Search Results (27)

Silicon (Si) and silicon carbide (SiC) are second- and third-generation semiconductor materials with excellent properties that are particularly suitable for applications in scenarios such as high temperature, high voltage, and high frequency. Si/SiC wafers face warpage and bending problems during production, which can seriously affect subsequent processing. Fast, accurate, and comprehensive detection of thickness, thickness variation, and flatness (including bow and warpage) of SiC and Si wafers is an industry-recognized challenge. Frequency scanning interferometry (FSI) can synchronize the upper and lower surfaces and thickness information of transparent parallel thin wafers, but it is still affected by multiple interfacial harmonic reflections, reflectivity asymmetry, and phase modulation uncertainty when measuring SiC thin wafers, which leads to thickness calculation errors and face reconstruction deviations. To this end, this paper proposes a high-precision facet reconstruction method for SiC/Si structures, which combines harmonic spectral domain decomposition, refractive index gradient constraints, and partitioning optimization strategy, and introduces interferometric signal “oversampling” and weighted fusion of multiple sets of data to effectively suppress higher-order harmonic interferences, and to enhance the accuracy of phase resolution. The multi-layer iterative optimization model further enhances the measurement accuracy and robustness of the system. The flatness measurement system constructed based on this method can realize the simultaneous acquisition of three-dimensional top and bottom surfaces on 6-inch Si/SiC wafers, and accurately reconstruct the key parameters, such as flatness, warpage, and thickness variation (TTV). A comparison with the Corning Tropel FlatMaster commercial system shows that this method has high consistency and good applicability. Full article

With the widespread use of encryption protocols on network data, fast and effective encryption traffic classification can improve the efficiency of traffic analysis. A resampling method combining Wasserstein GAN and random selection is proposed for solving the dataset imbalance problem, and it uses Wasserstein GAN for oversampling and random selection for undersampling to achieve class equalization. Based on Mamba, an ultra-low parametric quantity model, we propose an encrypted traffic classification model, ET-Mamba, which has a pre-training phase and a fine-tuning phase. During the pre-training phase, positional embedding is used to characterize the blocks of the traffic grayscale image, and random masking is used to strengthen the learning of the intrinsic correlation among the blocks of the traffic grayscale image. During the fine-tuning phase, the agent attention mechanism is adopted in the feature extraction phase to achieve global information modeling at a low computational cost, and the SmoothLoss function is designed to solve the problem of the insufficient generalization ability of cross-entropy loss function during training. The experimental results show that the proposed model significantly reduces the number of parameters and outperforms other models in terms of classification accuracy on non-VPN datasets. Full article

Predictive maintenance of induction motors continues to be a significant challenge in ensuring industrial reliability and minimizing downtime. In this study, machine learning techniques are utilized to enhance fault diagnosis through the use of the Machinery Fault Database (MAFAULDA). A detailed extraction of statistical features was performed on multivariate time-series data to capture essential patterns that could indicate potential faults. Three machine learning algorithms—deep neural networks (DNNs), support vector machines (SVMs), and K-nearest neighbors (KNNs)—were applied to the dataset. Optimization strategies were carefully implemented along with oversampling techniques to improve model performance and handle imbalanced data. The results achieved through these models are highly promising. The SVM model demonstrated an accuracy of 95.4%, while KNN achieved an accuracy of 92.8%. Notably, the combination of deep neural networks with fast Fourier transform (FFT)-based autocorrelation features produced the highest performance, reaching an impressive accuracy of 99.7%. These results provide a novel approach to machine learning techniques in enhancing operational health and predictive maintenance of induction motor systems. Full article

Incremental Delta-Sigma (I-DS) analog-to-digital converters (ADCs) are one of the best candidates for integrated sensor interface systems when it comes to high resolution and power efficiency. Advanced architectures such as Multistage noise shaping (MASH) or extended counting (EC) I-DS ADCs can be used to achieve a high resolution and fast conversion times and avoid stability issues. Different architectures have been proposed in the state of the art (SoA), but there exists no extensive quantitative or qualitative comparison between them. This manuscript fills this gap by providing a detailed system-level comparison between MASH, EC, and other architectural options in I-DS ADCs, where different performances between these architectures are realized depending on the employed oversampling ratio (OSR) and the chosen number of quantizer bits. Also, for specific MASH designs, the appropriate choice of the digital filter improves the SQNR. The advantages, disadvantages, and limitations of the different architectures are presented including non-idealities such as coefficient mismatch showing that 2-1 MASH-LI is less sensitive to mismatch and provides a high maximum stable amplitude (MSA) relative to the simulated architectures. Furthermore, the 2-1 EC achieves good results and comes with the advantage of a lower noise penalty factor compared to the MASH architectures. This work is intended to assist designers in selecting the most appropriate enhanced I-DS MASH architecture for their specific requirements and applications. Full article

