Search

This article presents a software platform (SP) that helps generate convolutional and feed-forward neural networks and a parametric identification algorithm is proposed for a Cartesian robot using convolutional neural networks (CNN) generated and trained with the proposed SP. The user interface is friendly and aids in developing convolutional and feed-forward neural networks. The SP is made in LAB VIEW®. The user does not need the complete knowledge of mathematics used to perform the training or execution of the network or the programming skills to perform the code that performs such tasks. The SP is designed to reduce the deployment time of neural networks, providing the user with a graphical interface that guides the user through the formation and development of the neural network, this is one of the main contributions. The SP interface is described in this article, showing all the options offered. Five neural networks are presented to prove and demonstrate the software platform’s use. The dynamic models of the Cartesian robot are used; one has six parameters and the other ten parameters. The maximum similarity achieved by the identification algorithm with the six-parameter dynamic model was 99.15% and 99.92% for the ten-parameter model; both are numerical torque analyses. A real Cartesian robot was identified, and five experimental torques were reconstructed for each axis; the maximum similarity was 97.87%. The proposed algorithm is compared with least squares, which obtain 95.81%. Therefore, the results show that the proposed method generates better parametric identification.

In recent times, there has been a growing interest in the domain of computationally challenging problem solving within both scientific and organizational contexts. This study is primarily concerned with the extraction and comprehension of the methodologies and strategies employed by individuals when confronted with intricate problems, specifically those falling under the purview of NP-hard problems. The Facility Location Problem (FLP) serves as a prominent exemplar within this study’s framework. Traditionally, the handling of such complex problems has leaned upon intuitive reasoning and visual perception as the primary tools. However, these conventional approaches tend to provide only limited insight into the underlying processes employed in solving such problems. The present research seeks to bridge this knowledge gap through the utilization of advanced machine learning techniques for the purpose of categorizing and scrutinizing the strategies deployed by individuals in their attempts to tackle computationally challenging problems. The analysis conducted as part of this study unveils discernible and well-defined patterns and strategies that are employed by participants, some of whom have achieved notable levels of success. Remarkably, in certain instances, the outcomes achieved by these individuals have demonstrated a competitive edge when compared to the results produced by sophisticated computational methods, such as genetic algorithms. A fundamental component of our research methodology involves the application of heatmaps and clustering techniques. Through the normalization of results, our findings distinctly delineate two primary categories of games: those characterized by uniform player strategies and those characterized by a multitude of diverse and individualized tactics. Furthermore, our research employs a systematic approach to represent games by clustering them based on inherent similarities, utilizing cosine similarity as a metric for this purpose. By computing the averages of vectors within each cluster, we derive centroids that encapsulate the central tendencies exhibited by games belonging to that cluster. These centroids are then visually presented in a three-dimensional format, complemented by proportional spheres. These visual representations serve to vividly illustrate the dispersion and influence associated with each cluster. Our research significantly contributes to the understanding of human problem-solving strategies when confronted with computationally challenging problems. It unearths valuable insights regarding the potential for harnessing human intuition and expertise in addressing complex computational challenges. Through the integration of machine learning methodologies and intuitive visualizations, this work advances our comprehension of the approaches individuals employ to excel in solving computationally intricate problems.

The Vehicular Ad Hoc Networks paradigm, i.e., VANETs, is a promising technology to enable the Internet of Vehicles for future integration with 5G and 6G infrastructures in smart cities. However, VANETs can be susceptible to wireless interference, such as co-channel interference, which can cause communication quality and performance to deteriorate, threatening vehicular safety applications, including traffic management and accident prevention. Hence, VANET requires practical interference mitigation strategies, such as digital beamforming, to mitigate signals from unwanted directions while improving the reception from desired directions with adequate precision. In this paper, a review of the literature on digital beamforming techniques is carried out. Furthermore, a significant contribution is the development of an experimental platform for a roundabout VANET based on simulations, where different digital beamforming algorithms, such as phase-shift and Minimum-Variance Distortionless-Response, i.e., MVDR, are tested. The proposed simulation platform ensures a fair comparison of the digital beamforming methods by evaluating them under identical conditions. The results show that the phase-shift algorithm leads to improvement due to its dynamics and precise beam control in ad hoc environments, with co-channel interference as the predominant impairment, achieving a reduction of the outage compared to non-beamforming and MVDR beamforming scenarios. Therefore, digital beamforming can improve the performance and reliability of VANETs, offering the potential for advanced wireless network implementations in smart cities with 5G and 6G infrastructures.

Human activity recognition (HAR) has applications ranging from security to healthcare. Typically these systems are composed of data acquisition and activity recognition models. In this work, we compared the accuracy of two acquisition systems: Inertial Measurement Units (IMUs) vs Movement Analysis Systems (MAS). We trained models to recognize arm exercises using state-of-the-art deep learning architectures and compared their accuracy. MAS uses a camera array and reflective markers. IMU uses accelerometers, gyroscopes, and magnetometers. Sensors of both systems were attached to different locations of the upper limb. We captured and annotated 3 datasets, each one using both systems simultaneously. For activity recognition, we trained 8 architectures, each one with different operations and layers configurations. The best architectures were a combination of CNN, LSTM, and Transformer achieving test accuracy from 89% to 99% on average. We evaluated how feature selection reduced the sensors required. We found IMU and MAS data were able to distinguish correctly the arm exercises. CNN layers at the beginning produced better accuracy on challenging datasets. IMU had advantages over other acquisition systems for activity recognition. We analyzed the relations between models accuracy, signal waveforms, signals correlation, sampling rate, exercise duration, and window size. Finally, we proposed the use of a single IMU located at the wrist and a variable-size window extraction.

