The need for effective and portable parallel programming models has increased due to the complexity of high-performance computing systems, especially with the introduction of many-core processors like GPUs. Creating applications for various multi-core and many-core architectures frequently necessitates maintaining separate codebases, which increases development effort. In this paper, we evaluate Kokkos, a high-level C++ framework that allows for performance portability across architectures, by running five benchmark applications on a CPU and comparing their performance with OpenMP-based counterparts. The obtained results and corresponding observations are discussed in detail.
This work evaluatesFK-meansRA against abroader set of LEACH variants, including LEACH-C, MOD-LEACH, LEACH-B, MULTIHOP-LEACH, and I-LEACH,with a focus on metrics such as average residual energy, number of alive nodes, and throughput. The paper focuses oncomparative performance benchmarking across these popular LEACH variants, providing a rigorous validation of FK-meansRA efficiency. In this paper, we extend our prior study on the performance of the proposed algorithm, FK-meansRA, against famous protocols, such as LEACH, etc. Comparative analysis with LEACH variants ensures a fair, relevant, and focused performance assessment of clusteringhierarchical protocols in contrast to chain-based or flatrouting protocols like PEGASIS or HEED. The simulationresults demonstrate that the fuzzy logic-based modeloutperforms these protocols in terms of network stability and resource utilization.FK-meansRA has nearly 450 nodes alive after the end of 1250 rounds, 200% energy more left than multihop LEACH, and a 13.6% throughput improvementover multihop LEACH.
This paper presents a survey of Internet ofThings (IoT) applications using the Raspberry PI (RPi) SingleBoard Computer (SBC) alongside Computer Vision (CV) techniques from the field of Artificial Intelligence (AI). It presents and compares solutions across several IoTapplication areas, offering an overview of the associated hardware, software, CV methods, and algorithms. The studyexplores IoT applications in the following areas: Smart Healthcare, Face and emotion recognition, Wildlife, Smart Agriculture, Smart Homes, Security, Smart Cities, Autonomous Vehicles, Robotics, Manufacturing, and Retail. Each area is analyzed with respect to the integration of RPiand CV, showcasing their contributions to enhancingoperational efficiency and enabling innovative solutions. The presented solutions use different RPi boards, from RPi 3A+ up to the latest RPi 5.
This paper presents the engineering design and usability validation of SAVIRE (Social Anxiety-Virtual Reality-Serious Game), an immersive digital therapeutic platform for adolescents with Social Anxiety Disorder (SAD).Adopting an engineering-centered approach, SAVIRE is developed using a modular system architecture that incorporates a Behavioral Scoring Engine for real-time classification of user responses during virtual social interactions. The system is implemented using Unity 3D and the Oculus SDK and optimized for the Meta Quest 3S, leveraging inside-out tracking to enable high-fidelity immersion without reliance on external computing devices. Following the Rapid Game Development (RGD) methodology, clinically grounded therapeutic principles are translated into interactive mechanics and graded virtual exposure scenarios. Technical evaluation conducted by clinical experts using the User Experience Questionnaire (UEQ) demonstrated positive results, with mean scores exceeding 2.67 across all dimensions, including efficiency and stimulation. Usability testing involving 20 adolescents diagnosed with SAD yielded an average System Usability Scale (SUS) score of 80.38, indicating excellent usability. These findings position SAVIRE as an engineering-oriented framework for adaptive virtual simulations, contributing to the development of scalable digital therapeutic systems in mental healthcare.
The main goal of the study is to determine whether artificial intelligence can create mathematical and simulation models of the automatic room heating system, and therefore test the abilities of artificial intelligence in the field of engineering sciences. The scientific novelty of the article is the development of a simulation model of an automatic room heating system with relay temperature control using artificial intelligence. To fully and comprehensively disclose the purpose of the study, elementary physical laws are provided that describe the processes of heat transfer and thermal conductivity of the air in a confined space, such as JouleLenz's law or Ohm's law and describe the foundations of the theory of automatic control, and in particular the role of necessary links and negative feedback. Based on the operation of the built models, graphs of the resulting temperature are given, on the basis of which the operating time, the maximum achieved temperatures and the switching time of the heater are based.
