The miniaturization of field-effect transistors has reached a critical juncture at the 1.5 nm technology node, where conventional drift-diffusion models fail due to dominant quantum effects. We present a device engineering framework grounded in geometric collapse theory, wherein environmental coupling is harnessed as an active switching mechanism rather than treated as a parasitic effect. By exploiting the inherent structure of quantum state space–the Riemann sphere for two-level systems–we demonstrate that asymmetric environmental coupling can force deterministic wavefunction collapse to a desired pointer state within switching timescales. Our approach provides explicit design rules: (1) environmental coupling rates κ_S and κ_D must differ by at least 10^13 s ^-1 , (2) effective mass engineering using III-V materials achieves this condition, and (3) symmetry breaking through gate geometry or material composition yields collapse times τ_collapse < 0.1 ps. This framework reconciles quantum measurement theory with semiconductor device physics, offering a pathway to extend Moore’s Law beyond the 1.5 nm barrier. We analyze the validity of the geometric flow equation, competition with thermal effects at 300 K, the transition from deterministic to probabilistic regimes, and fundamental limits below 1 nm. A comprehensive NEGF simulation with realistic parameters ( α = 0.5 ) confirms the theoretical predictions, demonstrating a 3.86 × 10^5 improvement in I_on/I_off ratio compared to conventional tunneling transport. The comprehensive NEGF simulation results, including key parameters and performance metrics, are summarized in Tables 6-12.
This paper presents a cloud-native blueprint for high-throughput, compliant banking workflows by modeling each transaction, KYC check, and payment orchestration as a verifiable trace graph across microservices running in hybrid and multi-cloud environments. It introduces a method to aggregate structurally similar transaction paths to surface latency hotspots, third-party gateway bottlenecks, and anomaly routes, enabling rapid root-cause triage aligned with SLAs, auditability, and regulatory retention. The approach renders side-by-side diffs of service dependencies and critical paths before and after release trains or policy updates, highlighting regressions that affect batch cutoffs, settlement windows, and fraud-risk scoring. Designed for cloud adoption patterns common in banks—Kubernetes, managed databases, and regional failover—the system remains vendor-agnostic and integrates with observability stacks to support incident response, capacity planning, and change management in core banking, payments, and digital channels.
In this work, the modelling and microcontroller implementation of Nonlinear Resistor-Capacitor-Inductor Josephson Junction (RCLSJJ) with Hysteretic Iron-Core is investigated. The nonlinear inductor is modeled using a tangent interference term. In the first part of this study, we analyze the Josephson junction model with a nonlinear inductor. The dynamical behaviors of the system are investigated using classical tools such as one-parameter bifurcation diagrams, Fast Fourier Transform (FFT) spectrums, and phase portraits. Numerical simulations performed in MATLAB reveal that the JJ-based nonlinear inductor exhibits a variety of behaviors, including excitable modes, regular spiking, periodic bursting, chaotic attractors, and periodic attractors. To support both engineering applications and educational purposes, a microcontroller-based implementation of the Piecewise Resistor-Capacitor-Inductor-Shunted Josephson Junction (PRCL-SJJ) circuit is developed. The results from the digital implementation closely match those obtained from numerical simulations.
