Resource
2026 EN
Mudathir Ahmed Mohamud · Hamidah Ibrahim · Fatimah Sidi
+4 more
Skyline queries based on pareto dominance have gained significant attention for their ability to effectively identify interesting objects from large multi-dimensional datasets. They are particularly useful in applications that involve multi-criteria decision support. Recently, several techniques have been proposed for processing continuous skyline queries over uncertain data streams, however, they mainly focus on uncertainty that results from objects having many instances. Uncertainty owing to objects having range values wherein the exact values of the objects are not known at the point of processing has not been extensively explored. Moreover, identical objects may recur in the stream at different times while continuous skyline queries submitted by different users can overlap within the same time frame. Thus, it would be inefficient to repeatedly process the same segment of the data stream and the same objects in processing these multiple queries. We seek to overcome the aforementioned problems by introducing an efficient skyline model named MCSQ-UDS, for computing multiple continuous skyline queries over uncertain data stream. Several extensive experiments were conducted on synthetic and real datasets with different parameter settings. The results show that the proposed model outperforms the baseline techniques with regard to number of pairwise comparisons and execution time.
Resource
2026 EN
Hossam Ibrahim · Mahmoud Elnaggar · Abdelmomen Mahgoub
Ensuring the stability of cascaded systems with constant power loads (CPLs) remains a critical challenge due to their negative incremental input admittance at low frequencies, which induces instability—particularly at higher power levels. In tightly regulated SPMSM drives, the inverter–machine behaves as a CPL at the dc input; hence the source–dc-link–drive constitute a cascaded system. This paper presents a novel adaptive parallel virtual impedance control strategy to enhance the dc-link stability of Surface Permanent Magnet Synchronous Motors (SPMSMs) under tight speed regulation, where dc-link stability denotes a well-damped dc bus—i.e., absence of sustained low-frequency oscillations caused by source–load impedance interaction—maintained as V bus and load power change. The method achieves this by reshaping the input impedance across varying bus voltage and power. By employing the Middlebrook feedback theorem, the input impedance and sensitivity functions are derived to identify instability sources. A parallel virtual impedance with power and voltage adaptability is then formulated, alongside a compensation controller transfer function to realize the virtual impedance. To address implementation complexity, a simplified version of the controller is proposed, eliminating the reliance on circuit parameters or additional sensors while maintaining dynamic performance. A 160W SPMSM prototype system, validated experimentally with real-time deployment on a TI C2000 F28379D microcontroller, demonstrates the feasibility and robustness of the proposed approach across varying power and voltage levels.
Resource
2026 EN
Yehia Ibrahim Alzoubi · Ahmet E. Topcu · Ersin Elbasi
The increasing urgency of climate change mitigation has accelerated the adoption of tools to measure and monitor greenhouse gas emissions at individual, organizational, and infrastructural levels. While these tools enhance the accuracy and accountability of emissions measurement, they also raise privacy concerns due to the sensitive data they collect, the granularity of monitoring, and the potential for secondary use. Despite their growing adoption, no prior study has systematically evaluated the privacy implications of carbon footprint tools across categories or examined how data practices differ between consumer calculators, enterprise systems, and emerging Internet of Things (IoT), Artificial Intelligence (AI), and blockchain solutions. This study addresses this gap by conducting a structured, cross-category privacy evaluation of carbon footprint tools, reviewing 46 tools from academic and industry sources. A novel five-dimensional privacy risk framework, encompassing data type, granularity, storage, processing, secondary use, and user agency, is developed and applied across five tool categories: individual calculators, corporate platforms, infrastructure tools, IoT/AI/blockchain systems, and research tools. The findings reveal pronounced privacy–accuracy trade-offs: tools offering higher measurement precision often require more intrusive data collection. The study also identifies category-specific least- and highest-risk tools, showing that simple household calculators present low risk. In contrast, enterprise platforms and IoT/blockchain systems exhibit the highest exposure due to sensitive operational, financial, and real-time data flows. The study is limited by its reliance on literature and public documentation, suggesting the need for empirical field studies and technical audits. Future research should expand on these findings by investigating user perceptions, legal frameworks, and privacy-preserving design strategies for sustainable digital tools.
