Editor’s Choice Articles

Annotation and analysis of sports videos is a time-consuming task that, once automated, will provide benefits to coaches, players, and spectators. American football, as the most watched sport in the United States, could especially benefit from this automation. Manual annotation and analysis of recorded videos of American football games is an inefficient and tedious process. Currently, most college football programs focus on annotating offensive formations to help them develop game plans for their upcoming games. As a first step to further research for this unique application, we use computer vision and deep learning to analyze an overhead image of a football play immediately before the play begins. This analysis consists of locating individual football players and labeling their position or roles, as well as identifying the formation of the offensive team. We obtain greater than 90% accuracy on both player detection and labeling, and 84.8% accuracy on formation identification. These results prove the feasibility of building a complete American football strategy analysis system using artificial intelligence. Collecting a larger dataset in real-world situations will enable further improvements. This would likewise enable American football teams to analyze game footage quickly. Full article

Almost all rotating machinery in the industry has bearings as their key building block and most of these machines run 24 × 7. This makes bearing health prediction an active research area for predictive maintenance solutions. Many state of the art Deep Neural Network (DNN) models have been proposed to solve this. However, most of these high performance models are computationally expensive and have high memory requirements. This limits their use to very specific industrial applications with powerful hardwares deployed close the the machinery. In order to bring DNN-based solutions to a potential use in the industry, we need to deploy these models on Microcontroller Units (MCUs) which are cost effective and energy efficient. However, this step is typically neglected in literature as it poses new challenges. The primary concern when inferencing the DNN models on MCUs is the on chip memory of the MCU that has to fit the model, the data and additional code to run the system. Almost all the state of the art models fail this litmus test since they feature too many parameters. In this paper, we show the challenges related to the deployment, review possible solutions and evaluate one of them showing how the deployment can be realized and what steps are needed. The focus is on the steps required for the actual deployment rather than finding the optimal solution. This paper is among the first to show the deployment on MCUs for a predictive maintenance use case. We first analyze the gap between State Of The Art benchmark DNN models for bearing defect classification and the memory constraint of two MCU variants. Additionally, we review options to reduce the model size such as pruning and quantization. Afterwards, we evaluate a solution to deploy the DNN models by pruning them in order to fit them into microcontrollers. Our results show that most models under test can be reduced to fit MCU memory for a maximum loss of $3\%$ in average accuracy of the pruned models in comparison to the original models. Based on the results, we also discuss which methods are promising and which combination of model and feature work best for the given classification problem. Full article

Spam detection is an essential and unavoidable problem in today’s society. Most of the existing studies have used string-based detection methods with models and have been conducted on a single language, especially with English datasets. However, in the current global society, research on languages other than English is needed. String-based spam detection methods perform different preprocessing steps depending on language type due to differences in grammatical characteristics. Therefore, our study proposes a text-processing method and a string-imaging method. The CNN 2D visualization technology used in this paper can be applied to datasets of various languages by processing the data as images, so they can be equally applied to languages other than English. In this study, English and Korean spam data were used. As a result of this study, the string-based detection models of RNN, LSTM, and CNN 1D showed average accuracies of 0.9871, 0.9906, and 0.9912, respectively. On the other hand, the CNN 2D image-based detection model was confirmed to have an average accuracy of 0.9957. Through this study, we present a solution that shows that image-based processing is more effective than string-based processing for string data and that multilingual processing is possible based on the CNN 2D model. Full article

Aiming to differentiate various transportation modes and detect the means of transport an individual uses, is the focal point of transportation mode detection, one of the problems in the field of intelligent transport which receives the attention of researchers because of its interesting and useful applications. In this paper, we present TMD-BERT, a transformer-based model for transportation mode detection based on sensor data. The proposed transformer-based approach processes the entire sequence of data, understand the importance of each part of the input sequence and assigns weights accordingly, using attention mechanisms, to learn global dependencies in the sequence. The experimental evaluation shows the high performance of the model compared to the state of the art, demonstrating a prediction accuracy of 98.8%. Full article

As a distributed machine learning paradigm, federated learning (FL) enables participating clients to share only model gradients instead of local data and achieves the secure sharing of private data. However, the lack of clients’ willingness to participate in FL and the malicious influence of unreliable clients both seriously degrade the performance of FL. The current research on the incentive mechanism of FL lacks the accurate assessment of clients’ truthfulness and reliability, and the incentive mechanism based on untruthful and unreliable clients is unreliable and inefficient. To solve this problem, we propose an incentive mechanism based on the reputation mechanism and reverse auction to achieve a more truthful, more reliable, and more efficient FL. First, we introduce the reputation mechanism to measure clients’ truthfulness and reliability through multiple reputation evaluations and design a reliable client selection scheme. Then the reverse auction is introduced to select the optimal clients that maximize the social surplus while satisfying individual rationality, incentive compatibility, and weak budget balance. Extensive experimental results demonstrate that this incentive mechanism can motivate more clients with high-quality data and high reputations to participate in FL with less cost, which increases the FL tasks’ economic benefit by 31% and improves the accuracy from 0.9356 to 0.9813, and then promote the efficient and stable development of the FL service trading market. Full article

Machines in factories are typically operated 24 h a day to support production, which may result in malfunctions. Such mechanical malfunctions may disrupt factory output, resulting in financial losses or human casualties. Therefore, we investigate a deep learning model that can detect abnormalities in machines based on the operating noise. Various data preprocessing methods, including the discrete wavelet transform, the Hilbert transform, and short-time Fourier transform, were applied to extract characteristics from machine-operating noises. To create a model that can be used in factories, the environment of real factories was simulated by introducing noise and quality degradation to the sound dataset for Malfunctioning Industrial Machine Investigation and Inspection (MIMII). Thus, we proposed a lightweight model that runs reliably even in noisy and low-quality sound data environments, such as a real factory. We propose a Convolutional Neural Network–Long Short-Term Memory (CNN–LSTM) model using Short-Time Fourier Transforms (STFTs), and the proposed model can be very effective in terms of application because it is a lightweight model that requires only about 6.6% of the number of parameters used in the underlying CNN, and has only a performance difference within 0.5%. Full article

