International Journal of Electronics and Telecommunications

Magnetic Resonance Imaging (MRI) is modern diagnostic tool to obtain internal images of human body. Vibration and noise are major challenges inherited by the MRI since their development. The paper analyses vibration and sound signals produced from the Low-field MRI system for different scanning sequences. Results from spectral analysis revealed strong relation between the vibration and sound signal, showing scanning sequence specific peaks at multiple frequencies. The results are useful to understand the dynamic nature of vibration and sound signal from MRI to develop the attenuation strategy.

The underwater communication technique based on MBFSK (Multiple Binary Frequency Shift Keying) modulation is used in challenging propagation conditions. Despite the bandwidth limitations present underwater, it provides stable underwater connectivity. This paper presents a hardware implementation of the MBFSK modulation technique for underwater wireless communications. The implementation was carried out using an FPGA-based programmable logic device. The concept of the device design, the hardware solution, and the firmware description are presented. An overview and diagram of the underwater communication system are also provided. The transceiver system was tested in the tank of the Hydroacoustics Laboratory of the Gdynia Maritime University, and the results are presented and discussed. The aim of this work is to present the design, software algorithm, and test results of the developed underwater communication system.

In this paper, we propose a unified approach to acoustic and electromagnetic field theories which employs control engineering methods for their analysis and modelling. Both theories can be derived from the wave equation using factorisation and subsequently represented as a system with a feedback loop in control engineering. This allows for the formulation of properties and solutions useful for further analysis. Moreover, it provides a justification and explanation of similarities between acoustics and electromagnetism. Hopefully, our unified approach to acoustic and electromagnetic field theories carries implications for the foundational understanding of both theories as well as their practical applications.

This paper describes and tests a method for automatically determining the envelope of the blood flow velocity curve in ultrasonic Doppler imaging, which was implemented in a prototype of the 128-channel mobile ultrasound B-mode scanner. On the basis of the determined envelopes, algorithms were also developed for the automatic determination of the most important characteristic points of the Doppler blood flow spectrum in pulse wave Doppler imaging mode and the most relevant blood flow parameters. Sufficiently good repeatability and precision were obtained with low computational complexity.

Binaural technology has been known for decades. However, advancements in software and consumer electronics have facilitated its widespread adoption, primarily in the postmillennium era. As binaural sound becomes more popular, the demand for spatial analysis tools is expected to grow. This paper evaluates three methods for assessing ensemble width in binaural music recordings: (1) an auditory model with decision trees, (2) a neural network model, and (3) a spatial spectrogram approach. Under ideal, anechoic conditions, the auditory model performed best with a mean absolute error (MAE) of 6.59° (±0.11°), followed by the neural network (8.57° ±0.19°) and the technique based on spatial spectrograms (13.54° ±0.92°). Extending previous work, this study analyzes the methods’ robustness to reverberation and noise. Noise resilience tests indicate moderate resistance, with the auditory model yielding an MAE of 12.34° at a 10 dB signal-to-noise ratio. However, reverberation tests show a significant drop in accuracy even at an RT60 reverberation time of 0.1 seconds. The findings may contribute to the improvement of models for estimating ensemble width in binaural recordings of music, which could influence the development of binaural sound analysis tools, with potential applications in audio production.

This document describes numerical analyses performed on a SAW gas sensor in a non-steady state. Our work involved predicting SAW velocity changes in relation to the surface electrical conductivity of the sensing layer. We found that the conductivity of the rough sensing layer (above a piezoelectric waveguide or quartz) is determined by the diffused gas molecule concentration profile inside it. Specifically, we present numerical results for the DMMP gas concentration profile (CAS Number 756-79-6) within an (RR)-P3HT layer during the non-steady state recovery step. The core of these investigations was to understand thin film interaction with target gases in a SAW sensor configuration, using the diffusion equation for polymers. The outcomes of these numerical analyses provide valuable insights for selecting sensor design conditions, including the sensor layer's morphology, thickness, operating temperature, and type. The numerical results, generated using Python code, are then elaborated upon and examined.

We present a method for motion artifacts reduction during 3D volume reconstruction of free hand 2D ultrasound sequences. Motion estimation and additional filtering improves quality of free hand 3D ultrasound data resulting in improved imaging. Reconstructed 3D data was visualized using the OpenGL ES framework and GLSL shader language allowing real-time rendering on an embedded class GPU device.