Rotating machinery is widely used across various industries, making its reliable operation crucial for industrial production. However, in real-world settings, intelligent fault diagnosis faces challenges due to imbalanced fault data and the complexity of neural network models. These challenges are particularly pronounced when defining decision boundaries accurately and managing limited computational resources in real-time machine monitoring. To address these issues, this study presents KDE-ADASYN-based MobileNet with SENet (KAMS), a lightweight convolutional neural network designed for fault diagnosis in rotating machinery. KAMS effectively handles data imbalances commonly found in industrial applications and is optimized for real-time monitoring. The model employs the Kernel Density Estimation Adaptive Synthetic Sampling (KDE-ADASYN) algorithm for oversampling to balance the data, applies fast Fourier transform (FFT) to convert time-domain signals into frequency-domain signals, and utilizes a 1D-MobileNet network enhanced with a Squeeze-and-Excitation (SE) block for feature extraction and fault diagnosis. Experimental results across datasets with varying imbalance ratios demonstrate that KAMS achieves excellent performance, maintaining nearly 90% accuracy even on highly imbalanced datasets. Comparative experiments further demonstrate that KAMS not only delivers exceptional diagnostic performance but also significantly reduces network parameters and computational resource requirements. Full article

Background: Quantitative pupillometry has been proposed as an objective means to diagnose acute sports-related concussion (SRC). Objective: To assess the diagnostic accuracy of a smartphone-based quantitative pupillometer in the acute diagnosis of SRC. Methods: Division I college football players had baseline pupillometry including pupillary light reflex (PLR) parameters of maximum resting diameter, minimum diameter after light stimulus, percent change in pupil diameter, latency of pupil constriction onset, mean constriction velocity, maximum constriction velocity, and mean dilation velocity using a smartphone-based app. When an SRC occurred, athletes had the smartphone pupillometry repeated as part of their concussion testing. All combinations of the seven PLR parameters were tested in machine learning binary classification models to determine the optimal combination for differentiating between non-concussed and concussed athletes. Results: 93 football athletes underwent baseline pupillometry testing. Among these athletes, 11 suffered future SRC and had pupillometry recordings repeated at the time of diagnosis. In the machine learning pupillometry analysis that used the synthetic minority oversampling technique to account for the significant class imbalance in our dataset, the best-performing model was a random forest algorithm with the combination of latency, maximum diameter, minimum diameter, mean constriction velocity, and maximum constriction velocity PLR parameters as feature inputs. This model produced 91% overall accuracy, 98% sensitivity, 84.2% specificity, area under the curve (AUC) of 0.91, and an F1 score of 91.6% in differentiating between baseline and SRC recordings. In the machine learning analysis prior to oversampling of our imbalanced dataset, the best-performing model was k-nearest neighbors using latency, maximum diameter, maximum constriction velocity, and mean dilation velocity to produce 82% accuracy, 40% sensitivity, 87% specificity, AUC of 0.64, and F1 score of 24%. Conclusions: Smartphone pupillometry in combination with machine learning may provide fast and objective SRC diagnosis in football athletes. Full article

Quick and accurate prediction of crop yields is beneficial for guiding crop field management and genetic breeding. This paper utilizes the fast and non-destructive advantages of an unmanned aerial vehicle equipped with a multispectral camera to acquire spatial characteristics of rice and conducts research on yield estimation in an open environment. The study proposes a yield estimation framework that hybrids synthetic minority oversampling technique (SMOTE) and deep neural network (DNN). Firstly, the framework used the Pearson correlation coefficient to select 10 key vegetation indices and determine the optimal feature combination. Secondly, it created a dataset for data augmentation through SMOTE, addressing the challenge of long data collection cycles and small sample sizes caused by long growth cycles. Then, based on this dataset, a yield estimation model was trained using DNN and compared with partial least squares regression (PLSR), support vector regression (SVR), and random forest (RF). The experimental results indicate that the hybrid framework proposed in this study performs the best (R² = 0.810, RMSE = 0.69 t/ha), significantly improving the accuracy of yield estimation compared to other methods, with an R² improvement of at least 0.191. It demonstrates that the framework proposed in this study can be used for rice yield estimation. Additionally, it provides a new approach for future yield estimation with small sample sizes for other crops or for predicting numerical crop indicators. Full article