Food safety technologies are important in maintaining physical health for everyone. It is important to digitize the scents of foods to enable an effective human–computer interface for smells. In this work, an intelligent gas-sensing system is designed and integrated to capture the smells of food and convert them into digital scents. Fruit samples are used for testing as they release volatile organic components (VOCs) which can be detected by the gas sensors in the system. Decision tree, principal component analysis (PCA), linear discriminant analysis (LDA), and one-dimensional convolutional neural network (1D-CNN) algorithms were adopted and optimized to analyze and precisely classify the sensor responses. Furthermore, the proposed system and data processing algorithms can be used to precisely identify the digital scents and monitor the decomposition dynamics of different foods. Such a promising technology is important for mutual understanding between humans and computers to enable an interface for digital scents, which is very attractive for food identification and safety monitoring. Full article

Memristors, as new electronic elements, have been under rigorous study in recent years, owing to their good memory and switching properties, low power consumption, nano-dimensions and a good compatibility to present integrated circuits, related to their promising applications in electronic circuits and chips. The main purpose of this paper is the application and analysis of the operations of metal–oxide memristors in logic gates and complex schemes, using several standard and modified memristor models and a comparison between their behavior in LTSPICE at a hard-switching, paying attention to their fast operation and switching properties. Several basic logic gates—OR, AND, NOR, NAND, XOR, based on memristors and CMOS transistors are considered. The logic schemes based on memristors are applicable in electronic circuits with artificial intelligence. They are analyzed in LTSPICE for pulse signals and a hard-switching functioning of the memristors. The analyses confirm the proper, fast operation and good switching properties of the considered modified memristor models in logical circuits, compared to several standard models. The modified models are compared to several classical models, according to some significant criteria such as operating frequency, simulation time, accuracy, complexity and switching properties. Based on the basic memristor logic gates, a more complex logic scheme is analyzed. Full article

The selection of the optimizer is critical for convergence in the field of on-chip training. As one second moment optimizer, adaptive moment estimation (ADAM) shows a significant advantage compared with non-moment optimizers such as stochastic gradient descent (SGD) and first-moment optimizers such as Momentum. However, ADAM is hard to implement on hardware due to the computationally intensive operations, including square, root extraction, and division. This work proposed Hardware-ADAM (HW-ADAM), an efficient fixed-point accelerator for ADAM highlighting hardware-oriented mathematical optimizations. HW-ADAM has two designs: Efficient-ADAM (E-ADAM) unit reduced the hardware resource consumption by around 90% compared with the related work. E-ADAM achieved a throughput of 2.89 MUOP/s (Million Updating Operation per Second), which is 2.8× of the original ADAM. Fast-ADAM (F-ADAM) unit reduced 91.5% flip-flops, 65.7% look-up tables, and 50% DSPs compared with the related work. The F-ADAM unit achieved a throughput of 16.7 MUOP/s, which is 16.4× of the original ADAM. Full article

The hardware architecture of a deep-neural-network-based feature point extraction method is proposed for the simultaneous localization and mapping (SLAM) in robotic applications, which is named the Feature Point based SLAM Network (FPSNET). Some key techniques are deployed to improve the hardware and power efficiency. The data path is devised to reduce overall off-chip memory accesses. The intermediate data and partial sums resulting in the convolution process are stored in available on-chip memories, and optimized hardware is employed to compute the one-point activation function. Meanwhile, address generation units are used to avoid data overlapping in memories. The proposed FPSNET has been designed in 65 nm CMOS technology with a core area of 8.3 mm². This work reduces the memory overhead by 50% compared to traditional data storage for activation and overall by 35% for on-chip memories. The synthesis and simulation results show that it achieved a 2.0× higher performance compared with the previous design while achieving a power efficiency of 1.0 TOPS/W, which is 2.4× better than previous work. Compared to other ASIC designs with close peak throughput and power efficiency performance, the presented FPSNET has the smallest chip area (at least 42.4% reduction). Full article

In this paper, a single-chip front-end (SCFE) operating in Ku-band (12–17 GHz) is presented. It is designed exploiting a GaN on SiC technology featured by 150 nm gate length provided by UMS foundry. This MMIC integrates high power and low noise amplification functions enabled by a single-pole double-throw (SPDT) switch, occupying a total area of 20 mm². The transmitting chain (Tx) presents a 39 dBm output power, a power added efficiency (PAE) higher than 30% and a 22 dB power gain. The receive path (Rx) offers a low noise figure (NF) lower than 2.8 dB with 25 dB of linear gain. The Rx port output power leakage is limited on chip to be below 15 dBm even at high compression levels. Finally, a complete characterization of the SCFE in the Rx and Tx modes is presented, also showing the measurement of the recovery time in the presence of large-signal interferences. Full article

The binarized neural network (BNN) is one of the most promising candidates for low-cost convolutional neural networks (CNNs). This is because of its significant reduction in memory and computational costs, and reasonable classification accuracy. Content-addressable memory (CAM) can perform binarized convolution operations efficiently since the bitwise comparison in CAM matches well with the binarized multiply operation in a BNN. However, a significant design issue in CAM-based BNN accelerators is that the operational reliability is severely degraded by process variations during match-line (ML) sensing operations. In this paper, we proposed a novel ML sensing scheme to reduce the hardware error probability. Most errors occur when the difference between the number of matches in the evaluation ML and the reference ML is small; thus, the proposed hardware identified cases that are vulnerable to process variations using dual references. The proposed dual-reference sensing structure has >49% less ML sensing errors than that of the conventional design, leading to a >1.0% accuracy improvement for Fashion MNIST image classification. In addition, owing to the parallel convolution operation of the CAM-based BNN accelerator, the proposed hardware achieved >34% processing-time improvement compared with that of the digital logic implementation. Full article

High security has always been the ultimate goal of image encryption, and the closer the ciphertext image is to the true random number, the higher the security. Aiming at popular chaotic image encryption methods, particle swarm optimization (PSO) is studied to select the parameters and initial values of chaotic systems so that the chaotic sequence has higher entropy. Different from the other PSO-based image encryption methods, the proposed method takes the parameters and initial values of the chaotic system as particles instead of encrypted images, which makes it have lower complexity and therefore easier to be applied in real-time scenarios. To validate the optimization framework, this paper designs a new image encryption scheme. The algorithm mainly includes key selection, chaotic sequence preprocessing, block scrambling, expansion, confusion, and diffusion. The key is selected by PSO and brought into the chaotic map, and the generated chaotic sequence is preprocessed. Based on block theory, a new intrablock and interblock scrambling method is designed, which is combined with image expansion to encrypt the image. Subsequently, the confusion and diffusion framework is used as the last step of the encryption process, including row confusion diffusion and column confusion diffusion, which makes security go a step further. Several experimental tests manifest that the scenario has good encryption performance and higher security compared with some popular image encryption methods. Full article