Reservoir computing (RC), first applied to temporal signal processing, is a recurrent neural network in which neurons are randomly connected. Once initialized, the connection strengths remain unchanged. Such a simple structure turns RC into a non-linear dynamical system that maps low-dimensional inputs into a high-dimensional space. The model’s rich dynamics, linear separability, and memory capacity then enable a simple linear readout to generate adequate responses for various applications. RC spans areas far beyond machine learning, since it has been shown that the complex dynamics can be realized in various physical hardware implementations and biological devices. This yields greater flexibility and shorter computation time. Moreover, the neuronal responses triggered by the model’s dynamics shed light on understanding brain mechanisms that also exploit similar dynamical processes. While the literature on RC is vast and fragmented, here we conduct a unified review of RC’s recent developments from machine learning to physics, biology, and neuroscience. We first review the early RC models, and then survey the state-of-the-art models and their applications. We further introduce studies on modeling the brain’s mechanisms by RC. Finally, we offer new perspectives on RC development, including reservoir design, coding frameworks unification, physical RC implementations, and interaction between RC, cognitive neuroscience and evolution.

Biomimetic Autonomous Underwater Vehicles (BAUVs) navigate aquatic environments by mimicking natural propellants from fish species. These vehicles move part(s) of their bodies using various mechanisms to propel and swim forward or laterally. Their main goal is to follow and adjust defined paths to reach a target autonomously. Local path planning is of paramount importance during navigation tasks due to unexpected obstacles. Moreover, path planning strategies should consider the environment’s information obtained by the vehicle during its mission, as well as its dynamics and mechanical limitations, to define new routes properly. This article presents the development of a waypoint generator based on the $\text{D}\ast $ Lite algorithm. The proposed planner considers a frontal-short-sighted and tail-actuated BAUV with motion constraints to adjust the vehicle’s path towards a target coordinate. By identifying obstacles, the planner adjusts and defines inner waypoints inside the vehicle’s vision range by considering closeness to obstacles found and BAUV’s current position. The developed strategy reduces collision risks due to the discrimination of nodes near obstacles, prioritizing broad hallways and safer swimming distances between the vehicle’s current position and inner waypoints. The effectiveness of the proposed algorithm is simulated using the BAUV’s hydrodynamics model and by adding a waypoint tracking controller to correct the vehicle’s swimming performance inside three scenarios. The vehicle can reach the goal by properly defining inner waypoints while safely avoiding collisions, narrow hallways, and sharp turns.

Classification is a commonly used technique in data mining and is applied in various fields such as sentiment analysis, fraud detection, and fault diagnosis. Multiclass classification, which involves more than two classes, is more complex than binary classification. There are mainly two ways to approach multiclass classification, one is to expand the binary classifier into a multiclass classifier through various strategies and the other is to divide the multiclass classification problem into multiple binary problems (binarization). Two popular approaches for binarization are One vs One (OvO) and One vs All (OvA). It is simpler to aggregate the outputs of all binary classifiers as the number of classifiers decreases. However, it causes an imbalance of positive and negative sample numbers, which affects the classification effect of each binary classifier. In this article, we contribute to the field of ensemble learning and multi-class classification by proposing a new method called Ensemble Partition Sampling (EPS). This article presents a new approach to multiclass classification using an “Ensemble Partition Sampling” method within the “one-vs-all” (OvA) framework. The primary goal of this method is to tackle the problem of data imbalance by incorporating ensemble learning and preprocessing techniques into each binary dataset. The study found that Ensemble Partition Sampling (EPS) is the most effective method for imbalanced and multiclass imbalanced classification, outperforming other methods including OvA, SMOTE, k-means-SMOTE, Bagging-RB, DES-MI, OvO-EASY, and OvO-SMB. The study used CART, Random Forest, and SVM as classifiers, and the results consistently showed that EPS outperformed all other algorithms. Based on the findings, it can be concluded that the EPS approach is a highly effective method for improving classification performance in imbalanced and multiclass imbalanced datasets.

Slow-read Distributed Denial of Service (DDoS) attacks are complex to detect and mitigate. Although existing tools allow one to identify these attacks, these tools mainly generate alerts. However, in real scenarios, a large number of attack detection alerts will put the security workforce in a bottleneck, as they will not be able to implement mitigation actions in a complete and timely manner. Furthermore, since most existing security solutions for DDoS attack mitigation are tested using datasets and simulated scenarios, their applicability to production networks could be unfeasible or ineffective due to possibly incomplete assumptions in their design. Therefore, automated security solutions against DDoS attacks are needed not only to be designed but also to be implemented and evaluated in real scenarios. This study presents a Software-Defined Networking (SDN)-based security framework, which automates the monitoring, detection, and mitigation of slow-rate DDoS attacks. The framework is implemented in a physical network that uses equipment from the European Experimental Facility Smart Networks for Industry (SN4I). The results demonstrate that the framework effectively mitigates malicious connections, with a mitigation efficiency between 91.66%– 100% for different conditions of the number of attackers and victims. In addition, the SDN-SlowRate-DDoS dataset is presented, which contains multiple experiments of slow-rate DDoS attacks performed on the real testbed. The resources provided in this security dataset are useful to the scientific and industry communities in designing and testing realistic solutions for intrusion detection systems.

