alexandria engineering journal最新文献

{"title":"Incremental-learning-based graph neural networks on edge-forwarding devices for network intrusion detection","authors":"Qiang Gao ,&nbsp;Samina Kausar ,&nbsp;HuaXiong Zhang","doi":"10.1016/j.aej.2025.03.102","DOIUrl":"10.1016/j.aej.2025.03.102","url":null,"abstract":"<div><div>Graph neural networks have become one of the research hotspots for network intrusion detection due to their natural suitability for representing computer networks. However, most of the related research on training GNNs is centralized, and this approach involves long-distance transmission and dumping of network data, so it is inefficient to perform, has the potential for privacy leakage, and introduces an additional transmission burden to the network. To address these challenges, this paper investigates the feasibility of offloading both graph neural networks' training and inference phases to edge-forwarding devices such as switches. We propose a distributed framework that aggregates residual computational resources from edge-forwarding devices into a micro-computing network. This framework then migrates GNN execution to edge-forwarding devices through a hybrid parallelism paradigm, thus locally detecting network anomalies to reduce network data transmission significantly. Meanwhile, to address the problem of computational and memory constraints of edge-forwarding devices, we propose a novel attention heatmap-driven memoryless incremental learning algorithm that learns network features and detects anomalies with minimal resources while avoiding catastrophic forgetting. Finally, we implement and verify the feasibility of the above framework and algorithm using a general-purpose embedded system and open-source software. The experiments show that although each edge-forwarding device's computational and memory load is light, the framework performs similarly to traditional approaches. To the best of our knowledge, this is the first approach that offloads a graph neural network model to edge-forwarding devices.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 81-89"},"PeriodicalIF":6.2,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873306","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SAMFNet: Scene-aware sampling and multi-stage fusion for multimodal 3D object detection","authors":"Baotong Wang ,&nbsp;Chenxing Xia ,&nbsp;Xiuju Gao ,&nbsp;Bin Ge ,&nbsp;Kuan-Ching Li ,&nbsp;Xianjin Fang ,&nbsp;Yan Zhang ,&nbsp;Yuan Yang","doi":"10.1016/j.aej.2025.03.129","DOIUrl":"10.1016/j.aej.2025.03.129","url":null,"abstract":"<div><div>Recently, multimodal 3D object detection (M3OD) that fuses the complementary information from LiDAR data and RGB images has gained significant attention. However, the inherent structural differences between point clouds and images pose fusion challenges, significantly hindering the exploration of correlations within multimodal data. To address this issue, this paper introduces an enhanced multimodal 3D object detection framework (SAMFNet), which leverages virtual point clouds generated from depth completion. Specifically, we design a scene-aware sampling module (SASM) that employs tailored sampling strategies for different bins based on the density distribution of point clouds. This effectively alleviates the detection bias problem while ensuring the key information of virtual points, significantly reducing the computational cost. In addition, we introduce a multi-stage feature fusion module (MSFFM) that embeds point-level and regional-adaptive feature fusion strategies to generate more informative multimodal features by fusing features with different granularities. To further improve the accuracy of model detection, we also introduce a confidence prediction branch unit (CPBU), which improves the detection accuracy by predicting the confidence of feature classification in the intermediate stage. Extensive experiments on the challenging KITTI dataset demonstrate the validity of our model.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 90-104"},"PeriodicalIF":6.2,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873307","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Agricultural consumer Internet of Things devices: Methods for optimizing data aggregation","authors":"Hanan Abdullah Mengash ,&nbsp;Hany Mahgoub ,&nbsp;Asma Alshuhail ,&nbsp;Abdulbasit A. Darem ,&nbsp;Jihen Majdoubi ,&nbsp;Ayman Yafoz ,&nbsp;Raed Alsini ,&nbsp;Omar Alghushairy","doi":"10.1016/j.aej.2025.03.134","DOIUrl":"10.1016/j.aej.2025.03.134","url":null,"abstract":"<div><div>With the advent of state-of-the-art computer and digital technology, modern civilisation has been immensely facilitated and optimised. The Internet of Things (IoT) has grown in importance in recent years, allowing us to monitor our physical environments and broadening our horizons. The \"practice, science, or art\" of farming is defined as tending to land, growing crops with the use of different