Keyword spotting (KWS) plays a crucial role in human–machine interactions involving smart devices. In recent years, temporal convolutional networks (TCNs) have performed outstandingly with less computational complexity, in comparison with classical convolutional neural network (CNN) methods. However, it remains challenging to achieve a trade-off between a small-footprint model and high accuracy for the edge deployment of the KWS system. In this article, we propose a small-footprint model based on a modified temporal efficient neural network (TENet) and a simplified mel-frequency cepstrum coefficient (MFCC) algorithm. With the batch-norm folding and int8 quantization of the network, our model achieves the accuracy of 95.36% on Google Speech Command Dataset (GSCD) with only 18 K parameters and 461 K multiplications. Furthermore, following a hardware/model co-design approach, we propose an optimized dataflow and a configurable hardware architecture for TENet inference. The proposed accelerator implemented on Xilinx zynq 7z020 achieves an energy efficiency of 25.6 GOPS/W and reduces the runtime by 3.1× compared with state-of-the-art work. Full article

The artificial intelligence (AI) application in instruments such as impedance spectroscopy highlights the difficulty to choose an electronic technology that correctly solves the basic performance problems, adaptation to the context, flexibility, precision, autonomy, and speed of design. Present work demonstrates that FPGAs, in conjunction with an optimized high-level synthesis (HLS), allow us to have an efficient connection between the signals sensed by the instrument and the artificial neural network-based AI computing block that will analyze them. State-of-the-art comparisons and experimental results also demonstrate that our designed and developed architectures offer the best compromise between performance, efficiency, and system costs in terms of artificial neural networks implementation. In the present work, computational efficiency above 21 Mps/DSP and power efficiency below 1.24 mW/Mps are achieved. It is important to remark that these results are more relevant because the system can be implemented on a low-cost FPGA. Full article

Despite the tremendous socioeconomic importance of supply chains (SCs), security officers and operators are faced with no easy and integrated way for protecting their critical, and interconnected, infrastructures from cyber-attacks. As a result, solutions and methodologies that support the detection of malicious activity on SCs are constantly researched into and proposed. Hence, this work presents the implementation of a low-cost reconfigurable intrusion detection system (IDS), on the edge, that can be easily integrated into SC networks, thereby elevating the featured levels of security. Specifically, the proposed system offers real-time cybersecurity intrusion detection over high-speed networks and services by offloading elements of the security check workloads on dedicated reconfigurable hardware. Our solution uses a novel framework that implements the Aho–Corasick algorithm on the reconfigurable fabric of a multi-processor system-on-chip (MPSoC), which supports parallel matching for multiple network packet patterns. The initial performance evaluation of this proof-of-concept shows that it holds the potential to outperform existing software-based solutions while unburdening SC nodes from demanding cybersecurity check workloads. The proposed system performance and its efficiency were evaluated using a real-life environment in the context of European Union’s Horizon 2020 research and innovation program, i.e., CYRENE. Full article

Electrical capacitance tomography (ECT) is a technique of imaging the distribution of permittivity inside an object under test. Capacitance is measured between the electrodes surrounding the object, and the image is reconstructed from these data by solving the inverse problem. Although both sinusoidal excitation and pulse excitation are used in the sensing circuit, only the AC method is used to measure both components of complex capacitance. In this article, a novel method of complex capacitance measurement using pulse excitation is proposed for ECT. The real and imaginary components are calculated from digital samples of the integrator response. A pulse shape in the front-end circuit was analyzed using the Laplace transform. The numerical simulations of the electric field inside the imaging volume as well as simulations of a pulse excitation in the front-end circuit were performed. The calculation of real and imaginary components using digital samples of the output signal was verified. The permittivity and conductivity images reconstructed for the test object were presented. The method enables imaging of permittivity and conductivity spatial distributions using capacitively coupled electrodes and may be an alternative measurement method for ECT as well as for electrical impedance tomography. Full article

Biological neural networks demonstrate remarkable resilience and the ability to compensate for neuron losses over time. Thus, the effects of neural/synaptic losses in the brain go mostly unnoticed until the loss becomes profound. This study analyses the capacity of electronic spiking networks to compensate for the sudden, random neuron failure (“death”) due to reliability degradation or other external factors such as exposure to ionizing radiation. Electronic spiking neural networks with memristive synapses are designed to learn spatio-temporal patterns representing 25 or 100-pixel characters. The change in the pattern learning ability of the neural networks is observed as the afferents (input layer neurons) in the network fail/die during network training. Spike-timing-dependent plasticity (STDP) learning behavior is implemented using shaped action potentials with a realistic, non-linear memristor model. This work focuses on three cases: (1) when only neurons participating in the pattern are affected, (2) when non-participating neurons (those that never present spatio-temporal patterns) are disabled, and (3) when random/non-selective neuron death occurs in the network (the most realistic scenario). Case 3 is further analyzed to compare what happens when neuron death occurs over time versus when multiple afferents fail simultaneously. Simulation results emphasize the importance of non-participating neurons during the learning process, concluding that non-participating afferents contribute to improving the learning ability and stability of the neural network. Instantaneous neuron death proves to be more detrimental for the network compared to when afferents fail over time. To a surprising degree, the electronic spiking neural networks can sometimes retain their pattern recognition capability even in the case of significant neuron death. Full article

