{"title":"Adaptive security framework for multi-environment networks using ensemble data drift detection and incremental deep learning","authors":"Furqan Rustam, Anca Delia Jurcut","doi":"10.1016/j.jisa.2025.104219","DOIUrl":"10.1016/j.jisa.2025.104219","url":null,"abstract":"<div><div>Modern multi-environment (M-En) networks comprise diverse architectures such as IoT and traditional IP-based networks. These networks pose significant challenges for threat mitigation due to heterogeneous protocols and traffic patterns. This study proposes a unified incremental learning framework to efficiently secure M-En networks by reducing management overhead, improving scalability, and lowering costs. We designed this approach for real-time environments, enabling adaptation to new scenarios with high accuracy and efficiency. To develop the framework, we first generate an M-En dataset using partial least squares canonical analysis, synthesizing data from two benchmark datasets: IoT23 and CICDDoS2019, representing IoT and traditional IP-based networks, respectively. Our approach employs an ensemble data drift detection (EDDD) mechanism that combines ADaptive WINdowing and autoencoders, enabling adaptive model updates. A deep neural network is incrementally retrained only when data drift is detected, ensuring adaptability to evolving attacks while conserving computational resources. To avoid catastrophic forgetting, we incorporate replay-based memory, regularization, and an interpolation mechanism governed by a blending parameter <span><math><mrow><mi>α</mi><mo>∈</mo><mrow><mo>[</mo><mn>0</mn><mo>,</mo><mn>1</mn><mo>]</mo></mrow></mrow></math></span>, which balances the integration of new and historical knowledge. Furthermore, the explainable AI technique LIME is integrated to enhance the transparency of the model’s decision-making process. Experimental results indicate that our approach achieves a mean accuracy of 0.999 while maintaining low memory usage, approximately 32.1 MB, and a stable model size of 0.11 MB.</div></div>","PeriodicalId":48638,"journal":{"name":"Journal of Information Security and Applications","volume":"94 ","pages":"Article 104219"},"PeriodicalIF":3.7,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Qizhen Jia , Kai Kang , Beier Wang , Yuehao Wu , Yibo Zhang , Fusen Guo
{"title":"Research on wind power prediction with secondary decomposition and multi-algorithm fusion for complex nonlinear time series","authors":"Qizhen Jia , Kai Kang , Beier Wang , Yuehao Wu , Yibo Zhang , Fusen Guo","doi":"10.1016/j.compeleceng.2025.110688","DOIUrl":"10.1016/j.compeleceng.2025.110688","url":null,"abstract":"<div><div>Wind power forecasting plays a critical role in maintaining grid stability and enhancing electricity market competitiveness. However, the inherent nonlinearity and noise in wind power data—driven by multiple influencing factors such as wind speed and temperature—pose significant challenges for accurate prediction. To address this, we propose a hybrid forecasting framework that integrates secondary decomposition with multiple predictive algorithms. First, Singular Spectrum Analysis (SSA) is employed to extract the primary trend and periodic components, followed by Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) for further signal decomposition. Fuzzy entropy is then used to reconstruct the components into high-frequency, mid-frequency, and trend sub-series, effectively denoising the data and capturing multiscale features. For forecasting, an IDF-Informer model—incorporating Intra-layer Directional Fusion and optimized via a multi-strategy Improved Sand Cat Swarm Optimization (ISCSO)—is used for the high- and mid-frequency components, while the trend component is predicted using a Random Vector Functional Link Network without Direct(RVFLwoDL). Experimental results across the four seasons show that the proposed method achieves consistent improvements in RMSE, MAE, MAPE, and <span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> compared to other models.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"128 ","pages":"Article 110688"},"PeriodicalIF":4.9,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049325","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"NICGM: A non-intrusive CSI-guided modulation method for image semantic communication","authors":"Weiming Niu , Mingru Dong , Benzhai Hai , Ying Zhang","doi":"10.1016/j.phycom.2025.102832","DOIUrl":"10.1016/j.phycom.2025.102832","url":null,"abstract":"<div><div>The incorporation of channel state information (CSI) has been proven effective in enhancing the robustness of transmission in image semantic communication systems. However, existing studies have largely overlooked the modal inconsistency between CSI and semantic features. Specifically, the intrusion of physical features into the semantic space can significantly degrade the reconstruction quality and hinder practical deployment under complex channel conditions. To address this