Franklin Open最新文献

{"title":"A novel framework for intrusion detection in IOT networks using hybrid optimization algorithm and convolutional neural networks","authors":"M.V. Srikanth ,&nbsp;Pamarthi Sunitha ,&nbsp;Ravi Sankar Puppala ,&nbsp;Suneel Kumar Asileti ,&nbsp;A. Akshaykranth","doi":"10.1016/j.fraope.2025.100461","DOIUrl":"10.1016/j.fraope.2025.100461","url":null,"abstract":"<div><div>Due to extremely unpredictable and diverse network traffic, the rapid expansion of IoT devices has increased security risks. Conventional IDS models frequently fall short of maintaining high accuracy when dealing with unbalanced datasets and changing attack types. In order to increase the effectiveness and precision of network intrusion detection, this research presents an intrusion detection system for IoT networks. The proposed system makes use of a hybrid whale optimization algorithm-particle swarm optimization (WOA-PSO) for feature selection and convolutional neural networks (CNN) for network traffic classification. Robust and accurate attack classification is ensured by this hybrid optimization, which facilitates effective feature space search and exploitation. According to the comparative analysis, the suggested method obtains a much lower false alarm rate (FAR), a greater accuracy of 98.7 %, a precision of 99.3 %, and a recall of 98.3 %. These results demonstrate that the suggested approach is a trustworthy solution for IoT network security as it successfully improves intrusion detection, lowers false alarms, and guarantees a higher detection rate.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100461"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885155","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mathematical modelling of COVID-19 transmission dynamics with immigration and dual quarantine–isolation strategies","authors":"Queeneth Ojoma Ahman ,&nbsp;Remigius Okeke Aja ,&nbsp;Patrick Agwu Okpara ,&nbsp;Emmanuel Olorunfemi Senewo","doi":"10.1016/j.fraope.2026.100531","DOIUrl":"10.1016/j.fraope.2026.100531","url":null,"abstract":"<div><div>The COVID-19 pandemic remains a major global health challenge, particularly in low- and middle-income countries where healthcare capacity is limited and individual health-seeking behaviour strongly influences transmission dynamics. Motivated by this context, we formulate and analyse a deterministic compartmental model that incorporates dual intervention pathways — government-managed and personally adopted quarantine and isolation — together with immigration of exposed and infectious individuals and return-to-susceptible dynamics. The basic reproduction number <span><math><msub><mrow><mi>B</mi></mrow><mrow><mn>0</mn></mrow></msub></math></span> is derived and its dependence on key epidemiological and intervention parameters is investigated through sensitivity analysis. Numerical simulations, calibrated using early outbreak data from Nigeria, demonstrate how the interaction between quarantine, isolation, and mobility shapes epidemic trajectories under varying levels of compliance. Sensitivity results reveal that transmission intensity, natural mortality, and susceptible recruitment exert the strongest influence on <span><math><msub><mrow><mi>B</mi></mrow><mrow><mn>0</mn></mrow></msub></math></span>, while both government and personal isolation significantly mitigate transmission when effectively implemented. These results highlight the importance of accounting for heterogeneous behavioural responses and multi-route interventions in mathematical models of COVID-19, especially in resource-constrained settings. The proposed framework therefore provides modelling insights that can support the design and evaluation of context-specific control strategies for COVID-19 and other emerging infectious diseases.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100531"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146187860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring pancreatic cancer cell interactions: A fractional mathematical framework for siRNA therapy assessment","authors":"Mihir Thakkar,&nbsp;Nimisha Pathak ,&nbsp;Dhara Patel ,&nbsp;Anil Chavada","doi":"10.1016/j.fraope.2026.100530","DOIUrl":"10.1016/j.fraope.2026.100530","url":null,"abstract":"<div><div>This research investigates the efficacy of siRNA therapy in mitigating pancreatic cancer through a fractional-order mathematical framework. Utilizing the Caputo–Fabrizio derivative to capture memory-dependent tumor–immune–cytokine interactions, we formulate and rigorously analyze distinct systems for untreated evolution and siRNA-mediated intervention. Following the establishment of mathematical well-posedness and stability via the basic reproduction number, numerical simulations demonstrate that siRNA treatment effectively curtails tumor growth and bolsters sustained immune responses. Validating the efficacy of fractional-order modeling in portraying memory-driven dynamics, the findings biologically indicate that silencing tumor-promoting cytokines facilitates the restoration of immune function, thereby retarding cancer progression. Ultimately, this study underscores the critical role of immune feedback and temporal cellular dynamics in refining future siRNA therapeutic strategies.