tools and techniques, and then selling the harvested food. If farmers optimise their operations with the help of a Wireless Sensor Network (WSN), they will be able to work much more efficiently and effectively. Data aggregation involves collecting information from multiple sensors. The data aggregation process is optimised by applying metaheuristic techniques. A Genetic Algorithm (GA) is a method for modelling evolution that uses mutation, crossover, and natural selection as its building blocks. The key benefit of the Artificial Bee Colony (ABC) approach is that it simultaneously considers local and global search, and it doesn't get trapped calculating its local minima. Naturalistic algorithms like ALO model their hunting behaviour after that of ant-lions and doodlebugs. It manages to find a happy medium between exploration and exploitation with just one operator. Experimental evidences show that the proposed metaheuristic technique, ABC-ALO, which combines the best elements of Artificial Bee Colony and Ant Lion Optimisation, is superior to existing metaheuristic approaches in terms of lifetime computation, or the number of alive nodes at different round counts.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"125 ","pages":"Pages 692-699"},"PeriodicalIF":6.2,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873522","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Classification of human-written and AI-generated sentences using a hybrid CNN-GRU model optimized by the spotted hyena algorithm","authors":"Mahmoud Ragab ,&nbsp;Ehab Bahaudien Ashary ,&nbsp;Faris Kateb ,&nbsp;Abeer Hakeem ,&nbsp;Rayan Mosli ,&nbsp;Nasser N. Albogami ,&nbsp;Sameer Nooh","doi":"10.1016/j.aej.2025.04.071","DOIUrl":"10.1016/j.aej.2025.04.071","url":null,"abstract":"<div><div>The rapid advancement of artificial intelligence (AI) in generating human-like text poses significant challenges in distinguishing between human-written and AI-generated content. Recent advancements in natural language generation have significantly enhanced the quality and variety of AI-generated text, making it almost indistinguishable from human-written content. ChatGPT, a popular AI model, belongs to the generative pre-trained transformer family. While human content is created with a clear intent to convey meaning, AI-generated text aims to replicate human-like language. Classifying human-written and AI-generated sentences is crucial for addressing issues like fake news, plagiarism, and spamming. AI text often follows repetitive patterns, while human writing is more creative and original, making detection significant for combating misinformation. Therefore, this study proposes to classify human-written and AI-generated sentences using a hybrid CNN-GRU model optimized by the Spotted Hyena Algorithm (CHWAIG-DLSHO) approach. The approach involves preprocessing text data through tokenization, lemmatization, and data splitting, followed by word embedding using Latent Dirichlet Allocation (LDA). A hybrid convolutional neural network (CNN) and gated recurrent unit (GRU) model is employed for sentence classification. The spotted hyena optimizer (SHO) model is utilized to fine-tune the hyperparameters of the CNN-GRU model, enhancing its performance. The analysis of the CHWAIG-DLSHO method takes place utilizing AI vs. human text dataset. The performance validation of the CHWAIG-DLSHO method portrayed a superior accuracy value of 99.17 % over existing techniques.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 116-130"},"PeriodicalIF":6.2,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873309","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Neural networks and genetic algorithms-based self-adjustment system for a backstepping controller of an unmanned aerial vehicle","authors":"Omar Rodríguez-Abreo ,&nbsp;Marcos Aviles ,&nbsp;Juvenal Rodríguez-Reséndiz ,&nbsp;A. García-Cerezo","doi":"10.1016/j.aej.2025.04.034","DOIUrl":"10.1016/j.aej.2025.04.034","url":null,"abstract":"<div><div>Backstepping control has been widely used in drones because it considers the dynamic of the system when designing the control law and is robust to parametric uncertainties. However, the typical controller has twelve gains that must be adjusted for optimal results. This process is done manually and with a fixed value, which limits the performance of the controller. This article presents a backstepping intelligent self-tuning system for a multirotor drone. The autotuning is done based on the dynamic vehicle response, optimizing energy consumption, and minimizing its rise time, but without causing an overshoot that consumes unnecessary energy. A backpropagation neural network was trained with a database that considers the dynamic response of the system to achieve this effect. The database was obtained with a metaheuristic algorithm to ensure that only combinations that meet these conditions are used. Several independent tests were carried out to test the system. The results show that the proposed method is adequately adjusted and fulfilled, with the expected dynamic response for 95% of the tests and a dynamic response with minor overshoot and settling time, compared to a PID tuned by genetic algorithm.