This paper presents our experience implementing Collaborative Working Spheres (CWS) in the context of an international undergraduate course on Global Software Development (GSD) during the COVID-19 pandemic. Many universities around the world increased their efforts in creating educational alternatives for adequately addressing the educational challenges that this pandemic has introduced. A particular case is the training of software skilled graduates to work in globally distributed environments because learning this topic requires a lot of practical work when student motivation could have been negatively affected by the COVID-19 pandemic. However, providing highly practical GSD courses during a pandemic is a challenging task for many of these universities. It is against this backdrop that we have developed an educational tool to provide CWS, enabling undergraduate students to acquire practical experience in GSD and improve their communication and teamworking skills, even during the COVID-19 pandemic. An international empirical evaluation was conducted involving students and teachers from seven universities in different countries around the world. The obtained results showed that our approach can make a significant contribution to the development of practical projects on undergraduate GSD courses with students developing their knowledge and social skills associated with this topic. The data collected on the teachers’ perceptions suggested that our approach could also be useful in introducing the GSD approach at undergraduate level when social distancing is in place.

The lock-in amplifier (LIA) instruments are designed to provide signal conditioning for precision measurement systems to extract signals from extremely noisy environments. The digital LIAs design often requires a verification process to ensure hardware performance. Thus, hardware description language (HDL) with functional verification strategies offers a powerful tool to provide an field-programmable gate array (FPGA) integrated solution. In this letter, we propose a methodology of design and verification of all-digital LIA and an additive white Gaussian noise (AWGN) module able to measure extremely lower levels of signal-to-noise ratio (SNR) of $\approx $ ${10}^{-{15}}$ or down to −37 dB while a wide reserve of spurious-free dynamic range (SFDR) up to 90 dB on FPGA is ensured. To this end, the designed and implemented FPGA framework for quick, accurate, and comprehensive characterization of a given digital LIA is used to leverage the capabilities of the design under controllable AWGN noise patterns stimulus. [ABSTRACT FROM AUTHOR]

This article presents a power-dense, six-phase dual-interleaved DC-DC Buck-Boost converter for electric/hybrid vehicle applications. This converter uses six hard-switched arms at 75 kHz, three dual-interleaved interphase transformers together with three common inductors and two filter capacitors at the input and output. The steady-state current ripple frequency of the common inductors is twice the switching frequency due to the dual interleaving in the interphase transformers. The resultant ripple current frequency of the input and output filters is six times the switching frequency; size and weight reduction of the passive components are obtained due to this frequency increase. The design and construction details of a 32 kW, 315–385 V supply, SiC-based prototype are presented along with experimental investigation. The prototype converter was successfully operated up to full load with a 350 V output, being the phase currents relatively balanced in the continuous conduction mode without a closed-loop control system. The prototype efficiency is 98 % at full load, being the gravimetric power density 7.56 kW/kg.

This paper proposes a general methodology for the development of power system models suitable for distributed real-time simulations (D-RTS) based on topology, simulator interfaces and data exchange. D-RTS have risen as functional alternatives that can combine remote and multi-vendor resources for large-scale power system simulations via a virtual connection. However, previous work focused on combination of separate models and not the performance of D-RTS when splitting a single monolithic model, failing to study the behaviors of D-RTS, including, synchronization and accuracy. The proposed methodology is used to develop distributed models of two widely used testbed large power systems, the IEEE Australian Benchmark model and the IEEE 300-Bus system. These testbeds are selected as they can be simulated as both monolithic and distributed models in available simulators in order to validate both the methodology and resulting model performance. The obtained results and comparison between monolithic and distributed models support the proposed approach and demonstrate the performance of D-RTS under both steady-state and transient operations in multiple scenarios.

For decades, researchers have experimented with the possibility that machines can equal human linguistic capabilities. Recently, advances in the field of natural language processing (NLP) as well as a substantial increase in available naturally occurring linguistic data on social media platforms have made more advanced methodologies such as sentiment analysis (SA) gain substantial momentum on contemporary applications. This document compiles what the authors consider to be some of the most important concepts related to SA, as well as techniques and processes necessary for the various stages of its implementation. Furthermore, specific applications related to the extraction and classification of social media data using novel SA techniques are presented and quantified, with an emphasis on those used for the identification of mental health degradation during the COVID-19 pandemic. Finally, the authors present several conclusions highlighting the most prominent benefits and drawbacks of the methods discussed, followed by a brief discussion of possible future applications of certain methods of interest.