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100530"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146187856","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Modified fletcher–reeves spectral conjugate gradient algorithms with three-term structure for optimization and image restoration","authors":"Hisham M. Khudhur ,&nbsp;Sani Aji ,&nbsp;Sulaiman Mohammed Ibrahim","doi":"10.1016/j.fraope.2026.100517","DOIUrl":"10.1016/j.fraope.2026.100517","url":null,"abstract":"<div><div>The conjugate gradient (CG) method is a widely used technique for solving large-scale unconstrained optimization problems due to its simplicity and low memory requirements. The classical Fletcher-Reeves (FR) method, known for its strong theoretical convergence properties, often exhibits slow practical performance and is prone to jamming when used with inexact line searches or on poorly scaled problems. To overcome these limitations, we propose two new spectral three-term conjugate gradient algorithms, TTUHS1 and TTUHS2, which incorporate spectral scaling and a three-term update structure inspired by FR. The proposed algorithms are designed to retain the global convergence and descent properties of FR while significantly improving numerical efficiency. Extensive numerical experiments on standard test functions and image restoration problems demonstrate that TTUHS1 and TTUHS2 outperform the Three-Term Fletcher-Reeves (TTFR) algorithm in terms of convergence speed, iteration count, and robustness, particularly in large-scale problems.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100517"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146188030","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Semi-automatic approach utilizing Siamese Neural Network for forensic voice comparison","authors":"S.G. Kruthika ,&nbsp;Trisiladevi C. Nagavi ,&nbsp;P. Mahesha ,&nbsp;H.T. Chethana ,&nbsp;Vinayakumar Ravi ,&nbsp;Alanoud Al Mazroa","doi":"10.1016/j.fraope.2026.100527","DOIUrl":"10.1016/j.fraope.2026.100527","url":null,"abstract":"<div><div>Forensic Voice Comparison (FVC) remains a critical yet challenging task in digital forensics, often hindered by manual subjectivity, background noise, and speaker variability. This paper presents a novel semi-automatic FVC framework based on Siamese Neural Networks (SNN), a discriminative metric-learning architecture combined with stationary noise reduction for robust voice similarity assessment. Proposed framework leverages the SNN’s ability to learn a shared embedding space where Euclidean distance reflects speaker identity. Using a jurisdiction-specific dataset of 3899 Australian English speech samples (FLAC format), proposed framework achieves 96.02% accuracy, 94.00% precision, and 92.10% recall in distinguishing same vs different speaker pairs. The proposed framework is evaluated against strong baselines including Convolutional Neural Networks (CNN), Bidirectional Long Short-Term Memory (BiLSTM), Gaussian Mixture Model-Universal Background Model (GMM-UBM), and validated via 5-fold cross-validation (mean ± std. dev.) to ensure statistical robustness. Proposed framework fills a critical gap in forensic phonetics by demonstrating that lightweight, interpretable, pairwise deep learning models can outperform complex generative or ensemble systems in real-world FVC scenarios. All preprocessing, training protocols, and hyperparameters are documented for reproducibility.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100527"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147420946","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Federated micro-expression mining and multi-modal metadata fusion for Deepfake fraud detection in ubiquitous financial video-KYC systems at IoT network","authors":"Romil Rawat ,&nbsp;Anjali Rawat ,&nbsp;Shweta Gupta ,&nbsp;A. Samson Arun Raj ,&nbsp;T.M. Thiyagu ,&nbsp;Hitesh Rawat ,&nbsp;Anand Rajavat","doi":"10.1016/j.fraope.2026.100523","DOIUrl":"10.1016/j.fraope.2026.100523","url":null,"abstract":"&lt;div&gt;&lt;div&gt;&lt;strong&gt;Introduction &amp; Problem Statement-&lt;/strong&gt; The increasing sophistication of AI-generated deepfakes poses significant challenges for financial video-KYC systems, where identity verification relies on accurate and real-time analysis of user biometrics. Traditional centralized and unimodal detection models struggle to balance accuracy, privacy, and deployment scalability, particularly across heterogeneous IoT edge devices. &lt;strong&gt;Need