To alleviate the data imbalance problem caused by subjective and objective factors, scholars have developed different data-preprocessing algorithms, among which undersampling algorithms are widely used because of their fast and efficient performance. However, when the number of samples of some categories in a multi-classification dataset is too small to be processed via sampling or the number of minority class samples is only one or two, the traditional undersampling algorithms will be less effective. In this study, we select nine multi-classification time series datasets with extremely few samples as research objects, fully consider the characteristics of time series data, and use a three-stage algorithm to alleviate the data imbalance problem. In stage one, random oversampling with disturbance items is used to increase the number of sample points; in stage two, on the basis of the latter operation, SMOTE (synthetic minority oversampling technique) oversampling is employed; in stage three, the dynamic time-warping distance is used to calculate the distance between sample points, identify the sample points of Tomek links at the boundary, and clean up the boundary noise. This study proposes a new sampling algorithm. In the nine multi-classification time series datasets with extremely few samples, the new sampling algorithm is compared with four classic undersampling algorithms, namely, ENN (edited nearest neighbours), NCR (neighborhood cleaning rule), OSS (one-side selection), and RENN (repeated edited nearest neighbors), based on the macro accuracy, recall rate, and F1-score evaluation indicators. The results are as follows: of the nine datasets selected, for the dataset with the most categories and the fewest minority class samples, FiftyWords, the accuracy of the new sampling algorithm was 0.7156, far beyond that of ENN, RENN, OSS, and NCR; its recall rate was also better than that of the four undersampling algorithms used for comparison, corresponding to 0.7261; and its F1-score was 200.71%, 188.74%, 155.29%, and 85.61% better than that of ENN, RENN, OSS, and NCR, respectively. For the other eight datasets, this new sampling algorithm also showed good indicator scores. The new algorithm proposed in this study can effectively alleviate the data imbalance problem of multi-classification time series datasets with many categories and few minority class samples and, at the same time, clean up the boundary noise data between classes. Full article

Vehicular ad hoc networks (VANETs) are used for vehicle to vehicle (V2V) and vehicle to infrastructure (V2I) communications. They are a special type of mobile ad hoc networks (MANETs) that can share useful information to improve road traffic and safety. In VANETs, vehicles are interconnected through a wireless medium, making the network susceptible to various attacks, such as Denial of Service (DoS), Distributed Denial of Service (DDoS), or even black hole attacks that exploit the wireless medium to disrupt the network. These attacks degrade the network performance of VANETs and prevent legitimate users from accessing resources. VANETs face unique challenges due to the fast mobility of vehicles and dynamic changes in network topology. The high-speed movement of vehicles results in frequent alterations in the network structure, posing difficulties in establishing and maintaining stable communication. Moreover, the dynamic nature of VANETs, with vehicles joining and leaving the network regularly, adds complexity to implementing effective security measures. These inherent constraints necessitate the development of robust and efficient solutions tailored to VANETs, ensuring secure and reliable communication in dynamic and rapidly evolving environments. Therefore, securing communication in VANETs is a crucial requirement. Traditional security countermeasures are not pertinent to autonomous vehicles. However, many machine learning (ML) technologies are being utilized to classify malicious packet information and a variety of solutions have been suggested to improve security in VANETs. In this paper, we propose an enhanced intrusion detection framework for VANETs that leverages mutual information to select the most relevant features for building an effective model and synthetic minority oversampling (SMOTE) to deal with the class imbalance problem. Random Forest (RF) is applied as our classifier, and the proposed method is compared with different ML techniques such as logistic regression (LR), K-Nearest Neighbor (KNN), decision tree (DT), and Support Vector Machine (SVM). The model is tested on three datasets, namely ToN-IoT, NSL-KDD, and CICIDS2017, addressing challenges such as missing values, unbalanced data, and categorical values. Our model demonstrated great performance in comparison to other models. It achieved high accuracy, precision, recall, and f1 score, with a 100% accuracy rate on the ToN-IoT dataset and 99.9% on both NSL-KDD and CICIDS2017 datasets. Furthermore, the ROC curve analysis demonstrated our model’s exceptional performance, achieving a 100% AUC score. Full article