issue, a novel non-intrusive CSI-guided modulation mechanism is proposed in this paper. The mechanism integrates affine modulation and gating control to structurally guide the fusion of CSI into the semantic feature space, effectively mitigating the destructive interference of CSI. Furthermore, a lightweight CSI encoder is introduced to accommodate feedback distortion and compression constraints. A modulation perturbation regularization term is designed to dynamically control the modulation intensity. During system training, perceptual loss and structural reconstruction loss are jointly optimized, while a modulation deviation monitoring indicator is employed to enhance non-intrusive behavior constraints. Experimental results demonstrate that, under low signal-to-noise ratio and distorted CSI scenarios, the proposed NICGM significantly outperforms baseline methods such as JSCC and JPEG in terms of PSNR and perceptual consistency. These findings validate the structural stability and performance advantages of the proposed model in non-ideal wireless environments.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"73 ","pages":"Article 102832"},"PeriodicalIF":2.2,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049647","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cecilia Laschi, Li Wen, Fumiya Iida, Arsen Abdulali, Helmut Hauser, Yifan Wang, Ke Liu, Leonardo Ricotti, Matteo Cianchetti, Kaspar Althoefer, Pham Huy Nguyen, Mirko Kovac, Marcello Calisti
{"title":"Soft robotics: what's next in bioinspired design and applications of soft robots?","authors":"Cecilia Laschi, Li Wen, Fumiya Iida, Arsen Abdulali, Helmut Hauser, Yifan Wang, Ke Liu, Leonardo Ricotti, Matteo Cianchetti, Kaspar Althoefer, Pham Huy Nguyen, Mirko Kovac, Marcello Calisti","doi":"10.1088/1748-3190/ae066d","DOIUrl":"https://doi.org/10.1088/1748-3190/ae066d","url":null,"abstract":"<p><p>The field of soft robotics has shown unprecedented growth in research efforts, scientific achievements, and technological advancements. Bioinspiration and biomimetics have played an instrumental role in the birth and growth of soft robotics. What is next for this field? To promote soft robotics research to the next level and have a broader impact in robotics and engineering fields, in this roadmap, we argue that two research directions should be strengthened i) more structured, formal methods and tools for designing and developing soft robots and bioinspired robots ii) more concrete applications of bioinspired soft robots in diverse sectors of human activities. This article provides a roadmap for the design of bioinspired soft robots, the integration of soft robot systems, and their applications in industry and services. Scientists and experts describe the state-of-the art and the perspectives of bioinspired, model-informed design of soft robots, outlining the challenges in developing complex soft robotic systems, and applications of soft robots in diverse fields.
.</p>","PeriodicalId":55377,"journal":{"name":"Bioinspiration & Biomimetics","volume":" ","pages":""},"PeriodicalIF":3.0,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145056295","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A unified multispectral multitasking network with synchronous qualitative and quantitative analysis","authors":"Shui Yu , Qian Ni , Keyang Xia , Kewei Huan , Hongna Zhu","doi":"10.1016/j.chemolab.2025.105536","DOIUrl":"10.1016/j.chemolab.2025.105536","url":null,"abstract":"<div><div>With the advancement of spectral analysis technology, non-destructive testing plays a crucial role in various fields such as agriculture, petrochemical, medicine, food, and forage. The multispectral multitasking methods not only exhibit feasibility but also hold significant potential for enhancing model predicted accuracy and generalization capabilities. A unified multispectral multitasking network of convolutional neural network combined with Transformer (SQQMulSNet) is proposed with achieving synchronous qualitative and quantitative analysis. A parallel convolution module (PCM), a multi-dimensional feature fusion module (MFFM), with classification and regression modules, are designed to construct SQQMulSNet. The PCM extracts spectral feature information through convolution and pooling operations. The MFFM enhances predicted accuracy by analyzing the underlying structures of spectral data through CEncoder and CDecoder. The classification and regression modules synchronous predict the types and contents of substances. Moreover, SQQMulSNet is tested on two public datasets of mango and melamine, and conducts ablation experiments. Comparisons are given between SQQMulSNet and classical CNNs, as well as commonly employed qualitative/quantitative analysis models. The results indicate that SQQMulSNet provides improved results than other modeling methods. SQQMulSNet accomplishes synchronized predictions for qualitative and quantitative analysis, attaining high predicted accuracy and enhanced generalization capabilities. This study establishes a crucial foundation for developing a non-destructive and accurate multispectral multitasking network.</div></div>","PeriodicalId":9774,"journal":{"name":"Chemometrics and Intelligent Laboratory Systems","volume":"267 ","pages":"Article 105536"},"PeriodicalIF":3.8,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145057115","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"化学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wenbin Jiang , Dongmei Zhang , Hong Cao , Xiaofeng Wang