The paper presents a new switchable test set dedicated to dual delay line surface acoustic wave (SAW) gas sensors. It operates at frequency of 205 MHz utilizing dual delay line fabricated by SAW Components. The sensor detects traces of dimethyl methylphosphonate (DMMP) that is a simulant of combat poisoning agents. The test set described is built of a single oscillator adopting commutation of two SAW delay lines and this way has better thermal properties than commonly used networks consisting of two independent oscillators. Its measured frequency to temperature is ten times lower than obserwed for double oscillator networks. Additonaly due to alteranate operation of the sensors the interaction between delay lines built on common substrate is eliminated. This work was presented at the 53th Winter School on Wave and Quantum Acoustics [1] February, 24 – 27, 2025, DOI 10.5162/20WWonAE/A7.

This paper presents an algorithm for immersive processing of a multichannel recording for headphone listening. The material listened to with headphones should evoke impressions in the listener that are identical to those experienced when listening from a multi-speaker system. In order to allow the processing system to be adapted to the individual anatomical characteristics of the listener, an algorithm was developed, which is based on data in the form of directional room impulse responses acquired with an intensity probe, in the form of classical room pressure impulse responses at the excitation emitted by the individual loudspeakers of the listening system. The listening room characteristics recorded in this way are supplemented with data from the Head Related Transfer Function (HRTF) databases, which can be selected according to the listener's perception. The study compared the effects of an impulse response segmentation algorithm using publicly available HRTF averaging databases with the classic approach using individualized binaural room impulse responses (BRIR). Reference was also made to available binauralization algorithms using dummy head.

In this article spatial audio recording techniques are compared: scene-based audio and object-based audio. The study involved mixing recordings from a higher-order ambisonic microphone and support microphones, ambisonically encoded on a virtual sphere. The recordings were combined in different spatial resolution variations by manipulating the ambisonic order. A MUSHRA-like test was conducted, taking into consideration the room divergence effect. The experiment used binaural rendering with headtracking. The results were analyzed using linear mixed models, providing insights into spatial audio recording techniques.

The speaker cable used to connect a power amplifier to a speaker is a topic of much debate between researchers and a group of music enthusiasts known as audiophiles. To analyse this issue, both objective and subjective measurements were conducted to study the impact of changing the type of speaker cable on the listener's auditory experience. The results of the objective research include the frequency characteristics of the tested cables, THD+N values, and measurements of the frequency response of the speaker system connected to the amplifier using different speaker cables. For the subjective tests, 15-second audio samples from three different music genres were used, along with a comparative evaluation method. The tested perceptual attributes were dynamics, clarity of sound, transparency, and overall rating. Gathered results made it possible to conclude that the influence of the loudspeaker cable in a music system is negligible. Subjective tests carried out on an expert group confirmed the theory that the loudspeaker cable used in a home music system has no effect on music reception in any of the genres tested.

A novel, high surface encoding capacity compact planar multiresonator tailored for Chipless RFID tag applications is discussed in this article. The tag consists of three hexagonal open loop resonators that are etched on the ground plane of a 50Ω microstrip transmission line. It operates within the frequency range of 2.12 GHz to 5.45 GHz, with a bandwidth of 3.33 GHz. Frequency Shift Coding is employed to record the tag's identification in the spectral domain. A maximum of 343 distinct code words can be generated utilizing three resonators. A notable feature of this tag is its capability to achieve distinct resonating frequencies by adjusting the overall dimensions of the slot. The tag prototype is designed and fabricated on an RT5880 lossy substrate, characterized by loss tangent of 0.0009 and dielectric constant of 2.2. Experimental data from actual prototypes are presented to verify the dependability of the suggested design.

In this paper, the long-known bit representation of integer partitions was used in a novel way to develop an algorithm for generating the next partition of an integer n. This algorithm is both loopless and conditionless (and therefore has strictly constant execution time). These features lead to an efficient and highly scalable parallel implementation on a modern multi-core CPU processor or a modern FPGA chip. In the CPU case, just 30 processor clock cycles are required to generate the next partition when n < 128. In the latter case, a single one of the parallel instances uses only 1,595 look-up tables and 962 registers for n < 128. It can produce the next partition in just 6 clock cycles. Even cost-effective FPGAs can hold a few dozen such instances.

