordinates of our autonomous vehicle; X<sub>n</sub>, Y<sub>n</sub>—position coordinates of the n-th other autonomous vehicle.</p> Full article ">Figure 3
<p>Graphic design of the area of acceptable maneuvers (green color) and prohibited maneuvers (red color) to move at a safe passing distance δ<sub>s</sub> in an autonomous vehicles’ group.</p> Full article ">Figure 4
<p>Block diagram of the KO algorithm for determining a safe kinematic trajectory in a group of autonomous vehicles.</p> Full article ">Figure 5
<p>Block diagram of the DONN algorithm for determining a safe dynamic trajectory in a group of autonomous vehicles.</p> Full article ">Figure 6
<p>Flow chart of procedure 1 a computer program representing the computation of a hexagonal domain.</p> Full article ">Figure 7
<p>Block diagram of the GO algorithm for determining a safe game trajectory in a group of autonomous vehicles.</p> Full article ">Figure 8
<p>Safe kinematic trajectory of our autonomous vehicle in multidirectional traffic while passing seven other autonomous vehicles for three safe passing distance values.</p> Full article ">Figure 9
<p>Safe dynamic trajectory of our autonomous vehicle in multidirectional traffic while passing seven other autonomous vehicles for three safe passing distance values.</p> Full article ">Figure 10
<p>Safe game trajectory of our autonomous vehicle in multidirectional traffic when passing seven other non-cooperating autonomous vehicles for three safe passing distance values.</p> Full article ">Figure 11
<p>Safe game trajectory of our autonomous vehicle in multidirectional traffic when passing seven other cooperating autonomous vehicles for three safe passing distance values.</p> Full article ">Figure 12
<p>Safe kinematic trajectory of our autonomous vehicle in one-way traffic while passing seven other autonomous vehicles for three safe passing distance values.</p> Full article ">Figure 13
<p>Safe dynamic trajectory of our autonomous vehicle in one-way traffic while passing seven other autonomous vehicles for three safe passing distance values.</p> Full article ">Figure 14
<p>Safe game trajectory of our autonomous vehicle in one-way traffic when passing seven other non-cooperating autonomous vehicles for three safe passing distance values.</p> Full article ">Figure 15
<p>Safe game trajectory of our autonomous vehicle in one-way traffic when passing seven other cooperating autonomous vehicles for three safe passing distance values.</p> Full article ">Figure 16
<p>Comparison of the determined trajectories of the autonomous vehicles in multidirectional traffic according to the KO, DONN, and GO algorithms for three safe passing distance values.</p> Full article ">Figure 17
<p>Comparison of the determined trajectories of the autonomous vehicles in one-way traffic according to the KO, DONN, and GO algorithms for three safe passing distance values.</p> Full article ">Figure 18
<p>Values of the final deviation of the determined trajectories from their initial values for autonomous vehicles in multidirectional and one-way traffic depending on the safe passing distance for the KO, DONN, and GO algorithms.</p> Full article ">
21 pages, 667 KiB  
Article
Towards High-Performance Supersingular Isogeny Cryptographic Hardware Accelerator Design
by Guantong Su and Guoqiang Bai
Electronics 2023, 12(5), 1235; https://doi.org/10.3390/electronics12051235 - 4 Mar 2023
Cited by 1 | Viewed by 1480
Abstract
Cryptosystems based on supersingular isogeny are a novel tool in post-quantum cryptography. One compelling characteristic is their concise keys and ciphertexts. However, the performance of supersingular isogeny computation is currently worse than that of other schemes. This is primarily due to the following [...] Read more.
Cryptosystems based on supersingular isogeny are a novel tool in post-quantum cryptography. One compelling characteristic is their concise keys and ciphertexts. However, the performance of supersingular isogeny computation is currently worse than that of other schemes. This is primarily due to the following factors. Firstly, the underlying field is a quadratic extension of the finite field, resulting in higher computational complexity. Secondly, the strategy for large-degree isogeny evaluation is complex and dependent on the elementary arithmetic units employed. Thirdly, adapting the same hardware to different parameters is challenging. Considering the evolution of similar curve-based cryptosystems, we believe proper algorithm optimization and hardware acceleration will reduce its speed overhead. This paper describes a high-performance and flexible hardware architecture that accelerates isogeny computation. Specifically, we optimize the design by creating a dedicated quadratic Montgomery multiplier and an efficient scheduling strategy that are suitable for supersingular isogeny. The multiplier operates on Fp2 under projective coordinate formulas, and the scheduling is tailored to it. By exploiting additional parallelism through replicated multipliers and concurrent isogeny subroutines, our 65 nm SMIC technology cryptographic accelerator can generate ephemeral public keys in 2.40 ms for Alice and 2.79 ms for Bob with a 751-bit prime setting. Sharing the secret key costs another 2.04 ms and 2.35 ms, respectively. Full article
►▼ Show Figures