Resource
2026 EN
Fatma Ciftci · L. S. Diab · Yunus Akdogan
+2 more
This study investigates the estimation of the stress-strength reliability measure, ς = P ( X < Y ), assuming that both stress ( X ) and strength ( Y ) follow the new X-Lindley (NXL) distribution. The NXL distribution has gained attention in reliability and risk analysis due to its flexibility and long-tailed behavior. To mimic realistic life-testing scenarios, progressive Type-II censoring is employed. Three estimation approaches are compared: maximum likelihood estimation (MLE), Bayesian estimation via Markov Chain Monte Carlo (MCMC) with a gamma prior, and the Tierney–Kadane (TK) approximation using a secondorder Laplace method. A comprehensive Monte Carlo simulation with 10,000 replications is conducted to evaluate estimator performance in terms of mean squared error, coverage probabilities, and the average lengths of asymptotic confidence or credibility intervals at the 95% and 97.5% levels. The results highlight the efficiency and robustness of the MCMC-based Bayesian approach, particularly under higher censoring levels.
Resource
2026 EN
Chaymae Yahyati · Ismail Lamaakal · Yassine Maleh
+2 more
We present FastKAN-Tiny, a fully–int8 driver-distraction detector designed for privacy-preserving, real-time operation on microcontrollers. The backbone replaces costly spatial convolutions with per-channel spline nonlinearities implemented as compact lookup tables, followed by efficient 1×1 channel mixing; stacked at stride 1 and terminated by global average pooling and a small projection ( d ∈{64, 96}), it delivers strong expressivity at tiny width while respecting tight flash/SRAM budgets. Personalization is achieved on-device via a non-parametric prototype head that requires only K ∈ {1, 5} labeled frames per class from an unseen driver, adding O(Cd) dot products and storing a tiny d×C state. The model is trained with quantization-aware training and deployed with TensorFlow Lite Micro/CMSIS-NN; a streaming loop applies causal exponential moving average smoothing and a dual-threshold early-exit gate for deterministic accuracy–latency–energy trade-offs. On AUC Distracted Driver V2, the approach achieves high accuracy in the zero-shot setting and climbs to the high–90% range with just a handful of labeled supports (e.g., ∼98% accuracy at 5-shot), while sustaining real-time throughput on embedded targets (e.g., ∼27 Hz on Cortex-M7) within a compact footprint (∼118 kB flash, ≤160 kB RAM). Altogether, FastKAN-Tiny unifies an expressive tiny-width backbone, on-device few-shot personalization, and a measured MCU deployment pipeline for practical, private driver monitoring.
Resource
2026 EN
Rami Ibrahim
Neural networks have made significant achievements in areas like computer vision and natural language processing. However, there is an increasing demand to understand their outputs and decisions. They are considered black box models because it is hard to understand their complex architecture as they lack justification and transparency. In this paper, we present a novel approach by generating stylized explanations using neural style transfer. Unlike methods that use one input image to generate activation maps, we adopt the neural style transfer technique to generate artistic styles of the input image. We select eight styles from the STaDA experiments, “la_muse”, “rain_princess”, “sunflower”, “the_scream”, “the_shipwreck”, “udnie”, “wave”, and “your_name”. We pass the stylized images with the input image to Score-CAM and produce class activation maps for the nine images, the input image and the eight stylized images. We evaluate the stylized activation maps by conducting experiments like faithfulness, object localization, and sanity check. The input image activation map outperformed stylized activation maps in terms of faithfulness with a lower confidence drop. However, in terms of object localization, “sunflower”, “the_scream”, “the_shipwreck”, “udnie”, “wave”, and “your_name” stylized activation maps outperformed the input image activation map with a higher IoU value. In terms of sanity check, both the input image activation map and stylized activation maps were sensitive to VGG-16 model randomization.
Resource
2026 EN
Osama F. Hassan · Ahmed F. Ibrahim · Ahmed Gomaa
+4 more
Driver drowsiness is a leading cause of traffic accidents worldwide, highlighting the urgent need for accurate and real-time detection systems to enhance road safety. Recent advances in deep learning have shown great promise, particularly with attention-augmented convolutional neural networks (CNNs). However, most existing studies focus on a single attention mechanism in isolation, providing limited insight into their comparative effectiveness. In this work, we present a systematic and comprehensive framework for integrating, analyzing, and benchmarking multiple attention mechanisms within a custom CNN architecture tailored for driver drowsiness detection. Specifically, we investigate four representative modules: Squeeze-and-Excitation (SE) channel attention, Spatial Attention, Efficient Channel Attention (ECA), and the hybrid Convolutional Block Attention Module (CBAM). A baseline CNN without attention is implemented to enable direct assessment of the added value of each mechanism. Furthermore, we benchmark our framework against six widely adopted transfer learning models, including VGG19, DenseNet169, ResNet50V2, InceptionV3, MobileNet, and InceptionResNetV2. Extensive experiments conducted on the NTHU-DDD2 dataset under real-world driving conditions demonstrate that attention mechanisms significantly enhance classification performance. Among them, CBAM achieved the best trade-off between accuracy and efficiency, reaching 99.63% accuracy, an AUC of 0.9999, and competitive inference latency suitable for real-time deployment. ECA also performed strongly, validating the role of lightweight attention in improving recall and training stability. By providing the first comprehensive comparative study of diverse attention mechanisms in this domain, this work establishes a robust benchmark and demonstrates that hybrid attention-augmented CNNs, particularly CBAM, are highly effective for driver drowsiness detection. These findings advance the development of intelligent and safety-critical transportation systems, while outlining pathways for future research toward robust, efficient, and deployable solutions.