Traditionally, the estimation of water quality is realized through laboratory analysis which is time-consuming and requires specialized installations, equipment, and personnel. Nowadays, it is possible to make real-time water monitoring through electrochemical sensors, microcontrollers, and central processing units to detect water pollutants. This work proposes a system based on the Arduino platform for monitoring parameters associated with water quality, such as oxidation-reduction potential, pH, total dissolved solids, turbidity, temperature, electrical conductivity, and dissolved oxygen. A critical criterion for the sensor selection was its cost and availability, resulting in sensors from different companies. They were integrated without much complexity, thanks to the selected platform. In addition, a proposal is made for a signal conditioning circuit for the oxidation-reduction potential electrode. A stage of filtering is added to the pH and turbidity commercial circuits to improve their performance. Remote access to the data is done through a mini-PC with WIFI connectivity and a MySQL database. All the sensors were calibrated with reference solutions or against other commercial meters. Through the proposed system, time series having a sampling period of 20 s of all parameters were recorded for more than a week-long exhibiting circadian patterns for the same water sample. Pearson correlation for the parameters was carried on. The results show that the system successfully monitored the seven physicochemical variables through low-cost sensors. It also has remote access capabilities.

Wireless communication systems working in millimeter-wave (mmWave) frequency bands offer higher bandwidths than traditional radio frequency schemes. This technology allows multibeam steering and data multiplexing with the help of massive multiple-input multiple-output (MIMO) systems. However, supporting large bandwidths at mmWave frequencies is challenging due to the use of large antenna arrays with beamforming, sampling signals with large bandwidths, and baseband signal processing operations at gigabit data rates. Due to the wider bandwidth and higher signal processing requirements of mmWave systems, low-complexity receiver algorithms become important. Previously reported investigations assumed the use of hybrid beamforming structures that reduce power consumption and signal processing tasks. Therefore, the use of artificial neural networks (ANNs) becomes relevant for the processing of mmWave signals as reported in earlier works. In this article, to carry out MIMO combining processing for mmWave communications, we propose a fully complex multilayer extreme learning machine (M-ELM) neural network. We investigate the tuning of the number of neurons in each hidden layer for the proposed method to maximize the system performance and decrease the complexity of the receiver. We compare the results of the introduced M-ELM algorithm with a fully complex extreme learning machine (ELM), fully real ELM, and M-ELM defined in the real plane in terms of spectral efficiency, bit error rate, computational complexity, and processing time. Furthermore, we compare the novel M-ELM strategy with traditional linear MIMO receivers, such as Maximum Ratio and Minimum Mean Square Error, as well as to a multilayer perceptron (MLP) neural network trained offline. The numerical results show that with a good balance between the overall performance and computational cost of the ANN, the fully complex M-ELM MIMO receiver outperforms the other evaluated schemes.

Objective: To record and characterize features of levator ani muscles (LAM) activity in pregnancy and postpartum using non-invasive and novel Magnetomyography (MMG) technique with amplitude and spectral parameters. Methods: Nulliparous women with uncomplicated pregnancies participated in the MMG data collection during rest and voluntary LAM contractions (Kegels) with modulated intensity in third trimester and approximately 2 months postpartum (PP). Simultaneous surface electromyography was recorded to document the recruitment of accessory muscles. Moderate strength Kegel (MK) MMG trials were analyzed. Amplitude and spectral parameters including root-mean square (RMS) amplitude, power spectrum density (PSD) and normalized PSD (rPSD) in three frequency bands (low, middle, high) were computed on MK epochs. Statistical comparisons between pregnancy and postpartum were calculated. Results: MMG recordings were measured from 10 pregnant women. Results showed decreased RMS and power between third trimester and postpartum, trending towards significance. rPSD values in the low frequency band decreased significantly (p = 0.028) from third trimester to postpartum, while significant increase was observed in the middle frequency band (p = 0.018). Conclusions: This study shows that MMG as non-invasive tool has the ability to detect and characterize changes of LAM activity with amplitude and spectral parameters during pregnancy and postpartum.

A novel position location technique that requires a reduced number of scattered observation signals is presented in this paper. This technique infers the mobile’s location based on measurements of angle of arrival (AOA), path length obtainable by time of arrival (ToA) measurements and Doppler Shifted Frequency (DSF). Based upon the principle of same Doppler collinearity, the proposed methodology describes feasible locating regions for scatterers which are collinear with the transmitter. The novelty and advantage of the proposed technique makes it reliable and efficient from an optimality perspective since the location of the mobile is estimated using a single base station in a non-cooperative wireless networking scenario. Therefore, the main contribution is the development of a parametric model capable to provide a feasibility location region in non cooperative scenarios.

In this study, a modified linear technique is proposed for the controllability and observability of robotic arms, the modified linear technique consists of the following steps: a transformation is used to rewrite a nonlinear time-variant model as a quasi-linear time-variant model, this quasi-linear time-variant model is evaluated at origin to obtain a linear time-invariant model, and the rank condition tests the controllability and observability of the linear time-invariant model. The modified linear technique is better than the linearization technique because the modified linear technique does not use the Jacobian approximation, while the linearization technique needs the Jacobian approximation. The modified linear technique is better than the linear technique because the modified linear technique can be applied to robotic arms with rotational and prismatic joints, while the linear technique can only be applied to robotic arms with rotational joints.