The convergence of Quantum Artificial Intelligence (QAI) with known educational technologies like Augmented Reality (AR), Gamification, and Adaptive Storytelling (AS) offers revolutionary learning experiences. While AR, Gamification, and AS have been demonstrably shown to enhance learner engagement and comprehension, especially in higher education in the STEM, language, and health domains, integrating them into a comprehensive, adaptive system that QAI can optimize remains a significant challenge. Currently, QAI applications in education are largely conceptual. This systematic literature review, following PRISMA guidelines, analyzed 42 relevant studies from Scopus (2015–2025) selected from an initial 976 studies. The findings confirm the pedagogical benefits of integrating AR, Gamification, and AS. In contrast, QAI implementation is hampered by technical constraints, pedagogical integration difficulties, and a lack of validated evaluation methodologies. Recognizing the potential of QAI-optimized adaptive learning necessitates closing the gap between existing practices and future aspirations via focused research, comprehensive framework development, expanded educator abilities, and strong multidisciplinary cooperation.

This article presents an architecture for engineering reusable embedded software using modern C++ principles and a custom–built dependency injection framework. It details the framework’s design, specifically tailored for resource-constrained environments. The framework promotes modular and testable architecture. Its data–driven (via Json file) configuration defines component dependencies and determines their instantiation. The article demonstrates how such approach facilitates component decoupling and provides a viable path for developers to create scalable, portable, and high-quality embedded software, significantly reducing future development efforts.

This paper focuses on the problem of modeling and analyzing the actions of human operators who perform demanding tasks such as real-time control of a complex dynamic plant. To this end, an approach that bases on the idea of interactive fuzzy information systems is proposed. In particular, we discuss the problem of selecting and generating perception and action attributes to describe the aircraft control tasks realized by a skilled pilot. A method consisting of several stages is proposed for modeling pilot-airplane interactions. The initial information system with sensory attributes is replaced with more complex attributes at subsequent stages. To determine the decision rules of the pilot, we apply flow graphs that are suitable for representing fuzzy interactive information systems and for evaluating properties and quality of the human operator’s decision model.

In this paper, we analyze the performance of common machine learning (ML) algorithms executed in Google Cloud and Amazon Web Services environments. The primary metric is training and prediction time as a function of the number of virtual machine cores. For comparison, benchmarks also include a "bare metal" (i.e. - non-cloud) environment, with results adjusted using the "Multi-thread Score" to account for architectural differences among the tested platforms. Our focus is on CPU-intensive algorithms. The test suite includes Support Vector Machines, Decision Trees, K-Nearest Neighbors, Linear Models, and Ensemble Methods. The evaluated classifiers, sourced from the scikit-learn and ThunderSVM libraries, include: Extra Trees, Support Vector Machines, KNearest Neighbors, Random Forest, Gradient Boosting Classifier, and Stochastic Gradient Descent. GPU-accelerated deep learning models, such as large language models, are excluded due to the difficulty of establishing a common baseline across platforms. The dataset used is the widely known "Higgs dataset," which describes kinematic properties measured by particle detectors in the search for the Higgs boson. Benchmark results are best described as varied—there is no clear trend, as training and prediction times scale differently depending on both the cloud platform and the algorithm type. This paper provides practical insights and guidance for deploying and optimizing CPU-based ML workloads in cloud environments.

Constrained optimization is central to large-scale machine learning, particularly in parallel and distributed environments. This paper presents a comprehensive study of augmented Lagrangian–based algorithms for such problems, including classical Lagrangian relaxation, the method of multipliers, the Alternating Direction Method of Multipliers (ADMM), Bertsekas’ algorithm, Tatjewski’s method, and the Separable Augmented Lagrangian Algorithm (SALA). We develop a unified theoretical framework, analyze convergence properties and decomposition strategies, and evaluate these methods on two representative classes of tasks: regularized linear systems and K-means clustering. Numerical experiments on synthetic and real-world datasets show that Bertsekas’ method consistently achieves the best balance of convergence speed and solution quality, while ADMM offers practical scalability under decomposition but struggles in high-dimensional or ill-conditioned settings. Tatjewski’s method benefits significantly from partitioning, whereas the classical Augmented Lagrangian approach proves computationally inefficient for large-scale problems. These findings clarify the trade-offs among augmented Lagrangian algorithms, highlighting Bertsekas’ method as the most effective for distributed optimization and providing guidance for algorithm selection in large-scale machine learning applications.