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 70-80"},"PeriodicalIF":6.2,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873296","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An efficient gesture recognition for HCI based on semantics-guided GCNs and adaptive regularization graphs","authors":"Liheng Dong ,&nbsp;Xin Xu ,&nbsp;Guiqing He ,&nbsp;Yuelei Xu ,&nbsp;Jarhinbek Rasol ,&nbsp;Chengyang Tao ,&nbsp;Zhaoxiang Zhang","doi":"10.1016/j.aej.2025.04.019","DOIUrl":"10.1016/j.aej.2025.04.019","url":null,"abstract":"<div><div>In the embedded system, real-time gesture recognition is crucial to human–computer interaction (HCI). Recently, Graph Convolutional Networks (GCNs) have been applied to inertial measurement unit-based (IMU-based) gesture recognition. However, the disadvantage of these GCN-based methods is that they use very deep networks to capture deep motion features, without considering computational efficiency. In this paper, we propose a shallow GCN as the basic framework to ensure the real-time performance of gesture recognition. To solve the problem of shallow networks’ difficulty capturing deep motion features, we provide hand-crafted semantic information about the positions of nodes (sensors) and frames to guide deep feature extraction. Furthermore, we propose a regularization module named Double-Mask (2MASK) to enhance the network’s generalization. Experiments show that the average inference time on raspberry pi 4b is less than 4 ms. Extensive testing on the self-constructed dataset indicates that the proposed method outperforms previous state-of-the-art (SOTA) methods on multiple metrics. The accuracy reaches 89.47% and 98.70% on two public datasets, outperforming other methods. Experiments in an HCI application show that our method meets the high-precision and low-latency requirements for autonomous taxiing of UAVs. The code for this paper has been uploaded to <span><span>https://github.com/oldbowls/2MAGCN-FN</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 30-44"},"PeriodicalIF":6.2,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143869613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decentralized multi-agent federated and reinforcement learning for smart water management and disaster response","authors":"H. Mancy ,&nbsp;Naglaa E. Ghannam ,&nbsp;Amr Abozeid ,&nbsp;Ahmed I. Taloba","doi":"10.1016/j.aej.2025.04.033","DOIUrl":"10.1016/j.aej.2025.04.033","url":null,"abstract":"<div><div>Water resource management and disaster response have become some of the most challenging tasks, especially when disasters pose a threat, as delays could lead to more impacts. The centralized system used for water dynamics and disaster control usually presents itself as a scalability problem since more clients present a problem, the system's latency is high, and the system is always prone to a single-point failure. The previous approach lacks flexibility and does not synchronously guarantee the integration of several subjects in real time, especially during unpredictable disaster conditions. The proposed FL-MAPPO model surpasses current methods by facilitating decentralized, privacy-protecting decision-making minimizing latency and single-point failures. In contrast to LSTM, Bi-LSTM, and DRNN, which are based on centralized data processing, FL-MAPPO provides real-time adaptability and effective resource management. Experimental results validate that it has lower MSE, higher R² scores, and quicker response times, making it better suited for flood prediction and disaster response. To this end, this study advances a solution through a Decentralized Learning-Driven Multi-Agent Autonomous System (DL-MAAS). The new feature is a Decentralized Cooperation environment in which intelligent and self-managing agents learn utilizing Reinforcement Learning (RL) and Federated Learning (FL) algorithms for enhancing smart water management and real-time disaster relief. IoT devices are adopted for sensing and data acquisition, adaptive learning for decision-making, and optimization of energy use among the agents in the system through metaheuristic algorithms. The research methodology for implementing the proposed solution involves the design of a multi-layered architecture, including data acquisition, decentralized learning, and real-time execution. With a Mean Squared Error (MSE) of 0.112, R-squared (R²) of 0.953, and Mean Absolute Error (MAE) of 0.207, the proposed method is better than existing approaches for big, real-time flood predictive systems. Data show that decentralized systems provide orders of magnitude higher efficiency in water distribution, time of response to disasters, and energy usage compared to conventional centralized systems. These results indicate the significant opportunity for decentralized multi-agent systems in the sustainability of disaster management and water resources.