Figure 1

Figure 1
<p>Portion of the 2-isogeny graph for <math display="inline"><semantics> <mrow> <mi>p</mi> <mo>=</mo> <mn>431</mn> </mrow> </semantics></math>.</p> Full article ">Figure 2
<p>High-level SIDH illustration.</p> Full article ">Figure 3
<p>Underlying operations for supersingular isogeny cryptography and ECC.</p> Full article ">Figure 4
<p>Strategies for performing large degree isogeny evaluation comprised of 6 small degree isogenies. Horizontal lines represent <span class="html-italic">l</span>-isogeny evaluations of points. Vertical lines stand for <span class="html-italic">l</span> scalar point multiplications. (<b>a</b>) Multiplication-based strategy, (<b>b</b>) isogeny-based strategy, (<b>c</b>) optimal serial strategy with a more expensive point multiplication, and (<b>d</b>) optimal serial strategy with a more costly isogeny evaluation.</p> Full article ">Figure 5
<p>Schematic of quadratic field modular multiplier.</p> Full article ">Figure 6
<p>Schematic of quadratic field modular adder.</p> Full article ">
25 pages, 2636 KiB  
Article
The BciAi4SLA Project: Towards a User-Centered BCI
by Cristina Gena, Dize Hilviu, Giovanni Chiarion, Silvestro Roatta, Francesca M. Bosco, Andrea Calvo, Claudio Mattutino and Stefano Vincenzi
Electronics 2023, 12(5), 1234; https://doi.org/10.3390/electronics12051234 - 4 Mar 2023
Cited by 5 | Viewed by 2222
Abstract
The brain–computer interfaces (BCI) are interfaces that put the user in communication with an electronic device based on signals originating from the brain. In this paper, we describe a proof of concept that took place within the context of BciAi4Sla, a multidisciplinary project [...] Read more.
The brain–computer interfaces (BCI) are interfaces that put the user in communication with an electronic device based on signals originating from the brain. In this paper, we describe a proof of concept that took place within the context of BciAi4Sla, a multidisciplinary project involving computer scientists, physiologists, biomedical engineers, neurologists, and psychologists with the aim of designing and developing a BCI system following a user-centered approach, involving domain experts and users since initial prototyping steps in a design–test–redesign development cycle. The project intends to develop a software platform able to restore a communication channel in patients who have compromised their communication possibilities due to illness or accidents. The most common case is the patients with amyotrophic lateral sclerosis (ALS). In this paper, we describe the background and the main development steps of the project, also reporting some initial and promising user evaluation results, including real-time performance classification and a proof-of-concept prototype. Full article
(This article belongs to the Special Issue Development and Validation of Innovative Low Cost Brain-Computer-Interfaces)
►▼ Show Figures

Figure 1

Figure 1
<p>Calibration phase: In the calibration phase, the user watches a distant monitor, waiting for instructions. After 2 s, a left or right hand is displayed for 3 s and then removed. During this time, the user has to imagine moving their corresponding hand and then rest for 2 s. This cycle is repeated N times, with a randomic 0.1–0.8-s extension of the resting interval. Only a 2.5-s epoch of EEG signal is considered for further processing (labeled in green). Each trial, along with its ground truth label (right/left), is used to construct the calibration CSP matrix from which a set of features is extracted and used to train an SVM classifier.</p> Full article ">Figure 2
<p>Testing phase: In this phase, experimental protocol and data collection are identical to the training phase. However, the collected EEG epoch is now transformed with the CSP matrix obtained in the calibration phase to obtain a set of features that are used to classify the current trial as left or right movement.</p> Full article ">Figure 3
<p>Home page.</p> Full article ">Figure 4
<p>BciAi4Sla sections: communication, activities, health, music, literature.</p> Full article ">Figure 5
<p>The communication section: keyboard and phrasebook.</p> Full article ">Figure 6
<p>The phrasebook, showing Yes, No, I don’t know, Hello, Thank you; Keyboard.</p> Full article ">Figure 7
<p>The keyboard.</p> Full article ">
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-196
Previous Issue
Next Issue
Back to TopTop