Research shows that mobile support robots are becoming increasingly valuable in various situations, such as monitoring daily activities, providing medical services, and supporting elderly people. For interpreting human conduct and intention, these robots largely depend on human activity recognition (HAR). However, previous awareness of human appearance (human recognition) and recognition of humans for monitoring (human surveillance) are necessary to enable HAR to work with assistance robots. Al-so However, multimodal human behavior recognition is constrained by costly hardware and a rigorous setting, making it challenging to effectively balance inference accuracy and system expense. Naturally, a key problem in human pose or behavior detection is the ability to extract additional purposeful interpretations from easily accessible live videos. In this paper, we employ human pose detection to address the problem and provide well-crafted assessment measures to show demonstrate the effectiveness of our approach, which utilizes deep neural networks (DNNs) This article proposes a human intention detection system that anticipates human intentions in human- and robot-centered scenarios by utilizing the incorporation of visual information as well as input features, including human positions, head orientations, and critical skeletal key points. Our goal is to aid human-robot interactions by helping mobile robots through realtime human pose prediction using the recognition of 18 distinct key points in the body's structure. The effectiveness of this strategy is demonstrated by the suggested study using Python, and the results of simulations verify the reliability and accuracy of this method.

This paper examines Hybrid Multi-Agent Systems, integrating both human and non-human intelligent agents, as a new subject of management research. It presents original definitions of key concepts: intelligent agents, artificial intelligent agents, and Hybrid Multi-Agent Systems. These definitions are grounded in Distributed Artificial Intelligence and provide a foundation for exploring the collaboration between human and artificial intelligent agents. The study addresses fundamental research questions regarding the nature of intelligent agents and their role within Multi-Agent Systems, proposing Hybrid Multi-Agent Systems as a novel framework that allows for seamless cooperation between human and non-human entities. Through a narrative literature review, this paper highlights the potential implications of Hybrid Multi-Agent Systems for scientific research in management, offering a conceptual basis for future research in this evolving field.

In today's technology-driven era, innovative methods for predicting behaviors and patterns are crucial. Virtual Learning Environments (VLEs) represent a rich domain for exploration due to their abundant data and potential for enhancing learning experiences. Long Short-Term Memory (LSTM) models, while proficient with sequential data, face challenges such as overfitting and gradient issues. This study investigates the optimization of LSTM parameters and hyperparameters for VLE prediction. Adaptive gradient-based algorithms, including ADAM, NADAM, ADADELTA, ADAGRAD, and ADAMAX, exhibited superior performance. The LSTM model with ADADELTA achieved 91% accuracy for BBB course data, while ADAGRAD LSTM models attained average accuracies of 80% and 85% for DDD and FFF courses, respectively. Genetic algorithms for hyperparameter optimization significantly contributed, with the GA + LSTM + ADAGRAD model achieving 88% and 87% accuracy in the 7th and 9th models for BBB course data. The GA + LSTM + ADADELTA model produced average accuracy rates of 80% and 84% in DDD and FFF course data, with the highest accuracy rates of 86% and 93%, as well. These findings highlight the effectiveness of adaptive and genetic algorithms in enhancing LSTM model performance for VLE prediction, offering valuable insights for educational technology advancement.

This paper presents the design of the temperature sensor and analysis of different calibration methods. The main aspect of the analysis was optimization of the number and location of the measurement points needed to perform the calibration for a given accuracy. For this purpose, temperature sensors using proportional to the absolute temperature (PTAT) current and current controlled oscillator (CCO) have been designed in 180 nm technology. To reduce the sensitivity of the sensor to the supply voltage, the low dropout regulator (LDO) has been used. The bandgap circuit generates the stable reference voltage for the LDO and PTAT current for the CCO.

In this paper, the important characteristics of a Zigzag Phosphorene Nanoribbon Tunneling FET (ZPNR-TFET) are studied by inserting a single vacancy (SV) defect. After adjusting the positions of the defect in the length of the channel, it is found that the SV defect decreases on current in all three defect positions, and the biggest reduction is when the SV defect is in the center position. The off current decreases when the SV defect is located in the center of the channel and increases when the defect is located near the source and drain. The largest increase in off current is related to the location of the defect close to the source. The on-off current ratio decreases in all three defect positions. The greatest impact is related to the condition where the defect is located on the source side. Semi-empirical Slater–Koster approach using DFTB-CP2K parameters were used for the density functional based tight binding (DFTB) calculations of ZPNR.