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 8-29"},"PeriodicalIF":6.2,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870238","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Numerical investigation of a fractal oscillator arising from the microbeams-based microelectromechanical system","authors":"Bin Chen ,&nbsp;Junfeng Lu ,&nbsp;Lei Chen","doi":"10.1016/j.aej.2025.04.015","DOIUrl":"10.1016/j.aej.2025.04.015","url":null,"abstract":"<div><div>In this paper, we consider a electrically excited microbeams-based microelectromechanical system (MEMS) on a fractal time space. This MEMS problem can be modelled by a fractal nonlinear oscillator. A numerical approach by combining the fractal complex transformation and the spreading residue harmonic balance method is proposed for finding the approximations to the fractal vibration system. The approximated solutions and frequencies with high accuracy are given, and compared with the approximations by the existing methods such as Runge–Kutta method, energy balance method and Li-He’s modified homotopy perturbation method. Sensitivity analysis of the approximations concerning different amplitudes and other parameters is also investigated for understanding the numerical behaviour. Numerical results confirm the efficiency of the proposed approach over some existing methods.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 53-59"},"PeriodicalIF":6.2,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870195","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Low-light object detection via adaptive enhancement and dynamic feature fusion","authors":"Siyong Fu,&nbsp;Qinghua Zhao,&nbsp;Hesheng Liu,&nbsp;Qiuxiang Tao,&nbsp;Danjuan Liu","doi":"10.1016/j.aej.2025.04.047","DOIUrl":"10.1016/j.aej.2025.04.047","url":null,"abstract":"<div><div>Under low-light conditions, object detection tasks face challenges such as low brightness, low contrast, and noise, which can lead to missed or incorrect detections. To address this issue, this paper proposes a low-light enhancement algorithm, called DAMFCN, and an improved DarkYOLOv8 method, aimed at enhancing low-light image quality and object detection performance. DAMFCN significantly improves the quality of low-light images by integrating the Low-Light Adaptive Module and the Multi-Scale Feature Compensation Block, where LLAM effectively extracts fine details and suppresses noise, and MSFCB compensates for lost details by integrating multi-scale information. The DarkYOLOv8 framework, built on the EfficientNet backbone, combines a multi-scale attention mechanism and the Dynamic Feature Fusion Attention Module, demonstrating superior object detection performance under low-light conditions. Experimental results show that the proposed methods outperform existing state-of-the-art techniques in terms of accuracy, robustness, and efficiency, offering broad application potential.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 60-69"},"PeriodicalIF":6.2,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Lumps, solitons, modulation instability and stability analysis for the novel generalized (2+1)-dimensional nonlinear model arising in shallow water","authors":"Kalim U. Tariq ,&nbsp;Adil Jhangeer ,&nbsp;Muhammad Nasir Ali ,&nbsp;Hamza Ilyas ,&nbsp;R. Nadir Tufail","doi":"10.1016/j.aej.2025.03.110","DOIUrl":"10.1016/j.aej.2025.03.110","url":null,"abstract":"<div><div>In this study, the (2+1)-dimensional Kadomtsev–Petviashvili type equation is investigated that describes the nonlinear wave patterns of behavior and properties in oceanography, fluid dynamics, and shallow water. Firstly, the Hirota bilinear form is implemented to develop a variety of lump, strip soliton and periodic waves solutions for the governing model. Furthermore, some interesting traveling and semi-analytical solitons are generated by availing the extended modified auxiliary equation mapping technique and the Adomian decomposition algorithm. Moreover, in order to determine the absolute error, we have constructed a juxtapose of approximate and soliton results. Additionally, we deliberate the stability analysis and the modulation instability for the governing model extensively to validate the scientific computations. Moreover, the graphical portrayals which include contour plots, 2D and 3D models are illustrated that are useful for understanding the behaviors and dynamics presented by the model’s solutions. The findings of current study are quite novel and make a big contribution to soliton dynamics and mathematical physics.</div></div>","PeriodicalId":7484,"journal":{"name":"alexandria engineering journal","volume":"126 ","pages":"Pages 45-52"},"PeriodicalIF":6.2,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143869614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}