This study presents a universal reconfigurable hardware accelerator for efficient processing of sparse decision trees, artificial neural networks and support vector machines. The main idea is to develop a hardware accelerator that will be able to directly process sparse machine learning models, resulting in shorter inference times and lower power consumption compared to existing solutions. To the author’s best knowledge, this is the first hardware accelerator of this type. Additionally, this is the first accelerator that is capable of processing sparse machine learning models of different types. Besides the hardware accelerator itself, algorithms for induction of sparse decision trees, pruning of support vector machines and artificial neural networks are presented. Such sparse machine learning classifiers are attractive since they require significantly less memory resources for storing model parameters. This results in reduced data movement between the accelerator and the DRAM memory, as well as a reduced number of operations required to process input instances, leading to faster and more energy-efficient processing. This could be of a significant interest in edge-based applications, with severely constrained memory, computation resources and power consumption. The performance of algorithms and the developed hardware accelerator are demonstrated using standard benchmark datasets from the UCI Machine Learning Repository database. The results of the experimental study reveal that the proposed algorithms and presented hardware accelerator are superior when compared to some of the existing solutions. Throughput is increased up to 2 times for decision trees, 2.3 times for support vector machines and 38 times for artificial neural networks. When the processing latency is considered, maximum performance improvement is even higher: up to a 4.4 times reduction for decision trees, a 84.1 times reduction for support vector machines and a 22.2 times reduction for artificial neural networks. Finally, since it is capable of supporting sparse classifiers, the usage of the proposed hardware accelerator leads to a significant reduction in energy spent on DRAM data transfers and a reduction of 50.16% for decision trees, 93.65% for support vector machines and as much as 93.75% for artificial neural networks, respectively. Full article

The mentality of electricity consumers is one of the most important entities that must be addressed when dealing with issues in the operation of power systems. Consumers are used to being completely passive, but recently these things have changed as significant progress of Information and Communication Technologies (ICT) and Internet of Things (IoT) has gained momentum. In this paper, we propose a statistical measurement model using a covariance structure, specifically a first-order confirmatory factor analysis (CFA) using SAS CALIS procedure to identify the factors that could contribute to the change of attitude within energy communities. Furthermore, this research identifies latent constructs and indicates which observed variables load on or measure them. For the simulation, two complex data sets of questionnaires created by the Irish Commission for Energy Regulation (CER) were analyzed, demonstrating the influence of some exogenous variables on the items of the questionnaires. The results revealed that there is a relevant relationship between the social–economic and the behavioral factors and the observed variables. Furthermore, the models provided a good fit to the data, as measured by the performance indicators. Full article

In this paper, we propose a homomorphic encryption-based privacy protection scheme for DBSCAN clustering to reduce the risk of privacy leakage during data outsourcing computation. For the purpose of encrypting data in practical applications, we propose a variety of data preprocessing methods for different data accuracies. We also propose data preprocessing strategies based on different data precision and different computational overheads. In addition, we also design a protocol to implement the cipher text comparison function between users and cloud servers. Analysis of experimental results indicates that our proposed scheme has high clustering accuracy and can guarantee the privacy and security of the data. Full article

Chagas disease, caused by the Trypanosoma cruzi (T. cruzi) parasite, is the third most common parasitosis worldwide. Most of the infected subjects can remain asymptomatic without an opportune and early detection or an objective diagnostic is not conducted. Frequently, the disease manifests itself after a long time, accompanied by severe heart disease or by sudden death. Thus, the diagnosis is a complex and challenging process where several factors must be considered. In this paper, a novel pipeline is presented integrating temporal data from four modalities (electrocardiography signals, echocardiography images, Doppler spectrum, and ELISA antibody titers), multiple features selection analyses by a univariate analysis and a machine learning-based selection. The method includes an automatic dichotomous classification of animal status (control vs. infected) based on Random Forest, Extremely Randomized Trees, Decision Trees, and Support Vector Machine. The most relevant multimodal attributes found were ELISA (IgGT, IgG1, IgG2a), electrocardiography (SR mean, QT and ST intervals), ascending aorta Doppler signals, and echocardiography (left ventricle diameter during diastole). Concerning automatic classification from selected features, the best accuracy of control vs. acute infection groups was 93.3 ± 13.3% for cross-validation and 100% in the final test; for control vs. chronic infection groups, it was 100% and 100%, respectively. We conclude that the proposed machine learning-based approach can be of help to obtain a robust and objective diagnosis in early T. cruzi infection stages. Full article

With the expansion of the civil UAV (Unmanned Aerial Vehicle) market, UAVs are also increasingly being used in illegal activities such as espionage and snooping on privacy. Therefore, how to effectively control the activities of UAVs in cities has become an urgent problem to be solved. Considering the urban background and the radar performance of communication signals, a low-altitude target detection scheme based on 5G base stations is proposed in this paper. A 5G signal is used as the external radiation source, the method of transceiver separation is adopted, and the forward-scattered waves are used to complete the detection of UAV. This paper mainly analyzes the principle of forward scattering detection in an urban environment, where the forward-scattered wave of a target is stronger than the backward-reflected wave and contains both height difference and midline height information on the target. Based on the above theory, this paper proposes a forward-scattered wave recognition algorithm based on YOLOv3-FCWImageNet, which transforms the forward-scattered wave recognition problem into a target detection problem and accomplishes the recognition of forward-scattered waves by using the excellent performance of algorithms in the field of image recognition. Simulation results show that FCWImageNet can distinguish two different low-altitude targets effectively, and realize the monitoring and classification of UAVs. Full article

A desirable feature of an electrical capacitance tomography system is the adaptation possibility to any sensor configuration and measurement mode. A run-time reconfiguration of a system for electrical capacitance tomography is presented. An original mechanism is elaborated to reconfigure, on the fly, a modular EVT4 system with multiple FPGAs installed. The outlined system architecture is based on FPGA programmable logic devices (Xilinx Spartan) and PicoBlaze soft-core processors. Soft-core processors are used for communication, measurement control and data preprocessing. A novel method of FPGA partial reconfiguration is described, in which a PicoBlaze soft-core processor is used as a reconfiguration controller. Behavioral reconfiguration of the system is obtained by providing run-time access to the program code of a soft-core control processor. The tests using EVT4 hardware and different algorithms for tomographic scanning were performed. A test object was measured using 2D and 3D sensors. The time and resources required for the examined reconfiguration procedure are evaluated. Full article