Homicide prediction is a challenging task due to the spatio-temporal sparsity of these crime events. In this paper we report the results of using several approaches to mitigate this sparsity condition in machine learning models specially tailored towards modeling homicides events. Since spatial resolution is a direct determinant of sparsity, we focus on the performance of these models across different resolutions of interest to police authorities. We use a simple count model as benchmark and propose some enhancements of it directed towards improving prediction performance. We then compare the results to more complex models motivated by manifold learning and graph signal processing methods. We found that the simple benchmark models are as good as state of the art models for low resolution, but, as resolution increases, the performance of machine learning models outperform the benchmark. These results provide a rationality for the use of state of the art machine learning models for homicide prediction at the high resolution of interest for the deployment of police resources.

Link budgets are widely applied to evaluate communication links for low Earth Orbit (LEO) satellites. However, approaches to calculate the received power from LEO satellites have followed similar procedures to those for Geostationary (GEO) satellites and other fixed distance wireless systems, ignoring the satellite mobility that causes continuous changes in the path length and in the elevation angle. Link budgets found in the literature for LEO communication systems have commonly opted to characterize the best and worst-cases of the received signal; however, this common approach tells little about how often those cases occur, and little can be inferred about the expected received power and its measures of dispersion. This article introduces an innovative methodology to evaluate LEO link budgets using the long term statistics and probabilities of occurrence of the elevation angle, which is characterized in this work through a random variable. This characterization of the elevation angle through a random variable allows the calculation of the expected value of the received power, its standard deviation, quantiles, among other quantities. The received power is essential to calculate other link indicators such as the carrier-to-noise ratio (C/N), and with the proposed methodology it can be now calculated considering the path length and elevation angle variability that occur for LEO satellites.

The pressure swing adsorption (PSA) process, is a novel method for the purification and production of bioethanol. A highly non-linear rigorous model is implemented to simulate the cyclic dynamics of PSA, achieving purity of 99% wt of ethanol, which meets international standards to be used as fuel. The contribution of this work focuses on obtaining an identified model capable of capturing the important dynamics of the PSA process (with a fit above of 90%) and to be used for controller design purposes, since it is very complicated to design control in highly nonlinear models that are represented with partial differential equations (PDE). For proof of concept, a comparison between Hammersetein-Wienner and Artificial Neural-Networks showed the relevance of the proposed method, using the same input and output signals. Both identified models capture the important dynamics of the rigorous PSA model.

The perturbations are the unwanted and unknown inlets in nonlinear plants which can affect the outlets. In this article, an estimator is studied for the variables and perturbations estimation in nonlinear plants. The saturation map is used in our estimator instead of the signum map to decrease the chattering, and we ensure the estimator convergence by the Lyapunov analysis. The conditions required by our estimator gains are found to reach the variables error convergence, and these gains are used for the perturbations estimation. An algorithm is proposed to choose the gains for achieving a satisfactory performance in our estimator. The studied estimator is applied for the variables and perturbations estimation in the gas turbine and gasification plants.

The phenomenon of non-linearity is the main problem of a DC motor and optimum performance cannot be obtained by the calculation of the controller's parameters using conventional methods. However, a DC motor is considered an extremely common device by the low-cost and effective dynamic response in various applications. Thus, it has been a subject for research studies to take advantage of its maximum performance. This manuscript proposes an experimental methodology that consists of the following: The DC motor's characterization method for finding the ideal frequency. The design of the Firmware-based Pulse Width Modulation (PWM) generating module and the P, PD, PID controller's implementation in an own FPGA-based programmable microprocessor to obtain almost the same performance as a servo-amplifier commercial of direct-drive. The PWM is a technique widely used to regulate the speed of rotation of a DC motor, in this case, the duty cycle of the PWM is used to provide the torque necessary to the mechanics of the system in order to look for a linear relationship but using the right frequency of the characterized DC motor. Finally, based on a built prototype of a micro-positioning system using the characterized motors, and the mathematical model, in both cases the three controllers were applied in order to establish the comparison between the responses, seeking to observe if the experimental results show a great difference with respect to the simulation results. The main aim of this study is to show that the proposed methodology works. However, since there was no significant difference in both results, motors used in the closed-loop control present approximately the same linear response as that of the motor model used in the simulation.

In this paper, we compare two artificial neural networks (ANNs) approaches designed to perform channel equalization for millimeter-wave (mmWave) signals operating in the 28 GHz frequency band. We used an in-house deterministic Three-Dimensional Ray-Launching (3D-RL) code to simulate the spatial structure of mmWave channels considering the material properties of the obstacles within the scenario at the frequency under analysis. We performed offline training of a multilayer perceptron (MLP) neural network with the simulated mmWave channels to equalize the received signal. We also performed online training of an extreme learning machine (ELM) neural network to directly get the equalized symbols at the receiver, given as input the received mmWave signal. The ANN solutions were tested in terms of the achievable spectral efficiency, bit error rate, and time to process. We compared the ANN techniques to the minimum mean square error and the zero-forcing equalizers, considering an orthogonal frequency-division multiplexing communication based on the 5G New Radio standard. We present numerical results on the performance of the proposed ANNs and show that the ELM strategy outperforms the MLP method, requiring significantly less processing time than the reviewed equalization methods.