Artificial intelligence and emerging data science techniques are being leveraged to interpret medical image scans. Traditional image analysis relies on visual interpretation by a trained radiologist, which is time-consuming and can, to some degree, be subjective. The development of reliable, automated diagnostic tools is a key goal of radiomics, a fast-growing research field which combines medical imaging with personalized medicine. Radiomic studies have demonstrated potential for accurate lung cancer diagnoses and prognostications. The practice of delineating the tumor region of interest, known as segmentation, is a key bottleneck in the development of generalized classification models. In this study, the incremental multiple resolution residual network (iMRRN), a publicly available and trained deep learning segmentation model, was applied to automatically segment CT images collected from 355 lung cancer patients included in the dataset “Lung-PET-CT-Dx”, obtained from The Cancer Imaging Archive (TCIA), an open-access source for radiological images. We report a failure rate of 4.35% when using the iMRRN to segment tumor lesions within plain CT images in the lung cancer CT dataset. Seven classification algorithms were trained on the extracted radiomic features and tested for their ability to classify different lung cancer subtypes. Over-sampling was used to handle unbalanced data. Chi-square tests revealed the higher order texture features to be the most predictive when classifying lung cancers by subtype. The support vector machine showed the highest accuracy, 92.7% (0.97 AUC), when classifying three histological subtypes of lung cancer: adenocarcinoma, small cell carcinoma, and squamous cell carcinoma. The results demonstrate the potential of AI-based computer-aided diagnostic tools to automatically diagnose subtypes of lung cancer by coupling deep learning image segmentation with supervised classification. Our study demonstrated the integrated application of existing AI techniques in the non-invasive and effective diagnosis of lung cancer subtypes, and also shed light on several practical issues concerning the application of AI in biomedicine. Full article

For medical data mining, the development of a class prediction model has been widely used to deal with various kinds of data classification problems. Classification models especially for high-dimensional gene expression datasets have attracted many researchers in order to identify marker genes for distinguishing any type of cancer cells from their corresponding normal cells. However, skewed class distributions often occur in the medical datasets in which at least one of the classes has a relatively small number of observations. A classifier induced by such an imbalanced dataset typically has a high accuracy for the majority class and poor prediction for the minority class. In this study, we focus on an SVM classifier with a Gaussian radial basis kernel for a binary classification problem. In order to take advantage of an SVM and to achieve the best generalization ability for improving the classification performance, we will address two important problems: the class imbalance and parameter selection during SVM parameter optimization. First of all, we proposed a novel adjustment method called b-SVM, for adjusting the cutoff threshold of the SVM. Second, we proposed a fast and simple approach, called the Min-max gamma selection, to optimize the model parameters of SVMs without carrying out an extensive k-fold cross validation. An extensive comparison with a standard SVM and well-known existing methods are carried out to evaluate the performance of our proposed algorithms using simulated and real datasets. The experimental results show that our proposed algorithms outperform the over-sampling techniques and existing SVM-based solutions. This study also shows that the proposed Min-max gamma selection is at least 10 times faster than the cross-validation selection based on the average running time on six real datasets. Full article

By benefiting from one-bit sampling, the system deployment of synthetic aperture radar (SAR) can be greatly simplified. However, it usually requires a high oversampling rate to avoid the apparent degradation in imagery, which counteracts the storage-saving advantages. In this paper, a two-bit lightweight SAR imaging strategy is proposed to take the advantage of one-bit quantization in simplification but get rid of the requirement of sampling at a high rate. Specifically, based on one-bit quantization, an extra bit after an appropriate phase shifting is introduced to suppress the harmonics resulting from the nonlinear effect of quantization. In this way, the awkward nonlinearity in conventional one-bit schemes can be tackled by the nonlinearity generated with the newly introduced bit. Hence, this improves the imaging quality. In addition, the proposed method does not rely on fast sampling. The harmonic suppression effect is retained under low-sampling-rate conditions. Therefore, the amount of data acquired will decrease dramatically. This will benefit the whole process of imaging and, consequently, lighten the system burden and cost. The theoretical analysis and experimental results showcase the superiority of the proposed method. Full article