{"title":"A search method for fractured-vuggy reservoir inter-well connectivity path based on multi-modal multi-agent","authors":"Wenbin Jiang , Dongmei Zhang , Hong Cao , Xiaofeng Wang","doi":"10.1016/j.engappai.2025.112184","DOIUrl":"10.1016/j.engappai.2025.112184","url":null,"abstract":"<div><div>The complex geological structure of carbonate reservoirs and the intricate fracture-vuggy configurations obscure inter-well connectivity, making its evaluation challenging. Conventional studies primarily rely on seismic static data to delineate fracture-vuggy reservoirs, but the limited recognition accuracy hampers the precise characterization of inter-well connectivity and the spatial configuration of fractures and vugs. To address this, this study constructs a 3D (Three-Dimensional) search environment and use multi-modal static and dynamic data and proposes a multi-agent connected channel search model based on deep reinforcement learning. The model treats multiphase fluid as an agent and incorporates Swin Transformer (Shift Window Transformer) to extract large-scale fracture features from seismic data, providing global prior information for path search. A Graph Attention Network is established based on dynamic response relationships to extract spatial geological features, while a multi-head self-attention mechanism captures real-time fluid interactions in various directions. The model fuses multi-modal features, including seismic attributes and production data, to generate decisions and automatically search for inter-well connectivity channels. Experiments were conducted using the WE1 and WE5 well groups from the fault-controlled karst reservoirs in the Tahe oilfield, with results compared against tracer tests. The findings demonstrate that the proposed model's automatic search paths closely align with seismic data and tracer test results, effectively capturing the spatial distribution of fractures and vugs across different scales. This validates the model's effectiveness in evaluating inter-well connectivity in complex carbonate reservoirs.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"161 ","pages":"Article 112184"},"PeriodicalIF":8.0,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145046191","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"DR-MIM: Zero-shot cross-lingual transfer via disentangled representation and mutual information maximization","authors":"Wenwen Zhao, Zhisheng Yang, Li Li","doi":"10.1016/j.ipm.2025.104389","DOIUrl":"10.1016/j.ipm.2025.104389","url":null,"abstract":"<div><div>Multilingual models have made significant progress in cross-lingual transferability through large-scale pretraining. However, the generated global representations are often mixed with language-specific noise, limiting their effectiveness in low-resource language scenarios. This paper explores how to more efficiently utilize the representations learned by multilingual pretraining models by separating language-invariant features from language-specific ones. To this end, we propose a novel cross-lingual transfer framework, DR-MIM, which explicitly decouples universal and language-specific features, reduces noise interference, and improves model stability and accuracy. Additionally, we introduce a mutual information maximization mechanism to strengthen the correlation between universal features and model outputs, further optimizing the quality of semantic representations. We conducted a systematic evaluation of this method on three cross-lingual natural language understanding benchmark datasets. On the TyDiQA dataset, DR-MIM improved the F1 score by 1.7% and the EM score by 4.5% over the best baseline. To further validate the model’s generalization capability, we introduced two new tasks: paraphrase identification and natural language inference, and designed both within-language and cross-language analysis experiments. All experiments collectively covered 22 languages. Further ablation studies, generalization analysis, and visualization results all confirm the effectiveness and adaptability of our approach.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104389"},"PeriodicalIF":6.9,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049111","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Riski Kurniawan , Sri Redjeki Pudjaprasetya , Rani Sulvianuri