Managing transportation systems in modern cities is increasingly complex owing to growing urbanization, rising vehicle numbers, and limited road infrastructure. High traffic density during morning peak hours makes intelligent control systems based on real-time data processing essential. Traditional cloud-based architectures suffer from network latency, bandwidth saturation, and delayed decision-making when large data volumes are transmitted to remote servers. This article proposes a mathematical model based on the edge computing paradigm to minimize network latency in cyber-physical transportation systems. In the model, primary data processing is performed at edge nodes located near road intersections, thereby reducing the data volume transmitted to the cloud and increasing real-time decision-making speed. The model was experimentally evaluated on real Los Angeles traffic data (METR-LA, 207 sensors). Results show that edge–cloud integration reduces the mean end-to-end latency from 132.4 ms to 39.9 ms (approximately 69.8
Water quality monitoring plays a vital role in environmental protection, public health, agriculture, aquaculture and industrial automation. Conventional laboratory-based water analysis methods are often time-consuming, expensive and unsuitable for continuous real-time monitoring. This paper proposes an intelligent Internet of Things (IoT)-based multi-parameter water quality assessment and classification system integrated with fuzzy logic for real-time applications. The proposed system utilizes multiple sensors, including pH, turbidity, Total Dissolved Solids (TDS) and temperature sensors, to continuously monitor water category. An ESP32 microcontroller is employed for real-time data acquisition, processing, wireless communication and intelligent decision-making. To improve classification reliability under uncertain and dynamic environmental conditions, a fuzzy logic-based inference system is incorporated into the proposed framework. The fuzzy controller evaluates the sensor parameters simultaneously using adaptive membership functions and rule-based reasoning to classify water into different usability categories such as drinkable, washable, farming, aquarium, and non-usable. The classified results are displayed locally through an OLED display and remotely monitored through IoT cloud connectivity. Experimental validation was conducted using multiple real-world water samples collected from domestic, agricultural, pond, industrial and contaminated water sources. The obtained results demonstrate that the proposed fuzzy logic-based system provides improved classification accuracy, better adaptability, enhanced noise tolerance and reliable real-time monitoring performance compared to conventional threshold-based approaches. The developed system is portable, low-cost, energy-efficient and suitable for deployment in remote and industrial environments. The proposed intelligent framework offers a promising solution for next-generation smart water quality monitoring and environmental management systems.
Abstract Buck/Boost converters require robust control to handle nonlinearities and uncertainties. However, conventional sliding mode control (CSMC) and backstepping control (BSC) methods suffer from chattering, tedious tuning, and unresolved tradeoffs between rapid response and steady-state stability. To address these issues, we propose a backstepping double integral sliding mode control (BDISMC) strategy optimized via Particle Swarm Optimization (PSO). A hyperbolic tangent function replaces the signum term to suppress chattering and enhance smoothness within a Lyapunov-stable framework. When benchmarked against SMC, BSC, integral sliding mode control (ISMC), and double integral sliding mode control (DISMC), our approach significantly improves settling and rising times, boosting responsiveness and robustness. Minor overshoot and undershoot occur during transients. Overall, this research provides valuable insights into converter control and clearly illustrates the intrinsic performance trade-offs among speed, stability, and precision.
The emerging era of quantum computing has highlighted the importance of secure communication systems that combine quantum-resistant cryptography, quantum communication, and advanced security analytics. This systematic review critically examines hybrid models of quantum and classical artificial intelligence, focusing on architectures for quantum key distribution (QKD), intrusion detection, network management, and the integration of post-quantum cryptography. Following PRISMA 2020 guidelines, studies from January 2020 to July 2026 were sourced from IEEE Xplore, ACM Digital Library, ScienceDirect, SpringerLink, Wiley Online Library, and backward citation searches. Out of these, 30 primary studies met the inclusion criteria and were evaluated using an eight-item quality rubric and a five-level evidence-maturity framework. The findings are grouped into three categories: AI-enhanced QKD and secure communication, hybrid quantum–classical learning for intrusion detection, and post-quantum or hybrid cryptographic solutions. AI-supported QKD research has shown notable reductions in parameter search time while maintaining near-optimal secret-key rates, primarily in simulation settings. Hybrid quantum machine learning models showed competitive intrusion detection accuracy and F1 scores, but improvements over classical methods were modest, dataset-dependent, and often lacked comprehensive reporting on false positives, statistical significance, or computational costs. Post-quantum cryptography approaches demonstrated greater maturity, with evaluations involving Transport Layer Security (TLS), Internet Protocol Security (IPsec), embedded devices, wireless links, and hardware accelerators, although performance varied with platform resources, cryptographic object sizes, network conditions, and side-channel protections. Among the studies, half were simulation-based, 30
