Altınbaş University Institutional Repository
Not a member yet
5805 research outputs found
Sort by
End-to-End DAE-LDPC-OFDM Transceiver with Learned Belief Propagation Decoder for Robust and Power-Efficient Wireless Communication
Full text linkThis paper presents a Deep Autoencoder–LDPC–OFDM (DAE–LDPC–OFDM) transceiver architecture that integrates a learned belief propagation (BP) decoder to achieve robust, energy-efficient, and adaptive wireless communication. Unlike conventional modular systems that treat encoding, modulation, and decoding as independent stages, the proposed framework performs end-to-end joint optimization of all components, enabling dynamic adaptation to varying channel and noise conditions. The learned BP decoder introduces trainable parameters into the iterative message-passing process, allowing adaptive refinement of log-likelihood ratio (LLR) statistics and enhancing decoding accuracy across diverse SNR regimes. Extensive experimental results across multiple datasets and channel scenarios demonstrate the effectiveness of the proposed design. At 10 dB SNR, the DAE–LDPC–OFDM achieves a BER of 1.72% and BLER of 2.95%, outperforming state-of-the-art models such as Transformer–OFDM, CNN–OFDM, and GRU–OFDM by 25–30%, and surpassing traditional LDPC–OFDM systems by 38–42% across all tested datasets. The system also achieves a PAPR reduction of 26.6%, improving transmitter power amplifier efficiency, and maintains a low inference latency of 3.9 ms per frame, validating its suitability for real-time applications. Moreover, it maintains reliable performance under time-varying, interference-rich, and multipath fading channels, confirming its robustness in realistic wireless environments. The results establish the DAE–LDPC–OFDM as a high-performance, power-efficient, and scalable architecture capable of supporting the demands of 6G and beyond, delivering superior reliability, low-latency performance, and energy-efficient communication in next-generation intelligent networks
Blockchain-Integrated Secure Authentication Framework for Smart Grid IoT Using Energy-Aware Consensus Mechanisms
Full text linkIntegrating IoT devices into smart grids raises some hard problems related to safe data sharing, the ability to grow, and energy use. Blockchain provides a safe way to check identities without a central authority. Typical ways to confirm transactions, like Proof-of-Work (PoW), use a lot of power, making them bad for devices that cannot use much energy. This study introduces a safe authentication system using Blockchain, a Deep Neural Network (DNN), and a power-saving way to confirm transactions (EACM). The system picks validators based on how much power they have left and their trust scores to save power during confirmation. Using the IoT-Enabled Smart Grid Dataset, simulations showed a transaction speed of 372 TPS, which is 32% better than normal methods. The system correctly authenticates 98.69% of the time, with a confirmation delay of 5.9 milliseconds and an 18% drop in validator node energy use. Also, the system spots 98.4% of unauthorized access tries, with a false acceptance rate (FAR) of 1.7% and a false rejection rate (FRR) of 0.31%. These outcomes prove the system’s ability to offer safe, fast, and energy-saving authentication for big, real-time Smart Grid IoT setups
Hybrid Harris Hawks optimization with eagle strategy particle swarm optimization for stability and disturbance rejection in tethered UAV systems
Full text linkIn the paper, a new hybrid optimization method is proposed to design an advanced attitude controller of tethered UAVs. The suggested methodology builds on the use of a Fractional-Order Proportional-Derivative-Derivative Integral (FOPDD-I) controller, which is optimized with the help of a Hybrid Harris Hawks Optimization-Eagle Strategy-Particle Swarm Optimization (HHHOESPSO) algorithm. The FOPDD-I controller is compared with Conventional PID, Cascade PID, classical Active Disturbance Rejection Control (ADRC) and Advanced ADRC methodologies. Comprehensive MATLAB simulations indicate that the FOPDD-I controller, optimized through HHHOESPSO, is more stable, responds more quickly to disturbances, and can reject disturbances better than their traditional counterparts. Significant results are a 56.5 percent increase in Kp to achieve better overall stability, a 65.2 percent increase in Ki to achieve smaller steady-state errors and better roll control and a 98.4 percent decrease in Kd to achieve smaller yaw overshoot and oscillations. Also, the fractional-order parameters are adjusted to adaptability improvement of 12.5 and 14.7 in nonlinear dynamic environment. Conversely, traditional controllers including PID and Cascade PID demonstrate only minor tuning benefits, whereas both classical and state-of-the-art ADRC approaches realize moderate performance benefits, that are not as high as the stability and responsiveness demonstrated by FOPDD-I controller. This study underscores the effectiveness of combining fractional-order control with hybrid optimization strategies, establishing the proposed FOPDD-I controller as a robust solution for tethered UAV attitude control under dynamic and uncertain conditions
A Modified Empirical Model for Predicting Liquefaction in Partially Saturated Sands