The growing importance of creating appropriate safeguards in electronic systems forces design of integrated circuits dedicated for cryptographic purposes. The paper focuses on True Random Number Generator (TRNG) circuits design allowing generation of random bit stream. Presented TRNG architecture uses low frequency high-noise oscillator for sampling high frequency clock signal. The article also describes a method for obtaining a high noise level in the oscillator. Achieved bit rate of designed TRNG equals 1 Mb/s. The circuit dissipates 144 μW. The design of the TRNG, simulation and measurement results of the manufactured IC chips have been described in the paper also. TRNG circuit has been implemented in 180 nm CMOS technology.

Falls are a common problem in many environments and affect people of all ages. Although some people fall to be minor incidents, they can have serious consequences, especially for vulnerable groups like the elderly and stroke survivors. This study aimed to develop a system for detecting falls in patients using sensor fusion and machine learning methods to accurately identify the positions of the falls. The system combines data from accelerometers and gyroscopes using the Kalman filter to categorize falls into four types: supine, prone, left, and right. The system uses the k-Nearest Neighbors (k-NN) algorithm for threshold fall motion detection to reduce false detections. A fall detection triggers the system to send the position data via LoRaWAN communication, making the data accessible through Node-RED and Telegram. The system performance was evaluated through several tests: MPU6050 sensor measurement to calibrate and respond to the Euler accelerometer and gyroscope sensor, kalman filter measurement, threshold fall detection with the k-NN algorithm measurement, and performance LoRaWAN communication. The results showed that calibrating the MPU6050 sensor effectively minimized sensor drift and noise. The implementation of the kalman filter successfully reduced noise in the sensor readings, the k-NN algorithm provided optimal system values and performance, and data transmission via LoRaWAN to Node Red and Telegram was effective.

This paper presents a novel mathematical framework utilizing an M/M/1 queuing algorithm with non-preemptive priority, modeled using SimEvents in MATLAB. The proposed framework evaluates the main Quality of Services (QoS) metric, specifically average delays across high, medium, and low priority queues. Comparison between the simulated M/M/1 queuing model and its theoretical calculations demonstrated close alignment, with the simulated results closely approximating the theoretical values. This validation confirms the effectiveness of the simulation model in representing the theoretical framework. Additionally, the framework complies with the IEEE 802.15.6 standard by maintaining average delays below the 125 ms threshold across all priority levels.

The X-ray imaging systems dedicated for X-ray spectroscopy, based on a semiconductor strip sensors have been recently an important research topic. The most important research objective is working towards improvement of the spectroscopic and position resolution features [1]–[3]. In spectroscopic applications the short strip silicon detectors are widely used due to their relatively small capacitance and leakage current. Using strip pitch below 75 μm enables achievement of high spatial resolution. In this work, the analysis and design of the read-out electronics for the short silicon strip detectors are presented. The Charge Sensitive Amplifier (CSA) is optimized for the detector capacitance of about 1.5 pF, and the shaping amplifier default peaking time is about 1 μs (controlled by the sets of switches). To achieve the lowest possible noise level, the sources of noise in a radiation imaging system both internal (related to the frontend electronics itself), as well as external, were considered [4]. We target the noise level below 40 el. rms, considering low power consumption (a few mW) and limited channel area. To increase the speed of incoming hits processing, the continuoustime resistive CSA feedback together with a digital feedback reset are included. The prototype integrated circuit comprises of 8 charge processing channels, biasing circuits, reset and base-line restoration logic, and a calibration circuit.

The article presents the results of video quality assessment using two methods recommended by the International Telecommunication Union. These are the Single Stimulus (SS) and the Double Stimulus Impairment Scale (DSIS) methods. The results obtained by both methods were compared. The studies were performed for two coding techniques, H.264 and H.265, and for spatial resolutions of 1280 × 720 and 1920 × 1080. The studies showed a very high correlation between the MOS values obtained by SS and DSIS methods for H.265 coding and both video resolutions. Regarding the H.264 coding, a very good consistency of the results was obtained for bitrates starting from 3000 kbps.