Household water safety depends not only on contaminant removal efficiency but also on consistent user adherence and timely maintenance. However, usability barriers and lack of accessible guidance limit the real-world effectiveness of point-of-use purification systems, particularly in multilingual settings. This study presents Automated Quality Understanding and Voice Assistant i.e. AQUA-VA, a smart domestic water purifier integrating dual-stage RO + UV/UF filtration with edge artificial intelligence and multilingual voice-assisted interaction. The system employs low-power sensing of pH, total dissolved solids (TDS), turbidity, flow, and temperature, coupled with a machine-learning-based water quality index model and Bayesian filter health estimation for predictive maintenance. A mixed-integer programming scheduler optimizes energy usage and purification cycles, while an on-device speech interface enables real-time multilingual guidance. The system was evaluated over 12 weeks across 36 households under diverse water conditions. Results show a 71.3
Abstract The imminent threat posed by quantum computing to classical cryptographic systems necessitates the development of quantum-resistant, efficient, and scalable encryption techniques, especially for real-time distributed optimization networks critical to national infrastructure. This study introduces a novel lightweight post-quantum cryptographic algorithm tailored for secure real-time decision-making in decentralized systems such as smart grids, autonomous transport, and defense communication networks. This study proposes a lattice-based encryption scheme optimized for low-latency and bandwidth-constrained environments, integrating a parameterized Learning With Errors (LWE) framework with a compressed key encapsulation mechanism (KEM). The cryptographic algorithm is coupled with an adaptive distributed optimization protocol that dynamically adjusts computation and communication loads across agents to maintain performance under cryptographic overhead. The method is rigorously analyzed in terms of computational complexity, security assumptions, and operational feasibility. Simulation experiments are conducted over dynamic networks modeled on real-world distributed control systems, comparing performance metrics such as encryption latency, decision throughput, and fault tolerance against state-of-the-art schemes. Results demonstrate significant improvements in end-to-end delay, with cryptographic integrity maintained under adversarial conditions, establishing the scheme’s applicability for post-quantum real-time systems. The contributions of this work bridge the critical research gap between post-quantum cryptography and real-time optimization, reinforcing secure decision-making in systems of national interest.
Abstract The presence of partial shading in photovoltaic (PV) generates unequal irradiance across modules, leading to a multi-peak power–voltage profile with several local operating points, significantly degrading the performance of conventional gradient-based maximum power point tracking (MPPT) methods. This paper proposes a Partial Shading Conditions PSC -aware artificial neural network ANN-based global MPPT strategy that reformulates global maximum power point GMPP tracking as a direct voltage estimation problem. The method employs a P–V curve probing mechanism based on five normalized power samples measured at predefined voltage ratios, enabling shading pattern identification without irradiance or temperature sensors. A feed-forward MLP trained offline using 3000 operating cases—including uniform irradiance, static PSC, severe multi-peak PSC, and dynamic shading scenarios—predicts the optimal operating voltage in real time. Simulation results demonstrate improved tracking accuracy, faster convergence, and reduced steady-state oscillations compared with P&O, Incremental Conductance, and PSO algorithms. Under severe PSC, the proposed approach achieves 93.93% MPPT efficiency, while providing up to 12% higher energy extraction under dynamic shading conditions. The proposed framework shows strong suitability for real-time embedded PV control applications.
Abstract In last few years, integrating renewable energy sources (RES) with traditional energy sources became prominent applications to diversify the energy mix in the modern power grids. Nevertheless, the incorporation of RES poses additional challenges to power distribution networks because of their inherent intermittency and their randomness nature. Furthermore, maintaining grid operational efficiency during peak demand periods requires effective peak load shaving, which remains a critical challenge. A promising remedy of these challenges is to use energy storage technologies such as battery systems, green hydrogen generation, thermal storage, etc. Thus, the battery stored energy can be dispatched during periods of low renewable generation and high load demand, thereby enhancing grid reliability and efficiency. Consequently, energy storage solutions emerge as a compelling alternative to traditional, expensive grid challenges, due to their flexibility, declining costs, and rapid deployment. This paper proposes a novel hybrid algorithm for optimizing the charging and discharging schedule of a PV-battery storage system connected to microgrid. The proposed Hybrid optimization technique combines two metaheuristic algorithms, Particle swarm optimization (PSO) and gray wolf optimization (GWO) for utilizing the available PV power optimally targeting the reduction of the peak power demand. Historical data of solar PV generation unit and power demand are collected from a substation in UK. The proposed approach introduces a new optimization strategy that maximizes the use of photovoltaic energy to charge the energy storage unit which contributes to maximizing the daily peak load reduction. The simulation results conducted for two seasonal scenario days with different PV generation profiles achieve power load reductions between 21.6% and 28.1%. The results also prove that the proposed algorithm outperforms the other well-known optimization techniques introduced in the literature in terms of optimization techniques performance metrics with ranges from 1 to 3%.