Bringing emotion recognition (ER) out of the controlled laboratory setup into everyday life can enable applications targeted at a broader population, e.g., helping people with psychological disorders, assisting kids with autism, monitoring the elderly, and general improvement of well-being. This work reviews progress in sensors and machine learning methods and techniques that have made it possible to move ER from the lab to the field in recent years. In particular, the commercially available sensors collecting physiological data, signal processing techniques, and deep learning architectures used to predict emotions are discussed. A survey on existing systems for recognizing emotions in real-life scenarios—their possibilities, limitations, and identified problems—is also provided. The review is concluded with a debate on what challenges need to be overcome in the domain in the near future. Full article

Heart rate is a nonstationary signal and its variation may contain indicators of current disease or warnings about impending cardiac diseases. Hence, heart rate variation analysis has become a noninvasive tool to further study the activities of the autonomic nervous system. In this scenario, the Poincaré plot analysis has proven to be a valuable tool to support cardiac diseases diagnosis. The study’s aim is a preliminary exploration of the feasibility of machine learning to classify subjects belonging to five cardiac states (healthy, hypertension, myocardial infarction, congestive heart failure and heart transplanted) using ten unconventional quantitative parameters extracted from bidimensional and three-dimensional Poincaré maps. Knime Analytic Platform was used to implement several machine learning algorithms: Gradient Boosting, Adaptive Boosting, k-Nearest Neighbor and Naïve Bayes. Accuracy, sensitivity and specificity were computed to assess the performances of the predictive models using the leave-one-out cross-validation. The Synthetic Minority Oversampling technique was previously performed for data augmentation considering the small size of the dataset and the number of features. A feature importance, ranked on the basis of the Information Gain values, was computed. Preliminarily, a univariate statistical analysis was performed through one-way Kruskal Wallis plus post-hoc for all the features. Machine learning analysis achieved interesting results in terms of evaluation metrics, such as demonstrated by Adaptive Boosting and k-Nearest Neighbor (accuracies greater than 90%). Gradient Boosting and k-Nearest Neighbor reached even 100% score in sensitivity and specificity, respectively. The most important features according to information gain are in line with the results obtained from the statistical analysis confirming their predictive power. The study shows the proposed combination of unconventional features extracted from Poincaré maps and well-known machine learning algorithms represents a valuable approach to automatically classify patients with different cardiac diseases. Future investigations on enriched datasets will further confirm the potential application of this methodology in diagnostic. Full article

With the increasing use of multi-purpose artificial intelligence of things (AIOT) devices, embedded field-programmable gate arrays (FPGA) represent excellent platforms for deep neural network (DNN) acceleration on edge devices. FPGAs possess the advantages of low latency and high energy efficiency, but the scarcity of FPGA development resources challenges the deployment of DNN-based edge devices. Register-transfer level programming, hardware verification, and precise resource allocation are needed to build a high-performance FPGA accelerator for DNNs. These tasks present a challenge and are time consuming for even experienced hardware developers. Therefore, we propose an automated, collaborative design process employing an automatic design space exploration tool; an automatic DNN engine enables the tool to reshape and parse a DNN model from software to hardware. We also introduce a long short-term memory (LSTM)-based model to predict performance and generate a DNN model that suits the developer requirements automatically. We demonstrate our design scheme with three FPGAs: a zcu104, a zcu102, and a Cyclone V SoC (system on chip). The results show that our hardware-based edge accelerator exhibits superior throughput compared with the most advanced edge graphics processing unit. Full article

Due to the dynamic mobility feature, the proactive flow rule cache method has become one promising solution in software-defined networking (SDN)-based access networks to reduce the number of flow rule installation procedures between the forwarding nodes and SDN controller. However, since there is a flow rule cache limit for the forwarding node, an efficient flow rule cache strategy is required. To address this challenge, this paper proposes the mobility-aware hybrid flow rule cache scheme. Based on the comparison between the delay requirement of the incoming flow and the response delay of the controller, the proposed scheme decides to install the flow rule either proactively or reactively for the target candidate forwarding nodes. To find the optimal number of proactive flow rules considering the flow rule cache limits, an integer linear programming (ILP) problem is formulated and solved using the heuristic method. Extensive simulation results demonstrate that the proposed scheme outperforms the existing schemes in terms of the flow table utilization ratio, flow rule installation delay, and flow rules hit ratio under various settings. Full article

This paper reviews the research trends that link the advanced technical aspects of recommendation systems that are used in various service areas and the business aspects of these services. First, for a reliable analysis of recommendation models for recommendation systems, data mining technology, and related research by application service, more than 135 top-ranking articles and top-tier conferences published in Google Scholar between 2010 and 2021 were collected and reviewed. Based on this, studies on recommendation system models and the technology used in recommendation systems were systematized, and research trends by year were analyzed. In addition, the application service fields where recommendation systems were used were classified, and research on the recommendation system model and recommendation technique used in each field was analyzed. Furthermore, vast amounts of application service-related data used by recommendation systems were collected from 2010 to 2021 without taking the journal ranking into consideration and reviewed along with various recommendation system studies, as well as applied service field industry data. As a result of this study, it was found that the flow and quantitative growth of various detailed studies of recommendation systems interact with the business growth of the actual applied service field. While providing a comprehensive summary of recommendation systems, this study provides insight to many researchers interested in recommendation systems through the analysis of its various technologies and trends in the service field to which recommendation systems are applied. Full article

In this paper, a model-based approach for the design of a bidirectional onboard charger (OBC) device for modern hybrid and fully electrified vehicles is proposed. The main objective and contribution of our study is to incorporate in the same simulation environment both modelling of electrical and thermal behaviour of switching devices. This is because most (if not all) of the studies in the literature present analyses of thermal behaviour based on the use of FEM (Finite Element Method) SWs, which in fact require the definition of complicated models based on partial derivative equations. The simulation of such accurate models is computationally expensive and, therefore, cannot be incorporated into the same virtual environment in which the circuit equations are solved. This requires long waiting times and also means that electrical and thermal models do not interact with each other, limiting the completeness of the analysis in the design phase. As a case study, we take as reference the architecture of a modular bidirectional single-phase OBC, consisting of a Totem Pole-type AC/DC converter with Power Factor Correction (PFC) followed by a Dual Active Bridge (DAB) type DC/DC converter. Specifically, we consider a 7 kW OBC, for which its modules consist of switching devices made with modern 900 V GaN (Gallium Nitrade) and 1200 V SiC (Silicon Carbide) technologies, to achieve maximum performance and efficiency. We present a procedure for sizing and selecting electronic devices based on the analysis of behaviour through circuit models of the Totem Pole PFC and DAB converter in order to perform validation by using simulations that are as realistic as possible. The developed models are tested under various operating conditions of practical interest in order to validate the robustness of the implemented control algorithms under varying operating conditions. The validation of the models and control loops is also enhanced by an exhaustive robustness analysis of the parametric variations of the model with respect to the nominal case. All simulations obtained respect the operating limits of the selected devices and components, for which its characteristics are reported in data sheets both in terms of electrical and thermal behaviour. Full article