With the great increases in data transmission rate requirements, analog-to-digital converter (ADC)-based wireline receivers have received more and more attention due to their flexible and powerful equalization capabilities. Considering power consumption, baud-rate Mueller–Muller clock and data recovery (MM-CDR) circuits are widely used in ADC-based wireline receivers since MM-CDR circuits only need one sample signal per unit interval (UI). However, MM-CDR circuits need to set an additional Vref voltage to match the size of the main tap of the channel. If the Vref matching is not appropriate or the signal quality is good as a square wave, MM-CDR circuits cannot accurately lock on to a certain phase and instead drift within a phase range. Therefore, MM-CDR circuits are not as robust and stable as oversampled CDR circuits. In this study, a digital bang-bang clock and data recovery (DBB-CDR) circuit combined with an ADC-based wireline receiver was proposed. The DBB-CDR circuit could eliminate various unstable factors of MM-CDR circuits and achieve fast and robust phase locking without excessively increasing power consumption. A model of the DBB-CDR circuit was combined with an actual 32 Gb/s ADC-based wireline receiver, which was implemented in 28 nm CMOS technology to analyze the performance of the DBB-CDR circuit. The simulation results showed that the DBB-CDR circuit could achieve 0.42 UIpp JTOL@10MHz, and that the minimum JTOL value was 0.362 UIpp under a 0.04 UI variance of Gaussian jitter. The area and power consumption of the DBB-CDR circuit were only 64 μm² and 0.02 mW, respectively; and the DBB-CDR circuit could also obtain very stable phase locking and demonstrated a fast frequency offset tracking ability when there was a frequency offset. Full article

In the context of cognitive radio, smart cities and Internet-of-Things, the need for advanced radio spectrum monitoring becomes crucial. However, surveillance of a wide frequency band without using extremely expensive high sampling rate devices is a challenging task. The recent development of compressed sampling approaches offers a promising solution to these problems. In this context, the Modulated Wideband Converter (MWC), a blind sub-Nyquist sampling system, is probably the most realistic approach and was successfully validated in real-world conditions. The MWC can be realized with existing analog components, and there exist calibration methods that are able to integrate the imperfections of the mixers, filters and ADCs, hence allowing its use in the real world. The MWC underlying model is based on signal processing concepts such as filtering, modulation, Fourier series decomposition, oversampling and undersampling, spectrum aliasing, and so on, as well as in-flow data processing. In this paper, we develop an MWC model that is entirely based on linear algebra, matrix theory and block processing. We show that this approach has many interests: straightforward translation of mathematical equations into simple and efficient software programming, suppression of some constraints of the initial model, and providing a basis for the development of an extremely fast system calibration method. With a typical MWC acquisition device, we obtained a speed-up of the calibration computation time by a factor greater than 20 compared with a previous implementation. Full article

Android has become the leading mobile ecosystem because of its accessibility and adaptability. It has also become the primary target of widespread malicious apps. This situation needs the immediate implementation of an effective malware detection system. In this study, an explainable malware detection system was proposed using transfer learning and malware visual features. For effective malware detection, our technique leverages both textual and visual features. First, a pre-trained model called the Bidirectional Encoder Representations from Transformers (BERT) model was designed to extract the trained textual features. Second, the malware-to-image conversion algorithm was proposed to transform the network byte streams into a visual representation. In addition, the FAST (Features from Accelerated Segment Test) extractor and BRIEF (Binary Robust Independent Elementary Features) descriptor were used to efficiently extract and mark important features. Third, the trained and texture features were combined and balanced using the Synthetic Minority Over-Sampling (SMOTE) method; then, the CNN network was used to mine the deep features. The balanced features were then input into the ensemble model for efficient malware classification and detection. The proposed method was analyzed extensively using two public datasets, CICMalDroid 2020 and CIC-InvesAndMal2019. To explain and validate the proposed methodology, an interpretable artificial intelligence (AI) experiment was conducted. Full article