Abstract The article examines a coherent set of principles for constructing observability systems in cloud-based applications, employing the OpenTelemetry standard as a primary instrument for achieving transparency, predictability, and controllability across distributed computing environments. The relevance of the study is conditioned by the rapid proliferation of cloud-native architectures and the need for unified mechanisms to correlate metrics, traces, and logs within multiservice, elastically scalable systems. The objective is to identify and systematize architectural and methodological principles that enable the design of observable applications grounded in the OTLP (OpenTelemetry Protocol) unified protocol and OpenTelemetry’s native integration into the Microsoft ecosystem. The novelty lies in a holistic treatment of observability not as an isolated technical module but as an embedded engineering discipline spanning the entire application life cycle, from project templates and CI/CD pipelines to cloud operations. The article proposes viewing OTLP as a telemetry USB port for distributed systems, enabling signal portability across monitoring platforms (Grafana, Azure, Dynatrace) without code changes or violating architectural invariants. Key results include substantiating OpenTelemetry’s role as a lingua franca between applications and analytics platforms; distinguishing three principal telemetry-collection topologies (sidecar, gateway, and managed); and analyzing their trade-offs between contextual proximity and operational overhead. The article is intended for DevOps engineers, cloud solution architects, developers, and researchers in telemetry and cloud-native technologies seeking to build a predictable, interpretable, and scalable observability system.
Abstract Increasing heterogeneity, densification, mobility, and service diversification have rendered traditional static resource allocation strategies ineffective for contemporary wireless networks. The transition to 5G-Advanced and emerging 6G systems further introduces highly dynamic environments characterised by continuously evolving traffic demand, interference patterns, network topology, and application requirements. This study presents a PRISMA-based systematic review of adaptive resource allocation in dynamic wireless environments, synthesising evidence from 34 peer-reviewed studies published from 2020 onward across major databases including IEEE Xplore, Scopus, Web of Science, ACM Digital Library, ScienceDirect, and SpringerLink. Unlike prior surveys, this review provides a unified cross-domain perspective by integrating modelling approaches, allocation strategies, and architectural contexts across terrestrial, edge-enabled, non-terrestrial, and semantic communication systems. It further develops a structured taxonomy of adaptive resource allocation frameworks, encompassing optimisation-based, learning-based, and hybrid methods under non-stationary conditions. The findings reveal a strong shift toward explicit dynamic modelling and the dominance of learning-based approaches, particularly deep reinforcement learning and multi-agent reinforcement learning, due to their ability to handle uncertainty and sequential decision-making. However, persistent challenges related to training stability, generalisation, scalability, and deployment safety are driving the emergence of hybrid and constraint-aware frameworks. The review also highlights the increasing integration of adaptive allocation within advanced architectures such as Open RAN, multi-access edge computing, non-terrestrial networks, and digital-twin-assisted systems. Despite these advances, evaluation practices remain largely simulation-driven, limiting reproducibility and real-world applicability. This study uniquely identifies critical methodological gaps and outlines deployment-oriented research directions, including cross-domain optimisation, trustworthy AI integration, and reproducible validation frameworks, to support the development of reliable and scalable adaptive resource allocation mechanisms for future 6G systems.