This paper addresses the problem of selecting a cloud infrastructure configuration for a geo-distributed enterprise. It extends the well-known virtual machine (VM) placement problem to consider multiple datacenters so they can serve a distribution of end-users in their geographic locations in an optimal way in terms of low end-user latency, and acceptable costs. We approach this problem by formulating a multicriteria mixed integer linear program (MILP) that integrates an aspiration/reservation-based modeling of the client’s preferences. A newly proposed model supports the selection of virtual instances across cloud regions, ensuring flexible trade-offs among QoS objectives: total infrastructure cost, user distance, and edgeto- central latency. Case study results based on Google datacenters in Europe demonstrate the flexibility of our method in providing Pareto-optimal solutions aligned with varied preferences. The approach contributes to the growing preference-aware cloud resource allocation field and offers a scalable solution to the service composition problem in heterogeneous cloud environments.

Microelectromechanical systems (MEMS)-based inertial navigation units are widely utilized due to their low cost, small form factor, and low power consumption. However, they face critical limitations in high-speed rotating systems due to gyroscopic drift, saturation, and sensitivity to environmental conditions. This paper proposes a novel method for supporting inertial navigation by estimating angular velocity using ambient electromagnetic radiation detection, offering a drift-resilient and interference-immune solution.

Quantum information technologies QIT have two separate but entangled layers – material and ideological. Material layer embraces algorithms, physical devices and systems. Cognitive layer is related to the boundaries and structure of our knowledge, including the magic of nonlocal interactions and subtleties of contextuality. The paradox is that QIT should use quantum magic as a resource to turn into functional utility. The International Year of Quantum Science and Technology is celebrated in Europe, Poland and elsewhere. Opening conference at the UNESCO headquarters in Paris was attended by representatives of the Polish Ministry of Science and Higher Education, and some quantum science groups. Several scientific conferences organized in Poland were announced to the IYQ25 calendar, including: May Symposia of Information and Quantum Horizons in Gdańsk, August Max Born Optical Symposium in Wrocław, September Congress of the Polish Physicists in Katowice. Quantum research projects are realized in Poland, QuantERA, tasks included in the European Quantum Flagship EQF, construction of quantum equipment by university consortia MIKOK, quantum networks and others.

The paper presents and discusses the requirements related to testing cameras used in active safety systems. It also includes the results of temperature tests of the MTF (Modulation Transfer Function) parameter of a camera developed by Aptiv, conducted across a wide range of ambient temperature variations in the company's laboratory.

The paper describes the design process of an efficient model predictive control (MPC) algorithm based on fuzzy models. An interesting feature of the proposed approach is that it uses easy–to–obtain fuzzy Takagi–Sugeno (TS) models composed of a few step responses employed as local models; one of these models is used to derive the dynamic matrix, and the second one, being a skillful modification of the first one, to generate the free response. The designed MPC algorithm uses formulation as an efficient quadratic optimization task. Still, it offers control quality compared with the MPC algorithm formulated as a nonlinear optimization task, thanks to the skillful generation of the free response. The efficiency of the proposed approach is tested and demonstrated in the simulated control system of the nonlinear and non–minimum phase process of the chemical reactor with the van de Vusse reaction.

We address the multicommodity flow problem with a nonlinear goal function modeling queueing delay. It is well-known that linear programming solvers perform better than those used for nonlinear programming. We can leverage their performance by employing the Generalized Benders Decomposition (GBD) to partition the problem into master and primal subproblems. We prove that in the case of multiple subproblems, which is true in our case, we can split both the optimality and feasibility cuts and add them independently. Moreover, we extended a known proof of convergence to enable a wider range of problems to be solved using GBD. We use the split cuts technique to precompute feasibility cuts and analytically solve the subproblems to omit the use of nonlinear optimization software. Furthermore, we explore the possibilities of starting point selection through linear and quadratic approximation. We carry out tests on a classical network example to show that GBD can sometimes outperform nonlinear solvers, and also that quadratic approximation for starting point selection can provide strictly better solution times, dominating commercial solvers.