Different types and different size photovoltaic (PV) systems are in use. Assuming that these systems are equipped with some of the: internet of things (IoT) devices, intelligent electronic devices (IED), phasor measurement units (PMUs) and that they are properly communicated with control center, or that inverters in the PV systems contain network interface cards (NICs) which are connected to the control center through network, a solution for infrastructure of the monitoring and management (M&M) system is proposed. The main parts of the system are cloud infrastructure on which supervisory control and data acquisition (SCADA) system is implemented, network that connects all parts of cloud together and enables connections, through proprietary links or service providers networks, to the distributed devices as well as to the inverters at PV sites. Time synchronization between the central part of the SCADA system and devices at PV systems sites enables obtaining precise moments of the events occurrence and timely response. All parts of the SCADA system and network must be secured from the inside and outside threats.
In this paper presents the design and radiation characteristics of a stepped printed dipole antenna array element. The array element is designed to operate in the 5G millimeter wavelength (mm-Wave) range (from 24.25 to 27.50 GHz). In the required frequency band, the value of the VSWR does not exceed 1.30. The value of the realized gain ranges from 4.53 to 5.59 dB. The overall dimensions (width × depth × height) are 6 mm × 6 mm × 2 mm. Based on it, a model of a 4 ×4 antenna array was developed. Its radiation characteristics were compared with the characteristics of the arrays of the differently shaped dipole antennas (a horizontal flat, a horizontal flat with matching inserts, a vertical flat, and a volumetric). It is shown that the stepped shape of the dipole makes it possible to obtain the best characteristics.
Securely and reliably connecting multiple operating sites using existing Internet access lines is a common requirement for companies as well as individuals. Towalink is a new open and flexible solution for interconnecting multiple network sites reliably and securely. It is built around proven open-source projects and by itself published open-source. We present in theory and based on a practical setup the central management it provides and show how it is following the infrastructure-as-code paradigm.
When the limit miniaturization of the coupler is achieved when the elements are located in one conductive layer, then miniaturization can be continued by changing the design of the capacitor that is part of the filter: reducing the gap between the plates or increasing the dielectric constant between the capacitor plates, and due to the movable plates, you can change the operating frequency of the coupler.
The paper presents an approach to automated conceptual database design that combines speech processing and text processing techniques for the automated derivation of conceptual database models from recorded speech. In the first phase, the recorded speech is converted to the corresponding text by applying speech processing techniques. In the second phase, the text is converted to the corresponding conceptual database model by applying text processing techniques. The proposed approach is supported by an online tool named SpeeD, which is the first tool enabling automated derivation of conceptual database models from recorded speech, whereby several different natural languages are supported.
With the growing impact of climate change, the occurrence of hazardous spatial events increases. Wireless sensor networks are suitable to sense, monitor, and report such events in remote or inaccessible locations. Hazardous events are rare compared to the network's lifetime, thus maintaining its consistency must be realized energy efficiently. During the impact, the network must monitor the event with precision, and report the incidence, while mitigating the loss of perishing nodes. To fulfill these requirements, we propose the Self-healing Multipath Routing Protocol that is based on the Heterogeneous Disjoint Multipath Routing Protocol and introduces application-specific extensions to improve network stability, resiliency, and failover. To realize the monitoring of spatially extended hazardous events we introduce an event-based, application-level protocol. To evaluate the routing protocol, we perform simulations utilizing a cellular automaton-based wildfire model as the spatial event and provide measurement results including delivery ratio, consumed energy, and protocol-specific metrics.
The performance of the class of sparse reconstruction algorithms which is based on the iterative thresholding is highly dependent on a selection of the appropriate threshold value, controlling a trade-off between the algorithm execution time and the solution accuracy. This is why most of the state-of-the-art reconstruction algorithms employ some method of decreasing the threshold value as the solution converges toward the optimal one. To address this problem we propose a data-driven adaptive threshold selection method based on the fast intersection of confidence intervals (FICI) method, with which we have augmented the two-step iterative shrinkage thresholding (TwIST) algorithm. The performance of the proposed algorithm, denoted as the FICI-TwIST algorithm, has been evaluated on a problem of image reconstruction with the missing pixels, exploiting image sparsity in the discrete cosine transformation domain. The obtained results have shown competitive performance in comparison with a number of state-of-the-art sparse reconstruction algorithms, even outperforming them in some scenarios.
Interaction channels are special opportunities to improve customer satisfaction by offering a consistent problem-solving experience. Contact center employees are the link between the company and the customer. They are responsible for maintaining an appropriate relationship between the company and the customer. So, they are personally responsible for the customer experience. In this paper, we present an objective evaluation method for evaluating customer-agent interaction, i.e., evaluating the effectiveness of the realization of customer requests from calls. The evaluation method is automatic and does not depend on the relationship between the call center manager and the employees. The motivation for evaluating calls stems from the key performance characteristics of a contact center, of which we particularly emphasize service time, first call resolution, handling time, and others.