Power Hardware-in-the-Loop (PHiL) simulation is an emerging testing methodology of real hardware equipment within an emulated virtual environment. The closed loop interfacing between the Hardware under Test (HuT) and the Real Time Simulation (RTS) enables a realistic simulation but can also result in an unstable system. In addition to fundamentals in PHiL simulation and interfacing, this paper therefore provides a consistent and comprehensive study of PHiL stability. An analytic analysis is compared with a simulative approach and is supplemented by practical validations of the stability limits in PHiL simulation. Special focus is given on the differences between a switching and a linear amplifier as power interface (PI). Stability limits and the respective factors of influence (e.g., Feedback Current Filtering) are elaborated with a minimal example circuit with voltage-type Ideal Transformer Model (ITM) PHiL interface algorithm (IA). Finally, the findings are transferred to a real low-voltage grid PHiL application with residential load and photovoltaic system. Full article

In CMOS integrated circuit (IC) design, operational amplifiers are one of the most useful active devices to enhance applications in analog signal processing, signal conditioning and so on. However, due to the CMOS technology downscaling, along the very large number of design variables and their trade-offs, it results difficult to reach target specifications without the application of optimization methods. For this reason, this work shows the advantages of performing many-objective optimization and this algorithm is compared to the well-known mono- and multi-objective metaheuristics, which have demonstrated their usefulness in sizing CMOS ICs. Three CMOS operational transconductance amplifiers are the case study in this work; they were sized by applying mono-, multi- and many-objective algorithms. The well-known non-dominated sorting genetic algorithm version 3 (NSGA-III) and the many-objective metaheuristic-based on the R2 indicator (MOMBI-II) were applied to size CMOS amplifiers and their sized solutions were compared to mono- and multi-objective algorithms. The CMOS amplifiers were optimized considering five targets, associated to a figure of merit (FoM), differential gain, power consumption, common-mode rejection ratio and total silicon area. The designs were performed using UMC 180 nm CMOS technology. To show the advantage of applying many-objective optimization algorithms to size CMOS amplifiers, the amplifier with the best performance was used to design a fractional-order integrator based on OTA-C filters. A variation analysis considering the process, the voltage and temperature (PVT) and a Monte Carlo analysis were performed to verify design robustness. Finally, the OTA-based fractional-order integrator was used to design a fractional-order chaotic oscillator, showing good agreement between numerical and SPICE simulations. Full article

Nowadays, the development, implementation and deployment of smart meters (SMs) is increasing in importance, and its expansion is exponential. The use of SMs in electrical engineering covers a multitude of applications ranging from real-time monitoring to the study of load profiles in homes. The use of wireless technologies has helped this development. Various problems arise in the implementation of SMs, such as coverage, locations without Internet access, etc. LoRa (long range) technology has great coverage and equipment with low power consumption that allows the installation of SMs in all types of locations, including those without Internet access. The objective of this research is to create an SM network under the LoRa specification that solves the problems presented by other wireless networks. For this purpose, a gateway for residential electricity metering networks using LoRa (GREMNL) and an electrical variable measuring device for households using LoRa (EVMDHL) have been created, which allow the development of SM networks with large coverage and low consumption. Full article

Due to the COVID-19 pandemic, most universities around the world started to employ distance-learning tools. To cope with these emergency conditions, some universities in Jordan have developed “mobile learning platforms” as a new tool for distance teaching and learning for students. This experience in Jordan is still new and needs to be evaluated in order to identify its advantages and challenges. Therefore, this study aims to investigate students’ perceptions about mobile learning platforms as well as to identify the crucial factors that influence students’ use of mobile learning platforms. An online quantitative survey technique using Twitter was employed to collect the data. A two-staged ANN-SEM modelling technique was adopted to analyze the causal relationships among constructs in the research model. The results of the study indicate that content quality and service quality significantly influenced perceived usefulness of mobile learning platforms. In addition, perceived ease of use and perceived usefulness significantly influenced behavioral intention to use mobile learning platforms. The study findings provide useful suggestions for decision makers, service providers, developers, and designers in the ministry of education as to how to assess and enhance mobile learning platform quality and understanding of multidimensional factors for effectively using mobile learning platforms. Full article

Quality-of-service (QoS) is the term used to evaluate the overall performance of a service. In healthcare applications, efficient computation of QoS is one of the mandatory requirements during the processing of medical records through smart measurement methods. Medical services often involve the transmission of demanding information. Thus, there are stringent requirements for secure, intelligent, public-network quality-of-service. This paper contributes to three different aspects. First, we propose a novel metaheuristic approach for medical cost-efficient task schedules, where an intelligent scheduler manages the tasks, such as the rate of service schedule, and lists items utilized by users during the data processing and computation through the fog node. Second, the QoS efficient-computation algorithm, which effectively monitors performance according to the indicator (parameter) with the analysis mechanism of quality-of-experience (QoE), has been developed. Third, a framework of blockchain-distributed technology-enabled QoS (QoS-ledger) computation in healthcare applications is proposed in a permissionless public peer-to-peer (P2P) network, which stores medical processed information in a distributed ledger. We have designed and deployed smart contracts for secure medical-data transmission and processing in serverless peering networks and handled overall node-protected interactions and preserved logs in a blockchain distributed ledger. The simulation result shows that QoS is computed on the blockchain public network with transmission power = average of −10 to −17 dBm, jitter = 34 ms, delay = average of 87 to 95 ms, throughput = 185 bytes, duty cycle = 8%, route of delivery and response back variable. Thus, the proposed QoS-ledger is a potential candidate for the computation of quality-of-service that is not limited to e-healthcare distributed applications. Full article