The aim of this research is to enhance the effectiveness of Android malware detection systems by implementing dimensionality reduction techniques on Boolean data. Algorithms such as Linear Discriminant Analysis (LDA), Principal Component Analysis (PCA), and Multi-Correspondence Analysis (MCA) serve as operations preceding the classification stage. The analysis is carried out using multiple classifiers such as Random Forest Classifier, Logistic Regression, and Support Vector Machines to measure how effective they can detect cyber threats. Results show that the Decision Tree Classifier, implemented without dimensionality reduction, achieved the optimal results with 100% accuracy. Efficient feature selection and rapid computation in the context of malware detection are necessary for real-time mobile cyber environment applications.

This paper presents a statistical analysis of the enhanced SDEx (Secure Data Exchange) encryption method, using a version that incorporates two session keys. This method has not previously been combined with the BLAKE3 hash function. The statistical analysis was conducted using the NIST Statistical Test Suite. Several real-world sample files were encrypted using the proposed method and then subjected to statistical analysis through selected tests from the NIST suite. These tests aimed to determine whether the resulting ciphertexts meet the criteria for pseudorandomness. Additionally, compression tests were performed using WinRAR, which confirmed that the ciphertexts are not compressible.

High-Level Synthesis (HLS) has become an established methodology to accelerate the development of FPGAbased systems by allowing algorithms to be written in high-level languages (HLLs) such as C/C++ or Python. Yet, for real-time physics experiments—including fusion plasma diagnostics, highenergy physics (HEP) detectors, and rare-event astrophysical triggers—conventional HLS still falls short in three essential aspects: determinism, portability, and auditability. Pragmas embedded in HLL code blur the separation between algorithmic intent and implementation details, coupling scientific software to a particular device or compiler version. This is particularly problematic in long-lived scientific projects such as ITER or the Pierre Auger Observatory, where systems must remain functional and maintainable over decades [4]–[6]. To address these challenges, we propose an Intermediate Representation (IR)-centric HLS flow—PyHLS—that explicitly introduces an abstraction layer between algorithm and Register- Transfer Level (RTL) design. The IR centralizes all performancecritical aspects: timing contracts (initiation interval, latency, jitter), concurrency (loop unrolling, pipelining), memory layout (banking, tiling, port allocation), and resource binding (DSPs, BRAMs, AI tiles). In this model, the algorithm is expressed in clean, testable Python code [1], [2], while device-specific optimizations are described in a structured IR graph. This IR is then lowered into a reusable VHDL microinstruction library [3], which serves as a portable middle layer across devices. By versioning and auditing IR graphs and instruction streams, PyHLS ensures reproducibility and traceability—critical properties in scientific computing where results must be verifiable years after deployment. The methodology builds upon earlier work in Python-based high-level synthesis, parameterizable metamodels, and algorithmic synthesis with multi-level compilers [8], [9], [11]. It incorporates systematic design space exploration (DSE), allowing parameter sweeps over IR attributes and early feasibility checks. The flow is complemented by a cycle-accurate microinstruction emulator, which validates both functionality and timing contracts before vendor toolchains are invoked, reducing iteration time and catching infeasible designs early. We demonstrate the motivation and applicability of this approach in two demanding domains. First, in plasma diagnostics at JET/ITER, where spectrometer and data acquisition systems must combine high bandwidth with deterministic latency [6]. Second, in trigger pipelines for astroparticle physics, where artificial neural networks (ANNs) and fuzzy-logic algorithms have been implemented directly in FPGA logic to discriminate rare events from large backgrounds [4], [5]. These use-cases highlight the need for explicit IR-level contracts and modularity: the same high-level algorithm must be portable across device generations, yet adapted to exploit specialized hardware resources such as DSP slices, systolic AI engines, or high-bandwidth memories. The contribution of this work is therefore threefold: 1) We formalize the role of an explicit IR in HLS, decoupling algorithms from implementation decisions and introducing contract-driven determinism. 2) We present a reusable VHDL microinstruction library and emulator that stabilize implementation and provide auditable artifacts. 