No full textOne of the theoretical judgments for liquefaction triggering is that the maximum excess pore pressure ratio (ru,max) becomes unity. Some studies show that reducing the degree of saturation and creating partially saturated zones in the soil can lower ru,max, preventing liquefaction. Geotechnical engineers are interested in predicting ru,max as a control parameter for liquefaction triggering. Empirical models to predict ru,max of a free-field layer of partially saturated sand are mostly based on large and small-scale experimental setups, which cannot exactly explore the effect of high effective stresses. Also, the liquefaction responses obtained from the experimental studies mostly focus on shallow soil layers, limiting their applicability to deep field layers. This paper modifies an existing empirical model (RuPSS model) to make it able to capture the effect of high effective stresses up to 100 kPa in the prediction of ru,max of partially saturated sands in free-field conditions. A comparison of ru,max obtained from RuPSS model and the modified RuPSS model with the experimental test results confirms that the RuPSS model is unable to predict ru,max response under high effective stress while the modified RuPSS model can acceptably predict ru,max of partially saturated sands
Comparative evaluation of pediatric rotary file systems and hand files for root canal preparation in primary molars: an in vitro study
Full text linkBackground: The purpose of the study was to compare pediatric rotary files with hand K-files regarding the amount of dentin removal, root canal transportation, root canal surface area and volume in primary mandibular second molars using Cone-Beam Computed Tomography (CBCT). Methods: A total of 36 primary teeth were randomly divided into four groups; K-file, Fanta AF Baby file, EndoArt Ni-Ti Pedo file Gold Kit, MiniSCOPE Ni-Ti Gold Pediatric file. Samples were imaged with CBCT before and after canal instrumentation. For root canal surface area and volume measurements, 3Matic (Materialize, Belgium) software was used. Linear measurements were performed using NNT iRYS software. Data analysis was conducted using the Dunn’s test and the Kruskal Wallis test. A significance level of p < 0.05 was used. Results: Compared to pediatric rotary files, the K-file was shown to remove a statistically greater amount of dentin at the coronal level (p = 0.032). The difference in dentin thickness with the K-file was significantly greater than with the EndoArt file (p = 0.017) and the MiniSCOPE file (p = 0.007). The volume difference with the MiniSCOPE file was significantly less than with the Fanta file (p = 0.002) and the EndoArt file (p = 0.032). Root canal transportation was significantly greater with the K-file compared to the Fanta AF Baby file in both the oblique (p = 0.031) and buccal-lingual (p = 0.006) directions. Conclusions: Pediatric rotary files could be considered an efficient alternative to the hand K-file in biomechanical instrumentation. Three dimensional analysis can provide better comprehensive approach to evaluating the pediatric rotary instruments
Bandgap tuned indaceno based polymer design by machine learning for chemical space by descriptor designing and data guidelines
No full textArticle number : 109899This research introduces an innovative approach for developing bandgap-tuned polymers by machine learning assisted modeling of indaceno polymer chemical space. Using a dataset of indaceno donor moieties, the descriptors that encode fundamental molecular features are designed and trained to predict their bandgaps. After three modeling rounds of breaking retrosynthesis in Python, 1000 new polymers with their bandgaps are designed. The study shows that descriptors like valence electrons, Labute ASA, and Morgan Density has a significant impact on model performance to highlight their key role. Additionally, the synthetic accessibility scores of newly designed polymers reach a maximum of 17 for RDkit based descriptors to indicate their ease of synthesis and promising practical applicability. This work not only deepens the understanding of indaceno polymers but also lays the groundwork indaceno based polymer design of through data-driven approaches.Funding agency : Taif University
Grant number : TU-DSPP-2024-9
Harnessing AI for Leadership Development: Predictive Model for Leadership Assessment
No full textThe present paper has been devoted to the study conducted with the purpose of examining the possibility of applying Machine Learning techniques in classifying leadership based on structured survey data. The objective was to create a predictive model that would allow classifying leadership into three groups – Low, Medium, and High – based on behavior scores. The model was expected to offer a reliable tool for improving leadership development programs and recruitment processes by providing a precise and scalable leadership classification, The study illustrates the potential of advanced ML techniques for rethinking the traditional approaches to the assessment of leadership. Due to the use of advanced ensemble modeling, it was possible to ensure the high accuracy of 93.3% in leadership predicting. Such outcomes can generate considerable advantages for organizational development strategies. The use of ensemble machine learning in the domain of organizational behavior studies can be considered as a valuable academic contribution as it has demonstrated the capacity of determining the application of ensemble techniques for enhancing leadership studies. at the same time, it offers a useful instrument to develop more sophisticated and data-driven practices for leadership development
Trastuzumab deruxtecan (T-DXd) plus pertuzumab (P) vs taxane plus trastuzumab plus pertuzumab (THP) for first-line (1L) treatment of patients (pts) with human epidermal growth factor receptor 2-positive (HER2+) advanced/metastatic breast cancer (a/mBC): Interim results from DESTINY-Breast09.