Blockchain-based electronic health system growth is hindered by privacy, confidentiality, and security. By protecting against them, this research aims to develop cybersecurity measurement approaches to ensure the security and privacy of patient information using blockchain technology in healthcare. Blockchains need huge resources to store big data. This paper presents an innovative solution, namely patient-centric healthcare data management (PCHDM). It comprises the following: (i) in an on-chain health record database, hashes of health records are stored as health record chains in Hyperledger fabric, and (ii) off-chain solutions that encrypt actual health data and store it securely over the interplanetary file system (IPFS) which is the decentralized cloud storage system that ensures scalability, confidentiality, and resolves the problem of blockchain data storage. A security smart contract hosted through container technology with Byzantine Fault Tolerance consensus ensures patient privacy by verifying patient preferences before sharing health records. The Distributed Ledger technology performance is tested under hyper ledger caliper benchmarks in terms of transaction latency, resource utilization, and transaction per second. The model provides stakeholders with increased confidence in collaborating and sharing their health records. Full article

Stochastic computing is an emerging scientific field pushed by the need for developing high-performance artificial intelligence systems in hardware to quickly solve complex data processing problems. This is the case of virtual screening, a computational task aimed at searching across huge molecular databases for new drug leads. In this work, we show a classification framework in which molecules are described by an energy-based vector. This vector is then processed by an ultra-fast artificial neural network implemented through FPGA by using stochastic computing techniques. Compared to other previously published virtual screening methods, this proposal provides similar or higher accuracy, while it improves processing speed by about two or three orders of magnitude. Full article

Early and objective autism spectrum disorder (ASD) assessment, as well as early intervention are particularly important and may have long term benefits in the lives of ASD people. ASD assessment relies on subjective rather on objective criteria, whereas advances in research point to up-to-date procedures for early ASD assessment comprising eye-tracking technology, machine learning, as well as other assessment tools. This systematic review, the first to our knowledge of its kind, provides a comprehensive discussion of 30 studies irrespective of the stimuli/tasks and dataset used, the algorithms applied, the eye-tracking tools utilised and their goals. Evidence indicates that the combination of machine learning and eye-tracking technology could be considered a promising tool in autism research regarding early and objective diagnosis. Limitations and suggestions for future research are also presented. Full article

This paper presents the design flow of an advanced non-linear control strategy, able to absorb the effects that the main causes of torque oscillations, concerning synchronous electrical drives, cause on the positioning of the end-effector of a manipulator robot. The control technique used requires an exhaustive modelling of the physical phenomena that cause the electromagnetic torque oscillations. In particular, the Cogging and Stribeck effects are taken into account, whose mathematical model is incorporated in the whole system of dynamic equations representing the complex mechatronic system, formed by the mechanics of the robot links and the dynamics of the actuators. Both the modelling procedure of the robot, directly incorporating the dynamics of the actuators and the electrical drive, consisting of the modulation system and inverter, and the systematic procedure necessary to obtain the equations of the components of the control vector are described in detail. Using the Processor-In-the-Loop (PIL) paradigm for a Cortex-A53 based embedded system, the beneficial effect of the proposed advanced control strategy is validated in terms of end-effector position control, in which we compare classic control system with the proposed algorithm, in order to highlight the better performance in precision and in reducing oscillations. Full article

This study is supposed to introduce a valid strategy for increasing the sensitivity of chipless radio frequency identification (RFID) encoders. The idea is to properly select the dielectric substrate in order to enhance the contribution of the sensitive layer and to maximize the frequency shift of the resonance peak. The specific case of a chipless sensor suitable for the detection of humidity in low-humidity regimes will be investigated both with numerical and experimental tests. Full article

Marine knowledge is such an important part of education that it has been integrated into various subjects and courses across educational levels. Previous research has indicated the importance of AR assisted students’ learning during the learning process. This study proposed the AR Oyster Learning System (AROLS) that integrates mobile AR with a marine education teaching strategy for teachers in primary schools. To evaluate the effectiveness of the proposed approach, an experiment was conducted in a primary school natural science course regarding oysters. The participants consisted of 22 fourth-grade students. The experimental group comprised 11 students who learned with the AROLS learning approach and the control group comprised 11 students who learned with the conventional multimedia learning approach. The results indicate that (1) students were interested in the AR learning approach, (2) students’ learning achievement and motivation were improved by the AR learning approach, (3) students acquired the target knowledge through the oyster course, and (4) students learned the importance of sustainability when taking online courses at home during the pandemic. Full article

This paper investigates the operation of each power conversion system (PCS) for efficient energy management systems (EMSs) of microgrids (MGs). When MGs are linked to renewable energy sources (RESs), the reduction in power conversion efficiency can be minimized. Furthermore, energy storage systems (ESSs) are utilized to manage the surplus power of RESs. Thus, the present work presents a method to minimize the use of the existing power grid and increase the utilization rate of energy generated through RESs. To minimize the use of the existing power grid, a PCS operation method for photovoltaics (PV) and ESS used in MGs is proposed. PV, when it is directly connected as an intermittent energy source, induces voltage fluctuations in the distribution network. Thus, to overcome this shortcoming, this paper utilizes a system that connects PV and a distributed energy storage system (DESS). A PV-DESS integrated module is designed and controlled for tracking constant power. In addition, the DESS serves to compensate for the insufficient power generation of PV. The main energy storage systems (MESSs) used in MGs affect all aspects of the power management in the system. Because MGs perform their operations based on the capacity of the MESS, a PCS designed with a large capacity is utilized to stably operate the system. Because the MESS performs energy management through operations under various load conditions, it must have constant efficiency under all load conditions. Therefore, this paper proposes a PCS operation algorithm with constant efficiency for the MESS. Utilizing the operation algorithm of each PCS, this paper describes the efficient energy management of the MG and further proposes an algorithm for operating the existing power grid at the minimum level. Full article