3) We show how PyHLS extends naturally to heterogeneous FPGAs, mapping operators to emerging AI/ML blocks while maintaining scientific reproducibility and portability across decades. By unifying algorithmic specification, IR-based parameterization, and reusable microinstructions, PyHLS establishes a sustainable methodology for real-time physics experiments and beyond. In short: write the science once, retarget the hardware many times.High-Level Synthesis (HLS) has become an established methodology to accelerate the development of FPGAbased systems by allowing algorithms to be written in high-level languages (HLLs) such as C/C++ or Python. Yet, for real-time physics experiments—including fusion plasma diagnostics, highenergy physics (HEP) detectors, and rare-event astrophysical triggers—conventional HLS still falls short in three essential aspects: determinism, portability, and auditability. Pragmas embedded in HLL code blur the separation between algorithmic intent and implementation details, coupling scientific software to a particular device or compiler version. This is particularly problematic in long-lived scientific projects such as ITER or the Pierre Auger Observatory, where systems must remain functional and maintainable over decades [4]–[6]. To address these challenges, we propose an Intermediate Representation (IR)-centric HLS flow—PyHLS—that explicitly introduces an abstraction layer between algorithm and Register- Transfer Level (RTL) design. The IR centralizes all performancecritical aspects: timing contracts (initiation interval, latency, jitter), concurrency (loop unrolling, pipelining), memory layout (banking, tiling, port allocation), and resource binding (DSPs, BRAMs, AI tiles). In this model, the algorithm is expressed in clean, testable Python code [1], [2], while device-specific optimizations are described in a structured IR graph. This IR is then lowered into a reusable VHDL microinstruction library [3], which serves as a portable middle layer across devices. By versioning and auditing IR graphs and instruction streams, PyHLS ensures reproducibility and traceability—critical properties in scientific computing where results must be verifiable years after deployment. The methodology builds upon earlier work in Python-based high-level synthesis, parameterizable metamodels, and algorith-incorporates systematic design space exploration (DSE), allowing parameter sweeps over IR attributes and early feasibility checks. The flow is complemented by a cycle-accurate microinstruction emulator, which validates both functionality and timing contracts before vendor toolchains are invoked, reducing iteration time and catching infeasible designs early. We demonstrate the motivation and applicability of this approach in two demanding domains. First, in plasma diagnostics at JET/ITER, where spectrometer and data acquisition systems must combine high bandwidth with deterministic latency [6]. Second, in trigger pipelines for astroparticle physics, where artificial neural networks (ANNs) and fuzzy-logic algorithms have been implemented directly in FPGA logic to discriminate rare events from large backgrounds [4], [5]. These use-cases highlight the need for explicit IR-level contracts and modularity: the same high-level algorithm must be portable across device generations, yet adapted to exploit specialized hardware resources such as DSP slices, systolic AI engines, or high-bandwidth memories. The contribution of this work is therefore threefold: 1) We formalize the role of an explicit IR in HLS, decoupling algorithms from implementation decisions and introducing contract-driven determinism. 2) We present a reusable VHDL microinstruction library and emulator that stabilize implementation and provide auditable artifacts. 3) We show how PyHLS extends naturally to heterogeneous FPGAs, mapping operators to emerging AI/ML blocks while maintaining scientific reproducibility and portability across decades. By unifying algorithmic specification, IR-based parameterization, and reusable microinstructions, PyHLS establishes a sustainable methodology for real-time physics experiments and beyond. In short: write the science once, retarget the hardware many times.H

In the forthcoming 6G wireless communication networks, it is crucial to address the challenges of massive information capacity and ultra-high data transmission rate. Terahertz (THz) band communication emerges as a promising candidate to meet these demands. However, the natural physical characteristics of THz signals, including significant attenuation in transmission and limited diffraction capabilities, pose substantial limitations on their practical applications. In recent years, reconfigurable intelligent surfaces (RIS) have garnered significant interest due to their exceptional potential in manipulating electromagnetic (EM) waves. RIS is anticipated to mitigate the challenges associated with THz communication. In this research, we introduce a RIS design featuring 1-bit phase modulation operating within THz bands. The RIS units leverage PIN diodes to dynamically adjust the unit structure, enabling dynamic encoding of the RIS metasurface based on the on and off states of diodes. In this paper, three typical application scenarios, including beam scanning, beam convolution, and radar cross section (RCS) reduction, are listed. Coding methodologies including fractional coding, convolution, Golay-Rudin-Shapiro (GRS) coding, and Genetic Algorithm (GA) coding, are deployed, and simulation results illustrate the effectiveness. This study lays the groundwork for the practical deployment and coding strategies of RIS in the THz bands, thereby facilitating the integration of THz technology into the next generation of 6G communication systems.