The incorporation of Distributed Energy Resources (DER), especially PV and BESS, is essential in making radial power systems stable. Unfortunately, most recent optimization algorithms make an assumption of an unconstrained utility supply, targeting minimization of losses during daylight or only profit-making energy trading. Such traditional methods will not work in underdeveloped areas that face significant capacity limitations in their main substations, forcing them to undertake mandatory load shedding. In such situations, having an unconstrained BESS leads to parasitic voltage drop at weak tail nodes, whereas reactive disconnect switches lead to excessive Energy Not Served (ENS) production. In order to fill this gap, an artificial intelligence-based approach of survivability control that changes the focus from economic dispatching to active grid survivability has been proposed in this study. By using Particle Swarm Optimization (PSO), the proposed approach incorporates modelling of artificial solar intermittency, constrained and asymmetric dispatching of BESS systems, and a demand-side management procedure. A penalization technique is used to ensure strict adherence to the statutory voltage constraints. Using the IEEE 33 bus network system with the maximum capacity for active power at 4.0 MW, the AI-based approach performed significantly better compared to the conventional distributed generation strategies. The new proposed approach was able to ensure that there were no violations of the absolute minimum network voltage (0.95 p.u.), reduced power losses to 1.696 MWh (a 31.6
Solar energy has become an important source of renewable energy towards supporting the increased electricity demand in the world and minimizing reliance on fossil energy. Nevertheless, solar irradiance is intermittent and unreliable, which necessitates the precise prediction of solar energy to promote effective integration and energy management into the grid. This review bridges a gap in the research to unify the different machine learning (ML) and deep learning (DL) methods to predict solar power and solar resource, with the need to have a comparative evaluation of the performance, interpretability, and adaptability of the methods. This paper presents a systematic review of the recent developments in ML and DL based models applied in the analysis of solar power and solar resources forecasting, such as standalone, hybrid, and ensemble models. The research topic is to determine the appropriate parameters of input, feature selection techniques, and forecasting horizons that improve the accuracy and precision of the model. The methodology that was adopted is a comprehensive comparative analysis of the previous research with an emphasis on the strengths, weaknesses, and possibilities of the various predictive models. The findings prove that hybrid and ensemble models are invariably more successful compared to traditional models, in terms of accuracy and reliability. The review shows that the synergy of AI-based solutions can determine significant improvements in terms of accuracy of forecasting, energy scheduling, and grid stability. Taking all of this into consideration, the provided review fits into the smart prediction model evolution for effective and environmentally friendly use of solar energy. The key contributions of this paper and the main highlights are as follows:
Abstract In the medical domain, most of the medical image datasets available today are struggling from the imbalance data problem, making it difficult to detect anomalies or rare healthcare events. Most of the rare events are found in the minority class, not in the majority class. Machine learning algorithms assume that the underlying dataset is evenly distributed. However, finding such rare medical events are difficult because of the imbalance nature of the data. Class imbalance is a major concern whether it belongs to medical or any other field. In any classification problem, the provided dataset should be equally distributed whether it is a majority class or minority class. However, in the real world we cannot observe such equal distribution in any datasets. In this review paper, we’ll discuss the primary issues involved in learning from the imbalanced dataset, as well as issues such as imbalance class problem in medical domain, by examining medical image records. In addition to that we will provide the most recent methods like GAN, OCC, DCNN, Federated Learning, Transfer Learning and Attention Mechanism for handling imbalanced datasets. Apart from this we will investigate the recent development and trends that we can use to tackle this imbalance learning problem.
In the realm of space technology, robust ground systems are crucial for effective satellite communication. This study focuses on the design and analysis of dynamic controllers for satellite antenna tracking systems utilizing DC servo motors. Through the evaluation of three controllers LQR (Linear Quadratic Regulator), PID (Proportional-Integral-Derivative), and Fuzzy-PI. The research investigates their performance in optimizing antenna positioning within Ground Data Receiving Stations (GDRS). Utilizing mathematical modeling and simulation techniques, the study reveals compelling insights. Results indicate that the Fuzzy-PI controller surpasses PID and LQR counterparts, exhibiting superior responsiveness and robustness. Specifically, the PID controller demonstrated a very big slew rate and a very big overshoot Where, the LQR controller exhibited a big slew rate, and a minimal overshoot. Remarkably, the Fuzzy-PI controller showcased the lowest slew rate and a mere overshoot. These findings underscore the importance of meticulous control strategy selection to enhance system stability and efficiency in ground data reception operations. Moreover, the study provides practical implications for improving satellite data receipt and communication reliability, offering significant contributions to the field of space technology.