With the advent of technological marvels like global digitization, the prediction of the stock market has entered a technologically advanced era, revamping the old model of trading. With the ceaseless increase in market capitalization, stock trading has become a center of investment for many financial investors. Many analysts and researchers have developed tools and techniques that predict stock price movements and help investors in proper decision-making. Advanced trading models enable researchers to predict the market using non-traditional textual data from social platforms. The application of advanced machine learning approaches such as text data analytics and ensemble methods have greatly increased the prediction accuracies. Meanwhile, the analysis and prediction of stock markets continue to be one of the most challenging research areas due to dynamic, erratic, and chaotic data. This study explains the systematics of machine learning-based approaches for stock market prediction based on the deployment of a generic framework. Findings from the last decade (2011–2021) were critically analyzed, having been retrieved from online digital libraries and databases like ACM digital library and Scopus. Furthermore, an extensive comparative analysis was carried out to identify the direction of significance. The study would be helpful for emerging researchers to understand the basics and advancements of this emerging area, and thus carry-on further research in promising directions. Full article

Diabetes is a chronic disease that can affect human health negatively when the glucose levels in the blood are elevated over the creatin range called hyperglycemia. The current devices for continuous glucose monitoring (CGM) supervise the glucose level in the blood and alert user to the type-1 Diabetes class once a certain critical level is surpassed. This can lead the body of the patient to work at critical levels until the medicine is taken in order to reduce the glucose level, consequently increasing the risk of causing considerable health damages in case of the intake is delayed. To overcome the latter, a new approach based on cutting-edge software and hardware technologies is proposed in this paper. Specifically, an artificial intelligence deep learning (DL) model is proposed to predict glucose levels in 30 min horizons. Moreover, Cloud computing and IoT technologies are considered to implement the prediction model and combine it with the existing wearable CGM model to provide the patients with the prediction of future glucose levels. Among the many DL methods in the state-of-the-art (SoTA) have been considered a cascaded RNN-RBM DL model based on both recurrent neural networks (RNNs) and restricted Boltzmann machines (RBM) due to their superior properties regarding improved prediction accuracy. From the conducted experimental results, it has been shown that the proposed Cloud&DL-based wearable approach achieves an average accuracy value of 15.589 in terms of RMSE, then outperforms similar existing blood glucose prediction methods in the SoTA. Full article

Recently, the emergence of the COVID-19 has caused a high acceleration towards the use of mobile learning applications in learning and education. Investigation of the adoption of mobile learning still needs more research. Therefore, this study seeks to understand the influencing factors of mobile learning adoption in higher education by employing the Information System Success Model (ISS). The proposed model is evaluated through an SEM approach. Subsequently, the findings show that the proposed research model of this study could explain 63.9% of the variance in the actual use of mobile learning systems, which offers important insight for understanding the impact of educational, environmental, and quality factors on mobile learning system actual use. The findings also indicate that institutional policy, change management, and top management support have positive effects on the actual use of mobile learning systems, mediated by quality factors. Furthermore, the results indicate that factors of functionality, design quality, and usability have positive effects on the actual use of mobile learning systems, mediated by student satisfaction. The findings of this study provide practical suggestions, for designers, developers, and decision makers in universities, on how to enhance the use of mobile learning applications and thus derive greater benefits from mobile learning systems. Full article

In 2020 and 2021, humanity lived in fear due to the COVID-19 pandemic. However, with the development of artificial intelligence technology, mankind is attempting to tackle many challenges from currently unpredictable epidemics. Korean society has been exposed to various infectious diseases since the Korean War in 1950, and to overcome them, the six most serious cases in National Notifiable Infectious Diseases (NNIDs) category I were defined. Although most infectious diseases have been overcome, viral hepatitis A has been on the rise in Korean society since 2010. Therefore, in this paper, the prediction of viral hepatitis A, which is rapidly spreading in Korean society, was predicted by region using the deep learning technique and a publicly available dataset. For this study, we gathered information from five organizations based on the open data policy: Korea Centers for Disease Control and Prevention (KCDC), National Institute of Environmental Research (NIER), Korea Meteorological Agency (KMA), Public Open Data Portal, and Korea Environment Corporation (KECO). Patient information, water environment information, weather information, population information, and air pollution information were acquired and correlations were identified. Next, an epidemic outbreak prediction was performed using data preprocessing and 3D LSTM. The experimental results were compared with various machine learning methods through RMSE. In this paper, we attempted to predict regional epidemic outbreaks of hepatitis A by linking the open data environment with deep learning. It is expected that the experimental process and results will be used to present the importance and usefulness of establishing an open data environment. Full article

Ontologies, and especially formal ones, have traditionally been investigated as a means to formalize an application domain so as to carry out automated reasoning on it. The union of the terminological part of an ontology and the corresponding assertional part is known as a Knowledge Graph. On the other hand, database technology has often focused on the optimal organization of data so as to boost efficiency in their storage, management and retrieval. Graph databases are a recent technology specifically focusing on element-driven data browsing rather than on batch processing. While the complementarity and connections between these technologies are patent and intuitive, little exists to bring them to full integration and cooperation. This paper aims at bridging this gap, by proposing an intermediate format that can be easily mapped onto the formal ontology on one hand, so as to allow complex reasoning, and onto the graph database on the other, so as to benefit from efficient data handling. Full article

In this paper, we develop end-to-end autonomous driving based on a 2D LiDAR sensor and camera sensor that predict the control value of the vehicle from the input data, instead of modeling rule-based autonomous driving. Different from many studies utilizing simulated data, we created an end-to-end autonomous driving algorithm with data obtained from real driving and analyzing the performance of our proposed algorithm. Based on the data obtained from an actual urban driving environment, end-to-end autonomous driving was possible in an informal environment such as a traffic signal by predicting the vehicle control value based on a convolution neural network. In addition, this paper solves the data imbalance problem by eliminating redundant data for each frame during stopping and driving in the driving environment so we can improve the performance of self-driving. Finally, we verified through the activation map how the network predicts the vertical and horizontal control values by recognizing the traffic facilities in the driving environment. Experiments and analysis will be shown to show the validity of the proposed algorithm. Full article

This review paper explores existing architectures, operating principles, performance metrics and applications of analog Gaussian function circuits. Architectures based on the translinear principle, the bulk-controlled approach, the floating gate approach, the use of multiple differential pairs, compositions of different fundamental blocks and others are considered. Applications involving analog implementations of Machine Learning algorithms, neuromorphic circuits, smart sensor systems and fuzzy/neuro-fuzzy systems are discussed, focusing on the role of the Gaussian function circuit. Finally, a general discussion and concluding remarks are provided. Full article