Archives
Abstract: Plant extract–mediated synthesis of zinc oxide nanoparticles (ZnO NPs) is an eco-friendly method that utilizes phytochemicals as natural reducing and stabilizing agents, to produce biocompatible and stablenanoparticles with tunable properties. This study reports the synthesis of ZnO Nanoparticles utilizing anaqueous extract of Bombax ceiba flower petals (BC). The characterization of BC-ZnO NPs was carried out using UV–visible (UV-Vis.), Powder XRD, FTIR, and FESEM-EDS spectroscopic methods. PXRD pattern confirms highly crystalline ZnONPs with a wurtzite hexagonal phase.The FESEM images revealed the flake-like or petal-shaped appearance of the particles, giving a flower-like or agglomerated granular morphology. BC-ZnO NPs with an average size ~41.89 nm were obtained. BC-ZnO NPs exhibited ~90% photodegradation of MB dye during initial 40 min of UV light irradiation. Further, the BC-ZnO NPs were also examined for their biological activities: antimicrobial, antioxidant, and antidiabetic. BC-ZnO NPs showed excellent antioxidant capability and good antimicrobial potential against C. albicans, S. aureus, and E. coli. They showed better antidiabetic activity (α-amylase activity) than the standard, with an IC50 value of 119.95 µg/mL. While for α-glucosidase, the BC-ZnO NPs exhibited excellent inhibition activity achievingan IC50 value of 98.94 µg/mL.The lower IC50 value (46.34 µg/mL) of the synthesized NPs was represented as their excellent antioxidant capability.
Abstract: The increasing penetration of renewable energy sources (RES) like solar and wind has introduced significant power quality issues, primarily due to harmonic distortions. Active Power Filters (APFs) have emerged as effective solutions for mitigating these challenges. This review explores the integration of Silicon Carbide (SiC) MOSFET technology with Park’s Transformation-based d–q control strategies in APF systems. SiC MOSFETs offer superior electrical characteristics, including high switching frequency, thermal efficiency, and voltage blocking capability, making them highly suitable for modern APFs. Park’s Transformation facilitates effective decoupling of active and reactive power components, enabling precise harmonic compensation. Various control methods—ranging from conventional PI to advanced predictive and AI-based controllers—are analyzed for their suitability in d–q reference frames. Case studies demonstrate significant Total Harmonic Distortion (THD) reduction using SiC-based APFs. The paper also highlights technical challenges such as EMI, thermal management, and cost, proposing research directions to enhance system performance, scalability, and cost-effectiveness in future renewable-integrated power systems.
Abstract: The Cyber Supply Chain (CSC) system is intricate, comprising various subsystems tasked with different functions. Securing this supply chain is challenging due to inherent vulnerabilities, which can be exploited anywhere within it, posing a significant risk to business continuity. Therefore, it's crucial to comprehend and anticipate potential threats to implement adequate security measures. Cyber Threat Intelligence (CTI) offers insights into identifying threats, utilizing factors such as threat actor skills, motivation, Tactics, Techniques, and Procedures (TTPs), and Indicators of Compromise (IoCs). This study aims to analyze and anticipate threats to enhance cyber supply chain security by leveraging CTI alongside Machine Learning (ML) techniques. By doing so, inherent vulnerabilities in the CSC can be pinpointed, enabling organizations to take appropriate control measures for overall cybersecurity enhancement.to validate our approach, CTI data was collected and several ML algorithms, including Logistic Regression (LG), Support Vector Machine (SVM), Random Forest (RF), Decision Tree (DT), Cat Boost, and Gradient Boost, were employed using the Microsoft Malware Prediction dataset. The experiment focused on input parameters such as attack and TTP, with vulnerabilities and Indicators of Compromise (IoC) as output parameters. Results from the predictive analytics indicated that Spyware/Ransomware and spear phishing were the most foreseeable threats within the CSC. Additionally, we suggested relevant controls to mitigate these threats. We advocate the utilization of CTI data for ML-based predictive modeling to bolster overall CSC cybersecurity.
Abstract: Deep learning is a subfield of machine learning that has been driving the technological advancements of recent times. It utilizes artificial neural networks to solve complex problems that were previously thought to be unsolvable by traditional machine learning methods. This paper provides an overview of deep learning techniques and its applications in various domains. The paper starts with a brief introduction to deep learning, followed by existing methods with a comparison of their strengths and limitations. The paper also highlights the research gaps in the existing methods and concludes with future scopes of research in deep learning. The paper concludes with a list of references for further reading.
Abstract: Satellite image classification plays a crucial role in remote sensing applications, ranging from land use monitoring to disaster management and environmental assessment. However, the inherent challenges of satellite imagery—such as high dimensionality, spectral variability, noise, and texture complexity—demand robust feature extraction and classification methods. This review provides a comprehensive analysis of feature-based satellite image classification techniques, with a special focus on Extended Local Binary Patterns (ELBP) and Support Vector Machines (SVM). ELBP enhances conventional texture descriptors by incorporating rotation-invariant and multi-radius neighborhood information, making it suitable for capturing fine-grained spatial patterns in satellite scenes. SVM, known for its strong generalization capabilities, is widely adopted for high-dimensional image classification, particularly when integrated with ELBP-derived features. The review contrasts handcrafted and deep learning-based feature extraction techniques, highlights their strengths and limitations, and evaluates classifier performances across various benchmark datasets. Recent literature using ELBP–SVM integration demonstrates consistent improvements in classification accuracy, especially in multiclass and noise-prone environments. Finally, key research gaps are identified, including the need for explainable models, real-time deployment, and unified benchmarking standards. This review serves as a valuable reference for researchers and practitioners seeking effective, interpretable, and scalable solutions for satellite image analysis.