Resource
2026 EN
Manal K. Fattoum · Heba El-Halabi · Sarah M. Ibrahim
+4 more
Water quality monitoring is critical for public health and environmental sustainability. This paper introduces a novel integration of radio frequency (RF) antenna sensing with machine learning to enable low-cost, real-time monitoring of water quality. The antenna, designed in High-Frequency Structure Simulator (HFSS) and fabricated on a Rogers RO4003C substrate (ε r = 3.55, thickness = 0.813 mm), is a compact wideband square loop microstrip antenna.The antenna’s copper-to-copper area is only 486.58 mm 2 (22 mm × 22.115 mm). The antenna operates across two distinct bands (3.88 – 8.77 GHz and 9.99 – 12.85 GHz), offering a total bandwidth of 7.75 GHz. It functions as a dielectric sensor, detecting salt and sugar concentrations in water by observing shifts in the reflection coefficient (S 11 ) induced by solute-dependent permittivity changes. Measurements were performed using a Vector Network Analyzer (9 kHz–13.6 GHz), and the results confirmed clear correlations between solute concentration and antenna response. To improve solute classification, a hybrid stacking ensemble machine learning pipeline is proposed that fuses extracted antenna-based features with statistical water quality data. Multiple base learners including Random Forest, Logistic Regression, KNN, LightGBM, and CatBoost were optimized and stacked using meta-learner. The ensemble achieved 92% accuracy, outperforming a baseline of a neural network. This work establishes a new paradigm in antenna-based sensing by merging RF hardware with interpretable AI, advancing the development of intelligent, real-time environmental monitoring systems.
Resource
2026 EN
Ibrahim Bagwan · Adheesh Mudgal · B. Nandini
Vishing, or voice phishing, is a major weakness in the world’s cybersecurity. The large data sets necessary for its detection are not publicly accessible, and real-time filtering is necessarily done in the cloud, which is a privacy issue for users. To overcome these issues, we developed a Domain-Adaptive Transfer Learning approach tailored for offline processing on edge devices. To cope with the lack of data, we built a two-step training process. First, we pre-trained a DistilRoBERTa model on a large dataset of 138,813 SMS messages to learn the semantic patterns of financial urgency. Second, the model was fine-tuned on our new Composite Call Transcript Dataset (N = 46, 982), which combines forensic call logs, translated fraud scripts, and artificially created scenarios. This approach proved highly effective, achieving a classification accuracy of 99.84% and an F1-Score of 0.99, thus proving that text patterns of fraud can be successfully transferred to the voice domain. To test its robustness, we conducted a SHAP value analysis, which verified that the model is concentrated on high-risk keywords such as “OTP,” “funds,” and not on random noise. Finally, the system was implemented as an offline Android app using the Vosk engine. Our system is fully offline and takes only 140 ms to complete the detection process on a standard mobile device, thus providing strong protection without sending the user’s audio or data to the cloud.
Resource
2026 EN
Nguyen Duc Tu · Ibrahim A. Elgendy · Abdukodir Khakimov
+4 more
Microservice migration on fog/edge infrastructure is a core operational mechanism to ensure low latency and high availability for time-sensitive applications. However, common single-path deployments are prone to network bottlenecks, which prolong migration time and increase downtime. This paper proposes MPMRP (Multipath Migration Routing Protocol), a cross-layer multipath migration routing protocol that directly targets downtime reduction. MPMRP performs path-level measurements and assesses computational resources at the path “bottleneck,” normalizes indicators, and scores paths using weights; it then selects the top-K paths and splits pre-copy data in proportion to bandwidth so that all paths complete simultaneously. The accompanying mathematical model establishes the convergence condition λ < 1, the stopping rule, and performance measures. Kubernetes/Minikube-based experiments show that MPMRP consistently shortens total migration time by about 61–74% compared with the single-path baselines AODV, EAODV, and the migration-aware routing scheme of Kuzmina et al., and by about 8–10% compared with the multipath transport baseline MPTCP. In terms of downtime, MPMRP achieves reductions of approximately 95–98% relative to AODV, 93–96% relative to EAODV, 94–97% relative to Kuzmina et al., and around 8–11% relative to MPTCP. These results confirm the benefits of cross-layer optimization grounded in path-level measurement and multipath routing, paving the way for seamless service migration in 6G URLLC scenarios, Metaverse/XR, and industrial digital twins.