for Research-&lt;/strong&gt;There is a pressing need for privacy-preserving, scalable, and robust deepfake detection mechanisms capable of identifying subtle manipulations in real-world financial environments. Current solutions often fail under domain-shift conditions, low-resolution inputs, or in scenarios involving complex micro-expression and behavioral cues. &lt;strong&gt;Proposed Work &amp; Objective-&lt;/strong&gt; This research proposes the &lt;strong&gt;Federated Micro-Expression Mining and Multi-Modal Metadata Fusion (FED-MEMF)&lt;/strong&gt; framework, designed to accurately detect deepfake fraud in decentralized video-KYC systems. The objectives are to (i) enhance detection accuracy by leveraging facial micro-expression dynamics, audio signals, and session metadata, and (ii) preserve user privacy through federated learning while ensuring low-latency real-time inference. &lt;strong&gt;Novelty-&lt;/strong&gt; The novelty lies in integrating fine-grained micro-expression analysis with behavioral metadata fusion in a &lt;strong&gt;federated learning environment&lt;/strong&gt;, combined with cross-modal attention mechanisms. This approach enables robust detection across multiple datasets while maintaining privacy and edge-device compatibility. &lt;strong&gt;Method-&lt;/strong&gt; The framework employs modality-specific encoders—μ-Transformer for micro-expressions, CNN for audio, and LSTM for metadata—with features fused via a cross-modal attention engine. Federated Averaging (FedAvg) aggregates local model updates from IoT edge devices without transferring sensitive data. Quantization and hardware optimizations enable real-time performance on low-power devices. &lt;strong&gt;Dataset-&lt;/strong&gt; Experiments utilized &lt;strong&gt;FaceForensics++, CAS(ME)^2&lt;/strong&gt;, and a proprietary &lt;strong&gt;KYC-FinVox2024&lt;/strong&gt; dataset comprising video, audio, and metadata streams, including micro-expression labels, to evaluate both intra- and cross-dataset performance. &lt;strong&gt;Results-&lt;/strong&gt; FED-MEMF achieved an overall accuracy of &lt;strong&gt;98.7%&lt;/strong&gt;, F1-score of &lt;strong&gt;0.987&lt;/strong&gt;, AUC of &lt;strong&gt;0.996&lt;/strong&gt;, and inference latency of &lt;strong&gt;82&lt;/strong&gt; &lt;strong&gt;ms&lt;/strong&gt;, outperforming XceptionNet, EfficientNet-B4, and CNN+LSTM baselines. Multi-modal fusion significantly reduced false positives and false negatives, demonstrating robustness under domain-shift conditions. &lt;strong&gt;Conclusion &amp; Future Work-&lt;/strong&gt; FED-MEMF provides a &lt;strong&gt;privacy-conscious, real-time, and scalable solution&lt;/strong&gt; for deepfake detection in financial video-KYC applications. Fu","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100523"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147421114","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Rough set based feature selection model for diabetic retinopathy classification","authors":"Abhishek Bhattacharya,&nbsp;Blerta Prevalla Etemi,&nbsp;Debabrata Samanta","doi":"10.1016/j.fraope.2026.100491","DOIUrl":"10.1016/j.fraope.2026.100491","url":null,"abstract":"<div><div>Medical image processing is an essential challenge in a wide range of applications in today’s clinical scenario. Such applications can be served throughout the clinical course of events, not just in the diagnostic environment but also in the planning, execution, and progression of surgical and radiation procedures. The role of medical imaging information retrieval and processing is significant in surgical planning and tracking the progress of diseases. So, using state of the art computing techniques, researchers have made efforts to propose an effective automated technique to determine Diabetic Retinopathy (DR) based on significant medical image features and the patient’s clinical history. In this work, an intelligent graph-based methodology is proposed, considering the concepts from Rough Set Theory for feature selection. Based on several centrality metrics of graphs, a voting method is proposed to identify important features, resulting in better classification outcomes for diabetic retinopathy. Proposed methods are compared with several existing baseline feature selection approaches and provide better feature selection outcomes than those existing approaches. The result shows significantly better classification outcomes with respect to classical classification approaches.