This study aims to find out how useful managers' past general and export experience is in predicting whether young manufacturing firms become fast internationalizers. Extant literature about the role of managerial experience in determining young firms' internationalization type is scant. This paper fills this gap by providing systematic evidence on which kinds of general and export experience can be used for accurate predictions of two firm types: born globals and general fast internationalizers. Our dataset encompasses information about managerial experience of the whole population of young Estonian manufacturing firms. Based on using four different prediction methods (logistic regression, rough sets, decision tree, neural networks) and a large variety of variables reflecting managers' past experience, the results indicate that in prediction models, export experience variables are more valuable than general experience variables. Born globals can be predicted with an accuracy of at least 90% in case of all applied machine learning methods, while the precision is lower in case of general fast internationalizers. The study leads to important implications for international business theory and practice.

Currently, smart buildings generate large amounts of data due to the many devices and equipment available. Hence, buildings implement building management systems (BMSs), which monitor, control, manage and analyze each of these components. However, current BMSs are incapable of managing a massive amount of data (big data) and therefore cannot extract knowledge or make intelligent decisions in quasi real time. In addition, there are serious limitations to integrating BMSs with other services since they generally use proprietary software. In this sense, service-oriented architecture (SOA) is an architectural style that allows one to build distributed systems and provide functionalities such as services to end users or other types of services. Therefore, an SOA has the great advantage of allowing the expansion of the functionalities of BMSs. In fact, there are several studies that address SOAs for building management. However, we have not found any description or systematic analysis in the literature that allows the development of a versatile and interoperable SOA focused on the energy efficiency of buildings and that can integrate massive data analysis features. For these reasons, this study seeks to fill this knowledge gap and, more specifically, to identify and analyze the various software requirements proposed in the literature and the characteristics of big data that allow for improving the energy efficiency of buildings. To this end, we performed an in-depth review of the literature according to the methodology proposed by Kitchenham. As a result of this review, we provide researchers with a specific vision of the requirements and characteristics to consider for software development aimed at the energy efficiency of unique or historic buildings.

The purpose of the this research work was to evaluate the impact of social support (through its two dimensions: co-worker support and supervisor support) on the reduction of the burnout syndrome (considering its three dimensions: emotional exhaustion, professional efficacy, and cynicism) from two proposed models that show clear differences. Social support was assessed using the Job Content Questionnaire instrument, and the burnout syndrome was measured through the Maslach Burnout Inventory - General Survey. A total of 467 middle and senior managers from the Mexican manufacturing industry participated in the research. Structural Equation Modeling was used to determine the relationships between the proposed variables. The results of model one show that social support helps mitigate the effects of burnout syndrome in its three dimensions. On the other hand, model two features better fit indices and specifically shows that co-worker support has a direct negative relationship with the dimension of emotional exhaustion, while supervisor support directly and negatively affects cynicism. These results show that, although social support helps in the mitigation of burnout syndrome in general, each type of support separately has a direct impact on a specific dimension of the burnout syndrome. Clarifying these relationships will help in the proposal of occupational hazard management policies in the Mexican industrial sector.

The multi-modal or multi-view integration of data has generated a wide range of applicability in pattern extraction, clustering, and data interpretation. Recently, variants of the Non-negative Matrix Factorization (NMF), such as joint NMF (jNMF), have allowed the integration of data from different sources and have facilitated the incorporation of prior knowledge such as the interactions between variables from different sources. However, in both NMF and jNMF, the factorization is carried out as a linear system, which does not identify non-linear patterns present in most real-world data. Therefore, we propose a new variant of jNMF called Kernel jNMF. This new method incorporates the factorization of the original matrices into a high-dimensional space. Applying our method to synthetic data and biological cancer data, we found that the method performed better in clustering and interpretation than the jNMF methods.

This paper studies vehicle platooning with communication channels subject to random data loss. We focus on homogeneous discrete-time platoons in a predecessor-following topology with a constant time headway policy. We assume that each agent in the platoon sends its current position to the immediate follower through a lossy channel modeled as a Bernoulli process. To reduce the negative effects of data loss over the string stability and performance of the platoon, we use simple strategies that modify the measurement, error, and control signals of the feedback control loop, in each vehicle, when a dropout occurs. Such strategies are based on holding the previous value, dropping to zero, or replacing with a prediction based on a simple linear extrapolation. We performed a simulation-based comparison among a set of different strategies, and found that some strategies are favorable in terms of performance, while some others present improvements for string stabilization. These results strongly suggest that proper design of compensation schemes for the communications of interconnected multi-agent systems plays an important role in their performance and their scalability properties.