Resource
2026 EN
Ibrahim Kircova · Munise Hayrun Saglam
Programmatic advertising has transformed digital marketing into an automated ecosystem governed by data, algorithms, and platforms. However, this transformation reproduces the core tensions of data capitalism, including the contradictions between precision and privacy, efficiency and ethics, and innovation and inequality. Grounded in the Paradox Theory framework, this study examines whether programmatic systems inevitably reinforce surveillance capitalism or can evolve toward a more participatory and accountable marketing order that enhances digital resilience. A three-round Delphi study was conducted with 34 senior practitioners representing advertisers, agencies, platforms, regulators, and civil society. The findings reveal eleven structural paradoxes that define the socio-technical dynamics of programmatic advertising. Scenario analysis evaluates nine possible futures along two axes: regulatory intensity and platform dominance. Results indicate that status-quo trajectories strengthen data monopolies, reduce algorithmic transparency, and increase exposure to misinformation and deepfakes. Transformative outcomes become feasible when four institutional conditions coincide: enforceable data commons, multi-effect accounting that includes attention and carbon, auditable algorithms, and open protocols under multi-stakeholder governance. Building on these insights, the study proposes a governance-oriented reframing of the marketing mix, termed Permissions, Participation, Portability, and Purpose (4P 2.0). This framework demonstrates how performance and ethics can reinforce one another, providing a roadmap for policymakers and industry leaders to move programmatic advertising from creative extraction to creative renewal.
Resource
2026 EN
Ahmed M. Gab Allah · Mohamed M.S. Gaballa · Dina M. Ibrahim
+1 more
The treatment of diabetic wounds presents a significant clinical obstacle. Traditional wound analysis methods are macroscopic and reliant on limited histological assessments. Biopsy sections stained with hematoxylin and eosin (H&E) and Masson’s trichrome are examined visually. This paper presents a plan for evaluating wound healing in diabetic patients. The classification model integrates a DiscreteWavelet Feature Extraction (DWTF) module for spatial-frequency feature extraction and a Spatial Feature Extraction (SFE) module; moreover, it shares information between the two modules to improve feature representation to ensure nuclear morphological features. Additionally, a segmentation method is developed to evaluate wound dimensions and contours for precise automated monitoring. To assess the efficacy of our method, we conducted histochemical and histopathological evaluations of full-thickness excisional wounds, utilizing wound biopsies fromWistar rats subjected to two distinct histochemical treatments.We utilized the identical deep neural network (DNN) architecture for both training and analysis, attaining a mean test set accuracy of 93.17%. We reassessed our evaluation of chronic wound healing by analyzing enhancements that improved our model’s performance. This resulted in a mean Dice score of 93.84% for segmentation and classification accuracy exceeding 90%. These findings underscore the promise of combining histopathological imaging with artificial intelligence.
Resource
2026 EN
Ayman Ibrahim Abouseda · Ali Hennache · Resat Doruk
+1 more
This study presents an experimental investigation of open- and closed-loop speed control applied to a 2.2 kW brushless DC (BLDC) motor integrated with a planetary gearbox and evaluated under variable-speed and fixed-load conditions. The aim is to characterize the dynamic response, energy-conversion efficiency, and battery state-of-charge (SOC) behavior of the powertrain under real-time control strategies. A complete laboratory platform was developed using a commercial BLDC motor controller, a 60 V–17.4 Ah lithium-ion battery pack, an optical speed sensor, and an eddy-current dynamometer capable of applying load torques from 1.5 to 7.5 N·m. Open-loop operation regulated speed through direct voltage modulation, whereas closed-loop control was implemented using a proportional–integral–derivative (PID) algorithm executed in real time on a microcontroller. Real-time measurements of voltage, current, torque, and speed were used to compute electrical power, mechanical power, and system efficiency, while battery SOC variation was estimated using the Coulomb-counting method. The results indicate that open-loop control exhibits substantial speed-tracking errors, pronounced transient power surges during speed transitions, reduced efficiency, and an increased SOC depletion rate over the investigated speed range of 1000–2600 rpm. In contrast, the closed-loop PID controller maintains steady-state speed errors within approximately 1 % across all load levels, suppresses current fluctuations, mitigates transient losses, and enhances overall system efficiency by approximately 14–19 percentage points. Closed-loop operation further reduces SOC depletion by 0.49–0.75 percentage points, depending on the applied load torque. These findings demonstrate the effectiveness of feedback control in improving stability, energy utilization, and electric powertrain performance in BLDC-based propulsion systems, with direct relevance to electric-vehicle and hybrid powertrain systems.