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100491"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977463","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Conv-ViT: An improved discrete convolution-based vision transformer for diabetic retinopathy detection","authors":"B. Chitradevi ,&nbsp;P. Mathiyalagan ,&nbsp;A. Ramachandran ,&nbsp;R. Dhanapal ,&nbsp;K. Sheikdavood ,&nbsp;S. Gnanamurugan","doi":"10.1016/j.fraope.2025.100477","DOIUrl":"10.1016/j.fraope.2025.100477","url":null,"abstract":"<div><div>Timely detection of Diabetic Retinopathy (DR), a major cause of irreversible blindness, is important to avert vision impairment. Present computer-aided diagnostic methods often suffer from poor segmentation, image noise, and a lack of generalization across datasets. This study proposes Conv-ViT, a hybrid model that integrates convolutional networks to overcome the disadvantages of the aforementioned models. Probability-based particle swarm optimization (PBPSO) was applied to achieve accurate segmentation, median filtering was applied to remove noise, and local binary pattern (LBP) was used to extract texture features. An innovative Electric Fish Optimization Arithmetic Algorithm (EFAOA), which enhances the compromise between exploration and exploitation during hyperparameter fine-tuning, was introduced to bolster model efficiency. Evaluation on the MESSIDOR dataset showed remarkable results, with an accuracy of 99.58 %, 98.86 %, 98.87 %, and an F1-score of 98.85 %, respectively. These results indicate the generalizability of the model beyond that of current state-of-the-art algorithms. The Conv-ViT framework represents a robust and scalable solution for the early detection of DR and holds great promise for use in automated cloud-based diagnostic systems.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100477"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine learning-based smishing detection using fuzzy logic and TF-IDF feature engineering","authors":"Santosh Kumar Birthriya,&nbsp;Priyanka Ahlawat,&nbsp;Ankit Kumar Jain","doi":"10.1016/j.fraope.2026.100506","DOIUrl":"10.1016/j.fraope.2026.100506","url":null,"abstract":"<div><div>Mobile communication security is increasingly threatened by smishing messages, necessitating advanced detection techniques to protect users from fraudulent and malicious content. This paper presents a hybrid approach that combines Term Frequency–Inverse Document Frequency (TF-IDF) with fuzzy membership–based linguistic and structural features to enhance smishing messages classification. The feature extraction process includes word count, punctuation usage, message length, sentiment polarity, capitalization patterns, and digit frequency. Fuzzy membership functions encode these attributes as gradual values rather than fixed thresholds, improving adaptability to evolving smishing patterns. These fuzzy features are concatenated with TF-IDF vectors to form a comprehensive representation that captures both semantic and stylistic characteristics. The proposed framework is evaluated on a dataset of 6119 SMS messages, comprising 5574 messages from the SMS Spam Collection v.1 and an additional 545 smishing messages from the Smishtank repository. Experimental results demonstrate that the proposed model achieves up to 99.10% accuracy, 99.30% precision, and 94% recall, outperforming existing methods such as SVM (97.40%) and Random Forest (98.15%). Furthermore, the incorporation of fuzzy membership concepts enhances adaptability to diverse smishing patterns, reduces false alarms, and improves the overall robustness of the classification framework.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100506"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146090637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comprehensive models for antenna measurement utilizing compressive sensing in multi-probe systems","authors":"Mohammad Reza Ghaderi","doi":"10.1016/j.fraope.2026.100498","DOIUrl":"10.1016/j.fraope.2026.100498","url":null,"abstract":"<div><div>Measuring the radiation pattern of complex antennas can often be a time-consuming process. As a result, researchers are seeking techniques that can reduce measurement time while maintaining accuracy. Compressive sensing is a powerful method in signal processing that can be applied to various types of measurements. It enhances systems by reducing energy consumption, optimizing memory usage, increasing measurement speed, and ultimately shortening measurement time. Although compressive sensing has seen significant advancements in theoretical research, designing systems and application models based on this technique requires a thorough understanding of its strengths and limitations. In this study, we evaluate CS-based antenna measurement models using simulated datasets of antenna. Results show that measurement time can be reduced by up to 60% while maintaining reconstruction accuracy within 5% RMSE. These findings highlight the efficiency of CS in antenna characterization.</div></div>","PeriodicalId":100554,"journal":{"name":"Franklin Open","volume":"14 ","pages":"Article 100498"},"PeriodicalIF":0.0,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146187858","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}