Distributed Denial of Service (DDoS) attacks represent the most common and critical attacks targeting conventional and new generation networks, such as the Internet of Things (IoT), cloud computing, and fifth-generation (5G) communication networks. In recent years, DDoS attacks have become not only massive but also sophisticated. Software-Defined Networking (SDN) technology has demonstrated effectiveness in counter-measuring complex attacks since it provides flexibility on global network monitoring and inline network configuration. Although several works have proposed to detect DDoS attacks, most of them did not use up-to-date datasets that contain the newest threats. Furthermore, only a few previous works assessed their solutions using simulated scenarios, easing the migration to production networks. This document presents the implementation of a modular and flexible SDN-based architecture to detect transport and application layer DDoS attacks using multiple Machine Learning (ML) and Deep Learning (DL) models. Exploring diverse ML/DL methods allowed us to resolve which methods perform better under different attack types and conditions. We tested the ML/DL models using two up-to-date security datasets, namely CICDoS2017 and CICDDoS2019 datasets, and they showed accuracy above 99% on classifying unseen traffic (testing set). We also deployed a simulated environment using the network emulator Mininet and the Open Network Operating System (ONOS) SDN controller. In this experimental setup, we demonstrated high detection rates, above 98% for transport DDoS attacks and up to 95% for application-layer DDoS attacks.

Nowadays, deep convolutional neural networks (CNNs) for face recognition exhibit a performance comparable to human ability in the presence of the appropriate amount of labelled training data. However, training CNNs remains as an arduous task due to the lack of training samples. To overcome this drawback, applications demand one-shot learning to improve the obtained performances over traditional machine learning approaches by learning representative information about data categories from few training samples. In this context, Siamese convolutional network (SiConvNet) provides an interesting deep architecture to tackle the data limitation. In this regard, applying the convolution operation on real world images by using the trainable correlative Gaussian kernel adds correlations to the output images, which hinder the recognition process due to the blurring effects introduced by the convolution kernel application. As a result the pixel-wise and channel-wise correlations or redundancies could appear in both single and multiple feature maps obtained by a hidden layer. In this sense, convolution-based models fail to generalize the feature representation because of both the strong correlations presence in neighboring pixels and the channel-wise high redundancies between different channels of the feature maps, which hamper the effective training. Deconvolution operation helps to overcome the shortcomings that limit the conventional SiConvNet performance, learning successfully correlation-free features representation. In this paper, a simple but efficient Siamese convolution deconvolution feature fusion network (SiCoDeF2Net) is proposed to learn the invariant and discriminative complementary features generated from both the (i) sub-convolution (SCoNet) and (ii) sub deconvolutional (SDeNet) networks using a concatenation operation which significantly improves the one-shot unconstrained facial recognition task. Extensive experiments performed on several widely used benchmarks, provide promising results, where the proposed SiCoDeF2Net model significantly outperforms the current state-of-art in terms of classification accuracy, F1, precision and recall. The code will be available on: https://github.com/purbayankar/SiCoDeF2Net.

In this work we propose a new biophysical computational model of brain regions relevant to Parkinson’s Disease (PD) based on local field potential data collected from the brain of marmoset monkeys. PD is a neurodegenerative disorder, linked to the death of dopaminergic neurons at the substantia nigra pars compacta, which affects the normal dynamics of the basal ganglia-thalamus-cortex (BG-T-C) neuronal circuit of the brain. Although there are multiple mechanisms underlying the disease, a complete description of those mechanisms and molecular pathogenesis are still missing, and there is still no cure. To address this gap, computational models that resemble neurobiological aspects found in animal models have been proposed. In our model, we performed a data-driven approach in which a set of biologically constrained parameters is optimised using differential evolution. Evolved models successfully resembled spectral signatures of local field potentials and single-neuron mean firing rates from healthy and parkinsonian marmoset brain data. This is the first computational model of PD based on simultaneous electrophysiological recordings from seven brain regions of Marmoset monkeys. Results indicate that the proposed model may facilitate the investigation of the mechanisms of PD and eventually support the development of new therapies. The DE method could also be applied to other computational neuroscience problems in which biological data is used to fit multi-scale models of brain circuits.

COVID-19 has not affected all countries equally: developing countries have been more disadvantaged by the pandemic. Regarding global development, the COVID-19 pandemic has forced a step back in the path to attaining the Sustainable Development Goals (SDGs). The SDGs most negatively affected by the pandemic are identified here: education, health, and work. Then using the SDGs as a reference, this research explores the new challenges faced by developing countries and the impact of the Internet of Things (IoT) after COVID-19’s emergence. IoT solutions carried out in developing countries during the pandemic have been identified and reviewed. Successful Internet of Things for Development (IoT4D) projects, in relation to the SDGs, are highlighted. New social and technical challenges that have emerged for the IoT4D as a consequence of the pandemic are then studied. This work concludes that the future of IoT4D in the wake of COVID-19 should focus on the use of low-cost IoT devices for the SDGs most affected by the pandemic. After an exhaustive study, the Intelligent Internet of Things (IIoT) has been determined to be a key actor in the pandemic’s wake, with a leading role in the health sector. The proposed approach includes an extensive study of the new role of the IoT4D for achieving the SDGs in our forever changed world.

Blockchain technology (BC) offers an innovation platform for decentralized and transparent transactions in the maritime port industry. This technology allows guaranteeing trust, transparency and traceability of cargo and data to be tracked. Today, in the port systems of emerging market countries, BC technology is increasingly being incorporated into information and communication processes. In parallel, the social domain has begun to be explored due to the lack of link between the port and the city it occupies, and the need to incorporate public actors in decision-making at the governance level. In this sense, the present work aims to promote BC technology in order to transform data and information into useful knowledge for effective decision making, through the use of Crowdsourcing. A Crowdsourcing Blockchain CrowdBC conceptual framework and its architecture are generated for a port system in which the cyber-technological, social and cognitive domains (CSTC) of smart ports, the knowledge generation process and Crowdsourcing technology are interrelated. Finally, opportunities are discussed for ports that are in permanent development to reduce the gaps with the smart industry. As a discussion, two possible scenarios and recommendations for future implementations that consider the social and cognitive aspects of Industry 4.0 are presented.