Resource
2026 EN
Mohamad Faizal Yusman Mohd Hanappi · Nor Azwan Mohamed Kamari · Mohd Asyraf Zulkifley
+1 more
This work proposes an equalization time-to-cycle conversion rate, T-rate eq , for determining the number of battery equivalent cycles. In most applications, batteries are subjected to repeated random, irregular, and imbalanced operating profiles that result in nonuniform cycle representation. Consequently, representing battery operating cycles in more accurate manner is important to estimate the battery state of health within the context of cyclic aging. T-rate eq is introduced to address this challenge and can be formulated from T-rate, a time-to-cycle conversion rate. In contrast to the existing approaches that use state of charge to determine the number of cycles, such as rain-flow cycle counting algorithm (RCCA) and area under curve (AUC), the T-rate, which is derived from the concept of battery C-rates and a general energy storage calculation, can convert operating time directly into cycles. T-rate eq then integrates battery degradation characteristics with the concept of T-rate to provide an equivalent cycle. The performance of the proposed T-rate is compared with that of existing approaches, RCCA and AUC. T-rate provides 19.4 cycles for a modified UDDS data of an ideal case study, showing a comparable cycle counting performance as AUC approach. In addition, T-rate improves the cycle counting accuracy when considering experimental data, yielding an improvement percentage of 10.0% and 3.7% from RCCA and AUC, respectively. On the other hand, T-rate eq demonstrates a normalized gradient with respect to the reference C-rate at 1C and gives a cumulative of 69.4 equivalent cycles for the idea case study. The findings clearly indicate that the proposed approach enables improved cycle counting to avoid either overestimating or underestimating the battery state of health.
Resource
2026 EN
Muhammad Rashid · Nazir Ahmad Zafar · Waheed Shahzad
+3 more
Polkadot is an emerging blockchain application in the cryptocurrency ecosystem, where a large number of participants actively take part to earn rewards. Block production in Polkadot is one of the most critical and technically complex components, which is governed by the Blind Assignment for Blockchain Extension (BABE) protocol. In BABE, validators compete in time slots to produce blocks by following the longest chain rule. To determine the eligibility of a specific validator for a given slot, several sub-operations are performed, including block lottery execution, block authoring, and block building procedures. Even a minor flaw in these processes can adversely affect the correctness of the chain and may compromise the overall blockchain operation. To ensure a reliable and error-free system, rigorous formal verification is essential to mathematically guarantee system correctness and reliability. Existing studies mainly rely on informal testing or limited assertion-based validation, while a comprehensive formal verification of BABE is still missing. To bridge this gap, this work presents a formal verification of the BABE protocol using model checking techniques. For this purpose, an integrated behavioral model of BABE is developed using the Process Meta Language (Promela), in which each core operational aspect of the protocol is modeled in detail. After modeling, key safety and liveness properties, such as fairness, availability, integrity, and consistency, are formally specified using Linear Temporal Logic (LTL). Both the Promela model and the corresponding LTL specifications are provided as input to the model checker, which verifies the properties over all possible executions of the model. Additionally, verification results include an analysis of the explored state space, verification time, and memory consumption. This study contributes toward the formal verification of other major Polkadot protocols by demonstrating the applicability of model checking, specifically using the SPIN model checker.
Resource
2026 EN
Emine Elif Yigit · Ramazan Yeniceri · Ibrahim Ozkol
This paper presents a model-based FPGA (a type of reconfigurable digital integrated circuit) accelerator for well-known classical channel flow problem CFD simulation and compares its performance with CPU and GPU implementations. A 100-core pipelined solver for Couette–Poiseuille flow is developed using a Model-Based Design approach in MATLAB/Simulink environment, where the high-level algorithm is developed and automatically translated into HDL code (the source code of the accelerator) via MATLAB HDL Coder. The design achieves a 6.34× speed-up over the GPU on a 350 × 350 grid with 177,824 iterations. The model-based workflow significantly reduces development time by enabling high-level algorithm design, automatic HDL generation, and early verification. Operating at 106 MHz, the FPGA solution offers high data processing capacity and low power consumption, making it an ideal choice for real-time, onboard CFD applications in energy-constrained aerospace systems. Experimental results demonstrate higher performance per watt and consistent timing compared to CPU and GPU alternatives. The results highlight the scalability and versatility of model-based FPGA acceleration for CFD and pave the way for future applications involving unstructured grids and more complex flow scenarios.
Resource
2026 EN
Alejandro Giraldo-Quintero · Daniel Sierra-Sosa · Ibrahim Imam
+1 more
This review provides a comprehensive introduction to the current state of optimization through quantum computing, as well as its main methods and highlights. It begins with an overview of the history and principles of quantum computing, followed by a concise explanation of classical optimization techniques. References to key mathematical tools essential for understanding quantum optimization are introduced to assist readers new to the field. The review then explores various quantum optimization techniques, categorized by hardware requirements, pipelining strategies, kernel methods, order, continuity, heuristic and metaheuristic approaches. Graphs and charts are included to illustrate the interconnections between these techniques, enhancing understanding of their complementarity. Finally, the review concludes with some considerations about societal, economic and environmental impacts and a detailed table summarizing the different methods and their applications.