Abstract Accurate electricity demand forecasting is essential for reliable power system operation, energy planning, and infrastructure development. In Saudi Arabia, electricity consumption is strongly affected by high temperatures, seasonal activities, religious events, and rapid industrial growth under Vision 2030. Traditional forecasting methods often struggle to capture the nonlinear and complex behaviour of regional electricity demand. This paper presents a comparative evaluation of four advanced ensemble and gradient boosting models, including CatBoost, XGBoost, LightGBM, and Random Forest, for electricity demand forecasting in the Makkah and Madinah regions of Saudi Arabia. A real long-term historical dataset spanning nearly two decades was used with climate-related and calendar-based features to model load variability. Model performance was evaluated using MAE, RMSE, MAPE, and $$R^2$$ . The results show that CatBoost achieved the best overall performance with an RMSE of 2029.85 GWh (out of a mean monthly consumption of approximately 37,800 GWh), MAE of 1458.10 GWh, MAPE of 3.86%, and an $$R^2$$ of 0.985. The findings confirm the effectiveness of CatBoost for electricity demand forecasting in climate-sensitive and event-driven regions.
Abstract With the development of generative artificial intelligence (AI) and the active implementation of large language models (LLMs) in the ubiquitous field, a very important task arises, which requires an objective evaluation of the quality of such AI systems. Traditional machine learning metrics turn out to be inapplicable, since solution responses of LLM-based solutions demonstrate high variability in wording while maintaining semantic correctness. This paper analyzes existing approaches to evaluate the quality of systems built on the basis of generative AI, such as lexical methods (term frequency–inverse document frequency, TF-IDF, and Best Matching 25, BM25), semantic embeddings, hybrid approaches based on LLM-as-a-Judge, and natural language inference (NLI) methods. Particular attention is paid to the development of an algorithm for selecting the optimal evaluation strategy depending on various tasks, including the latency of evaluation, the correctness and interpretability of the results, as well as the stability and reproducibility of the obtained evaluation results. For comparison, the work presents the results of various evaluation methods using the example of analyzing the accuracy and relevance of a response from an AI system on a set of 500 test examples, demonstrating a correlation with expert assessments in the range from 0.67 to 0.92, depending on the chosen approach. The proposed algorithm can be used to build a suitable evaluation process for AI systems in various domains.
Abstract Power factor deterioration in electrical systems with dynamic and nonlinear loads remains a significant challenge, leading to increased power losses, voltage instability, and reduced system efficiency. Conventional automatic power factor correction systems, typically based on fixed or binary capacitor switching, often exhibit slow response times and limited adaptability under rapidly varying load conditions. This study presents the development of an IoT-based monitoring and automatic power factor correction system designed to improve real-time compensation performance in dynamic load environments. The proposed system integrates real-time sensing, a microcontroller-based control unit, and IoT-enabled data transmission for continuous monitoring and adaptive capacitor switching, ranging from 2 to 37 µF, via relay control. A control algorithm dynamically adjusts reactive power compensation based on instantaneous load conditions. The experimental validation with resistive, inductive, and mixed loads demonstrated an improvement in power factor from 0.64 to 0.99. This results in a 17–34% boost in efficiency and a 31% reduction in current. This device exhibits a latency of under 10 s for cloud data synchronization and achieves an operational uptime exceeding 98%. To validate the capacitor selection process, the MATLAB Simulation was used to compute the capacitance necessary for the reactive power compensation. The average Pearson correlation score (> 0.97), MAE (2.2708), RMSE (3.1617) and MAPE (14.02) score in comparison to the theoretically based MATLAB simulation illustrates the system’s accuracy and reliability. A cost analysis revealed a payback period of approximately2714 operating hours, indicating the system’s economic viability. Overall, these findings present a scalable and cost-effective IoT-based automatic power factor correction, which is an innovative approach, enhancing power quality and energy efficiency in smart grid applications for residential and industrial settings.