{"title":"Numerical study of two-dimensional sediment transport using momentum-conserving staggered grid scheme","authors":"Riski Kurniawan , Sri Redjeki Pudjaprasetya , Rani Sulvianuri","doi":"10.1016/j.jocs.2025.102714","DOIUrl":"10.1016/j.jocs.2025.102714","url":null,"abstract":"<div><div>Sediment transport plays a crucial role in the evolution of bed morphology through deposition and erosion. This study presents numerical simulations of two-dimensional sediment transport induced by fluid flow. The fluid-sediment interaction is governed by a capacity model, i.e., the coupled system of shallow water and Exner equations, a simplification of more physically advanced non-capacity models. The system is solved using a momentum-conserving staggered grid (MCS) scheme. Model validation is performed using the Meyer-Peter and Müller (MPM) bedload transport formula, applied to experimental data from dam-break flows in various channel configurations. The proposed method successfully reproduces trends in the evolution of the water surface and quasi-steady sediment profiles. In general, the MCS scheme provides more accurate water level predictions than the numerical benchmark schemes. Although the predictions of maximum depths of deposition and erosion are less accurate, the overall results are consistent with those obtained from non-capacity models. Furthermore, the model is applied to the Kampar River estuary to simulate sediment transport due to the tidal bore.</div></div>","PeriodicalId":48907,"journal":{"name":"Journal of Computational Science","volume":"92 ","pages":"Article 102714"},"PeriodicalIF":3.7,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049295","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jin Ren , Qinling Zhou , Shunzhi Yang , Jinfeng Yang
{"title":"Multi-source Feature Map Distillation for enhanced low-resolution object recognition","authors":"Jin Ren , Qinling Zhou , Shunzhi Yang , Jinfeng Yang","doi":"10.1016/j.compeleceng.2025.110710","DOIUrl":"10.1016/j.compeleceng.2025.110710","url":null,"abstract":"<div><div>Knowledge distillation is an effective method for addressing the problem of low-resolution object recognition. However, due to resolution differences, the feature map sizes become inconsistent, making it difficult for the student model to fully learn the rich privileged information contained in the teacher model. Our previous work addressed this issue through a feature decoder, achieving cross-resolution feature map distillation. However, it fails to fully leverage both high-resolution samples and their feature maps to extract more privileged information at the distillation point. To this end, this paper proposes a Multi-source Feature Map Distillation (MsFMD) method to further improve the performance of low-resolution object recognition in practical applications such as intelligent video surveillance. We design a feature decoder with a channel attention mechanism to better leverage the privileged information from the teacher model and employ multi-level decoder modules to process deep features, achieving multi-level feature map distillation. Additionally, this paper introduces a joint data augmentation method, effectively enhancing the student model’s adaptability and robustness across samples with varying resolutions. The overall performance of MsFMD is verified in multiple recognition tasks by comparing it with state-of-the-art knowledge distillation methods on low-resolution and noisy objects.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"128 ","pages":"Article 110710"},"PeriodicalIF":4.9,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049320","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Satellite access points selection for cell-free LEO networks","authors":"Yuanbo Liu, Chaotian Lu, Weiyang Xu","doi":"10.1016/j.aeue.2025.156026","DOIUrl":"10.1016/j.aeue.2025.156026","url":null,"abstract":"<div><div>The proliferation of dense Low Earth Orbit (LEO) satellite constellations demands advanced architectures to manage their dynamic nature. This paper investigates a reconfigurable cell-free massive MIMO (CF-mMIMO) framework to overcome the limitations of static clustering. We develop optimization schemes using the alternating direction method of multipliers (ADMM), bisection, and interior-point methods to dynamically select satellite access points (SAPs) that maximize downlink sum-rate, uplink max–min fairness, and energy efficiency. Simulation results demonstrate that the proposed dynamic selection schemes yield substantial performance gains. Notably, our approach improves the downlink sum-rate by over 250% compared to a traditional single-satellite serving architecture. These findings validate dynamic clustering as a scalable and effective strategy for next-generation LEO systems.</div></div>","PeriodicalId":50844,"journal":{"name":"Aeu-International Journal of Electronics and Communications","volume":"202 ","pages":"Article 156026"},"PeriodicalIF":3.2,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049738","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}