Nowadays, Learning Arabic is perceived as drab by students, with 2.4% finding it very drab, 25.3% finding it drab, and 34.9% finding it quite drab. This research is a development of the MLATS ARABIC for easy Arabic language learning. The functional testing showed a 100% success rate for all 7 test cases, and the application performed well on five different computer browsers. The user satisfaction survey yielded a 95% average satisfaction rate for each indicator.
The purpose of this study is to develop a method for two-factor authentication of electronic documents using an enhanced encrypted non-certified digital signature with the use of a security token with biometric data (fingerprint image). In order to achieve the goal, the method of comparative analysis was used. Existing algorithms for the electronic signature operation were studied. The method of multi-factor authentication of an electronic signature using biometric data has been studied in detail. Biometric data included: a handwritten password, an autograph, typing biometrics when typing a pass phrase, a facial image, typing biometrics when typing a free text. An external storage medium with a biometric authentication method based on a fingerprint image was also studied. Information on this media was accessed by scanning a fingerprint image. After conducting a comparative analysis and studying in detail these methods, a method for two-factor authentication of an electronic signature using a security token with biometric data was developed.
Usage of electrical energy obtained from renewable sources is rapidly increasing. Distributed solar systems will be widely used. The photovoltaic (PV) systems can be used to produce active and reactive power or to compensate reactive power. This paper proposes a solution for monitoring and management of distributed PV systems from the integrated center. The main parts of the system are a cloud infrastructure at the Data Center and Internet of Things (IoT) devices at distributed sites. Also, networks that are used for communication between main parts are an important piece of the proposed solution. While IoT devices continuously send information about electrical parameters of the system and environmental conditions, the inverter, as a part of the PV system, can be controlled by influencing the Maximum Power Point Tracking (MPPT) algorithm operation and inverter functionality.
The perceptual quality of image is affected by distortions during compression, delivery and storage. Distortions also impact automatic image quality assessment (IQA) that needs to be highly correlated with subjective scores. In the absence of reference, which is a typical scenario in practice, no-reference (NR) metrics are necessary for quality measurements. Recently such methods are proposed, and they employ natural scene statistics (NSS). The experimental analysis performed in this paper takes into consideration two fitting or regression models of several NR-IQA metrics relying on different distortion types. The results show quadratic model as promising for making relations in terms of difference mean opinion score and Shannon entropy.
5G networks are already being implemented around the globe. One of the most important enablers of their penetration are the Software Defined Networking (SDN) technologies and the Network Functions Virtualization (NFV) architecture, which allow the needed flexibility of the network and the composing elements. In such circumstances, the Internet-of-Things (IoT), which has been long awaited, is becoming feasible and economically reasonable. This setup has its challenges, especially due to the network expansion toward the edge, where the number of networking elements and service consumers is rapidly rising. The compute resources and the storage have to be brought in the network proximity of the access network, so that the latency of the service is kept under 1ms, which is one of the base 5G requirements. For our research, we have made an experimental setup of a distributed NFV architecture on a multiple geo-location, with a main objective to review the network latency caused by the architectural distribution of the services that are built in it. The results can be used by researchers and network architects to build reliable and costeffective distributed services with the lowest possible latency, as well as to plan possible disaster recovery scenarios when some physical location is unavailable.
There is a constant push on agriculture to produce more food and other inputs for different industries. Precision agriculture is essential to meet these demands. The intake of this modern technology is rapidly increasing among large and medium-sized farms. However, small farms still struggle with their adaptation due to the expensive initial costs. A contribution in handling this challenge, this paper presents data gathering for testing an in-house made, cost-effective, multispectral camera to detect Flavescence dorée (FD). FD is a grapevine disease that, in the last few years, has become a major concern for grapevine producers across Europe. As a quarantine disease, mandatory control procedures, such as uprooting infected plants and removing all vineyard if the infection is higher than 20%, lead to an immense economic loss. Therefore, it is critical to detect each diseased plant promptly, thus reducing the expansion of Flavescence dorée. Data from two vineyards near Riva del Garda, Trentino, Italy, was acquired in 2022 using multispectral and hyperspectral cameras. The initial finding showed that there is a possibility to detect Flavescence dorée using Linear discriminant analysis (LDA) with hyperspectral data, obtaining an accuracy of 96.6 %. This result justifies future investigation on the use of multispectral images for Flavescence dorée detection.