Resource
2026 EN
Ezzidin Hassan Aboadla · Kushsairy A. Kadir · Ali Hassan
+1 more
The increasing demand for energy efficiency and renewable energy integration has created a strong need for accurate and reliable power measurement systems. This paper presents the design and implementation of a high-precision isolated AC power measurement system developed to enhance safety, accuracy, and reliability in electrical monitoring applications. The proposed design employs an isolation amplifier and an isolated DC–DC converter to achieve complete galvanic separation between the high- and low-voltage domains, thereby minimizing noise and electrical interference. Both simulation and experimental evaluations were conducted, demonstrating voltage measurement error of approximately 0.25% and current measurement error of 0.5%. Comparative analysis with conventional designs confirms the system’s improved precision and operational safety. Owing to its compact and cost-effective architecture, the developed system is suitable for use in residential, commercial, and industrial environments, particularly in smart grids, automation systems, and renewable energy monitoring applications.
Resource
2026 EN
Dilan Onat Alakut · Ibrahim Turkoglu
Non-Line-of-Sight (NLOS) object detection is an important research problem for inferring position, shape, and material properties in situations where direct optical access is not possible. This study proposes a multi-modal learning architecture with attention mechanisms to classify material types using only reflected acoustic waves. The proposed system consists of an attention-supported U-Net-based Reflection Isolation Network, which isolates reflection regions, and a multi-modal fusion classifier that combines spectral-temporal features. Within the scope of the study, the ANLOS-R (Acoustic Non-Line-of-Sight with Reflection) dataset was created, consisting of 1,440 echo samples collected using three different microphone–speaker configurations. This dataset was used for attention-based segmentation of echo intensity maps and extraction of acoustic features; the obtained spectral and temporal representations were evaluated using both deep learning and machine learning-based classifiers. Experimental results show that the single feature-based GRU (Gated Recurrent Unit) model achieved 65% accuracy, while the proposed fusion model reached 74% accuracy. Furthermore, the system demonstrated high generalization capability in both single and multiple material combinations. The proposed method improves NLOS detection in acoustic environments with low signal-to-noise ratios and provides a robust foundation for future acoustic imaging and autonomous detection systems.
Resource
2026 EN
Mucahit Altintas · Ismail Duru · Ibrahim Yazici
+2 more
The vision of 6G requires autonomous, intelligent, and adaptive systems that can manage complex and dynamic communication environments. In this context, artificial intelligence (AI)-based agents are seen as key enablers, combining perception, knowledge, decision, and execution functions to operate across distributed network layers. This study provides a comprehensive overview of agentic systems for 6G, examining their architecture, core capabilities, and enabling AI techniques. We propose a novel framework for agentic intelligence levels which conceptualizes agentic AI systems as a gradual and multidimensional spectrum. Additionally, trust and governance levels, which characterize system deployability independently of agentic competencies, are introduced. We analyze how methods such as reinforcement learning (RL), federated learning, graph neural networks, transfer learning (TL), and meta-learning can be integrated into agents to support tasks including network optimization, predictive traffic management, spectrum allocation, energy-efficient resource control, and security enhancement. The potential of foundation model-based agents in 6G networks is examined, comparing them to classical RL-based agents to highlight their complementarity. Practical 6G use cases—including smart resource orchestration, environmental coordination, industrial applications, reliability and security assurance, and situational awareness—are presented to demonstrate the applicability of agents in real-world scenarios. Their strengths, limitations, and complementary roles within agent-based frameworks are highlighted. The current technological and industrial maturity levels of use cases are mapped to the proposed levels of agentic intelligence and trust/governance. The study bridges theoretical advances and practical 6G implementations by linking AI methods with concrete deployment challenges. Finally, the study identifies challenges related to scalability, interoperability, explainability, and trust, and highlights future directions for building reliable, adaptive, and ethically aligned agent-based systems in 6G.