Waste from raw materials has a direct influence on the final price of a product. However, since waste has no added value to products, customers are unwilling to pay for it. Reducing waste along the production process and supply chain allows companies to decrease costs and remain competitive in product prices. This research is conducted in a medical device manufacturing company located in Tijuana (Mexico). The company found a negative variation in part number 9540, a metallic foil used for hot stamping. During inventory cycle counts, the company found that they were purchasing rawer material than necessary due to a 50% of waste generated along the production process. We implemented two process improvement methodologies, namely Practical Process Improvement (PPI) and the Plan-Do-Check-Act model, to eliminate 100% of the raw material waste, specifically regarding the foil’s waste. The improvement project comprised two phases: a) adjusting the parameters of the hot foil stamping machine and b) replacing the hot foil stamping machine with a pad printing machine. In our research, the PPI methodology is presented in 8 detailed stages as a simple problem-solving method, in contrast with five stages reported in other documented cases. This case study presents how a company can apply continuous improvement programs and its managerial implications, even without having a structured and defined systems for quality, such as Six Sigma or Lean Manufacturing implemented. After the two implementation phases, the improvement project led to economic savings of $\$ $ 165,000 in a year.

During the last two decades, terrestrial TV broadcasting has evolved from analog to digital technology, permitting a better spectrum efficiency, being more resistant to noise and interference as well as improving signal quality. High Definition TeleVison (HDTV) has been introduced to enhance the viewer's experience. HDTV and Ultra High Definition TeleVision bring the challenge of having a quasi-error free performance at a bit error rate as low as 10-12; i.e. less than one uncorrectable error during one hour's continuous transmission of 5 Mbps data stream. Such low error rates require robust standards, careful network planning and optimized service operation. Propagation phenomena have a direct (and possibly critical) impact on those three pillars. This paper fills the current gap of a thorough survey on propagation methods and models for broadcasting. We contribute to this field with a description and analysis of propagation phenomena concerning different aspects of the broadcast network planning, broadcast standard evaluation, and broadcast service operation. The paper provides also a technical perspective of state-of-the-art TV broadcast standards and discusses the relevancy of propagation studies with future and development and regulation challenges.

The instructional design of meaningful experimental activities completely integrated into on- line Engineering courses is still a significant challenge, by following the sustainable development goal of the United Nations for quality education, the SDG number 4. Remote and virtual laboratories can be a solution to improve this design in distance methodology. This fact has been more noticeable during the COVID-19 pandemic and its consequences with respect to education. Its transparent integration into the learning process and the inclusion of learning analytic features are essential features of these kinds of technologies. Without them, lecturers cannot track the performance and the satisfaction of their students. Our LoT@UNED (Laboratory-of-Things@UNED) platform has been implemented to include all these educational features for IoT environments, ranging from edge to cloud computing. Taking the advantages of the IoT paradigm, this platform involves acquiring practical competences within IoT and edge-cloud programming topics, among others. Furthermore, its flexibility allows us to host other areas, such as cybersecurity. Specifically, our platform incorporates a set of sustainable capabilities (scalability, availability, and security), so contributing to the SDG 4 challenge with a low-complex and efficient solution for quality distance education. To analyze the impact of our proposed solution, this work also focuses on studying the influence among several acceptance UTAUT/TAM factors during the students’ learning and teaching process. To achieve this goal, both exploratory and confirmatory analyses have been conducted. The reliability and validity of data are also analyzed. An improved structural equation model has also been studied, satisfying several statistical indicators according to the recommended thresholds in the current literature. The usefulness of our platform, easing the users’ attitude and having a high availability of resources affect positively the intention of use our IoT cloud solution and, as a consequence, the great quality of the education process.

We study the problem of semi-blind channel estimation and symbol detection in the uplink of multi-cell massive MIMO (multi-input multi-output) systems with spatially correlated time-varying channels. An algorithm based on expectation propagation (EP) is developed to iteratively approximate the joint a posteriori distribution of the unknown channel matrix and the transmitted data symbols with a distribution from an exponential family. This distribution is then used for direct estimation of the channel matrix and detection of data symbols. A modified version of the popular Kalman filtering algorithm referred to as KF-M is also proposed which emerges from our EP derivations. Performance of the Kalman smoothing algorithm followed by KF-M, referred here as KS-M, is also examined. Simulation results demonstrate that channel estimation error and the symbol error rate (SER) of the semi-blind KF-M, KS-M, and EP-based algorithms improve with the increase in the number of base station antennas and the length of the data symbols in the transmitted frame. In particular, by increasing the number of transmitted data symbols in the frame, the proposed semi-blind algorithms can mitigate the effects of pilot contamination as well as time-varying channels in a multi-cell massive MIMO system with pilot-overhead of around 5%.