No full text...Funding agency : AstraZeneca ; Daiichi Sankyo Company Limited
Optimization of Semi-Solid Lipid Nanoparticle Dispersions by Quality by Design Approach for Dermal Delivery of Curcumin
No full textCurcumin is an important anti-inflammatory agent for the treatment of skin disorders. However, its low water solubility, poor bioavailability, and instability limit the utilization of curcumin. Semi-solid lipid nanoparticle (SLN and NLC) dispersions, which maintain their colloidal particle size despite their high viscosity, offer a novel promising approach with high potential for dermal curcumin delivery. In this study, novel semi-solid SLN-NLC formulations of curcumin were manufactured using a one-step method, without the need to disperse the nanoparticles in an additional vehicle. Modde Pro 12 was used to examine the relationship between variables and quality attributes. QbD-based formulation optimization was successfully performed using artificial neural network program (ANN), and optimum semi-solid SLN-NLC formulations were prepared. The particle size of the optimum formulations was found to be 204.7 ± 1.5 nm for SS-SLN-Opt and 198.5 ± 0.81 nm for SS-NLC-Opt, indicating that the particle sizes were within the targeted range. The amount of curcumin released from the SS-NLC-Opt formulation was 33.72 ± 4.99% at 24th Hour, which was higher than the release obtained from the eight SS-NLC formulations entered as input into the ANN program. On the other hand, while the curcumin release percentage at the 24th Hour from the SS-SLN formulations entered into the program ranged between 11.13% and 44.31%, the release amount for the SS-SLN-Opt formulation was found to be 38.34 ± 3.48%, which was within this range and close to the maximum value. Rheological characterization results indicated that the optimum semi-solid SLN and NLC formulations were more elastic than viscous. The stability of the optimum semi-solid SLN formulation at 4 °C was higher than that of the optimum semi-solid NLC after one month. In vivo studies in rats revealed that the optimum semi-solid SLN formulation exhibited higher anti-inflammatory activity than both the optimum semi-solid NLC and the conventional gel. The SS-SLN-Opt formulation effectively reduced the inflammation in rats starting from the first hour. In conclusion, the optimum semi-solid SLN formulation, which is more stable and has higher anti-inflammatory activity, is a promising alternative for the dermal delivery of curcumin
Optimizing Smart Grid Load Forecasting via a Hybrid Long Short-Term Memory-XGBoost Framework: Enhancing Accuracy, Robustness, and Energy Management
No full textArticle number : 2842As renewable energy sources and distributed generation become more integrated into modern power systems, accurate short-term electricity load forecasting is increasingly critical for effective smart grid management. Most statistical techniques used in the analysis of time series models, conventional statistical models, often fail to account for temporal dependencies and inherent non-linear patterns found in real-world energy time series. Methods: To this end, merging the power of both the ML approaches, namely Long Short-Term Memory (LSTM) networks and XGBoost, into hybrid frameworks has become a powerful solution. This work aims to develop a new compound model of LSTM for time series pattern extraction from the temporal data and XGBoost for outstanding predictive performance. To assess the performance of the proposed model, we used the Elia Grid dataset from Belgium, which includes load data recorded every 15 min throughout 2022. Results: When compared to individual models, this hybrid approach outperformed them, achieving a Root Mean Square Error (RMSE) of 106.54 MW, a Mean Absolute Percentage Error (MAPE) of 1.18%, and a coefficient of determination (R2) of 0.994. Discussion: In addition, this study implements an ensemble learning strategy by combining LSTM and XGBoost to improve prediction accuracy and robustness. An experimental attempt to integrate attention mechanisms was also conducted, but it did not enhance the performance and was therefore excluded from the final model. The results extend the literature on the development of fusion-based machine learning models for time series forecasting, and the future work of energy consumption analysis, anomaly detection, and resource allocation in SM grids