{"title":"A robust IoT architecture for smart inverters in microgrids using hybrid deep learning and signal processing against adversarial attacks","authors":"Mahmoud Elsisi , Shimaa Bergies","doi":"10.1016/j.iot.2025.101576","DOIUrl":"10.1016/j.iot.2025.101576","url":null,"abstract":"<div><div>The increasing autonomy and deployment of cyber-physical systems, particularly power electronics-based inverters within microgrids, has heightened their vulnerability to cyber threats, such as False Data Injection (FDI) and adversarial attacks, which can compromise the integrity of data exchanged across communication networks. To address these security concerns, this paper proposes a new Internet of Things (IoT) architecture that integrates a hybrid approach combining 2-D Convolutional Neural Networks (2-D CNN) with Continuous Wavelet Transform (CWT) for enhanced cyberattack detection. The framework is designed to detect and mitigate adversarial perturbations, focusing on FDI and other attack vectors targeting the communication infrastructure of smart inverters. By transforming raw data into images using CWT, the framework enables efficient statistical feature extraction, enhancing learning accuracy to approximately 98.9 %, outperforming other models. Additionally, it reduces the computational load of signal processing, achieving a processing time of just 0.0548 s. The proposed deep learning model is tested against various levels of cyber perturbations, and its performance is benchmarked against other deep learning and machine learning techniques. The framework is validated using real-time data from a practical distribution system equipped with smart inverters, demonstrating its effectiveness in safeguarding microgrids from cyber threats.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101576"},"PeriodicalIF":6.0,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143697480","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}
Alberto Ferrero-López, Antonio Javier Gallego, Miguel Angel Lozano
{"title":"Bluetooth low energy indoor positioning: A fingerprinting neural network approach","authors":"Alberto Ferrero-López, Antonio Javier Gallego, Miguel Angel Lozano","doi":"10.1016/j.iot.2025.101565","DOIUrl":"10.1016/j.iot.2025.101565","url":null,"abstract":"<div><div>This study explores the application of neural networks in indoor positioning using BLE (Bluetooth Low Energy) and the Fingerprinting location technique. The methodology involves two main phases: the capture and filtering process, where received BLE signals are smoothed and combined into fingerprint vectors, and the subsequent location prediction phase, which compares the position estimation from eight neural network designs and the classical trilateration method. We conduct a performance comparative analysis of each prediction method and study the optimal parameter values for the capturing and filtering processes. The research underscores the limitations of training metrics in reflecting real-world performance, emphasizing the importance of testing models on actual trajectories. Results indicate that regression neural networks outperform classification ones, and a complex dense neural network model proves most versatile and stable across testing scenarios. Our approach achieves a mean error of 1.9 meters, surpassing existing accuracies of 3.7 meters for trilateration and 3.1 meters for state-of-the-art neural network designs, thus holding promise for significantly improving indoor positioning accuracy with practical implications across various domains.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101565"},"PeriodicalIF":6.0,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143628386","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":"Secure UAV routing with Gannet Optimization and Shepard Networks","authors":"R Yuvaraj , Velliangiri Sarveshwaran","doi":"10.1016/j.iot.2025.101575","DOIUrl":"10.1016/j.iot.2025.101575","url":null,"abstract":"<div><div>In recent times, Unmanned Aerial Vehicle (UAV) networks have been extensively employed in civilian and military scenarios. However, they are also highly susceptible to threats from adversaries owing to its distributed nature. To ensure reliable and secure functioning of smaller drones, designing a robust network architecture and applying tailored privacy as well as security mechanisms is important. This research presents a Gannet Weaving Optimization Algorithm based Adversarial Shepard Convolutional Spinal Network (GWOA+Adversarial ShCSpinalNet) for efficient routing and malicious detection in UAV. Initially, the UAV network is simulated, and then, routing is accomplished utilizing the Gannet Weaving Optimization Algorithm (GWOA) by considering the multi-objectives. The GWAO is designed by incorporating Gannet Optimization Algorithm (GOA) with Carpet Weaving Optimization (CWO). Here, energy prediction is accomplished by a Dilated Residual Network (DRN). Thereafter, data communication is performed by monitoring agents. Then, malicious detection is carried out employing Adversarial ShCSpinalNet by a decision-making agent, wherein packet delivery, round trip time, signal strength count of incoming packets and size of packet are considered as attributes. Moreover, Adversarial ShCSpinalNet is introduced by combining Shepard Convolutional Neural Network (ShCNN) and SpinalNet with an adversarial loss function. Thereafter, attack mitigation is conducted by a defensive agent. The GWOA+Adversarial ShCSpinalNet attained a maximal detection rate of 94.827 %, energy of 44.755J and Packet Delivery Ratio (PDR) of 76.446 % as well as a minimal delay of 0.553ms.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101575"},"PeriodicalIF":6.0,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143628385","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}
Mohamed Elgamal , Abdelfattah A. Eladl , Bishoy E. Sedhom , Ahmed N. Sheta , Ahmed Refaat , A. Abdel Menaem
{"title":"A Deep Learning-Based Cyberattack Detection Method for Transmission Line Differential Relays","authors":"Mohamed Elgamal , Abdelfattah A. Eladl , Bishoy E. Sedhom , Ahmed N. Sheta , Ahmed Refaat , A. Abdel Menaem","doi":"10.1016/j.iot.2025.101574","DOIUrl":"10.1016/j.iot.2025.101574","url":null,"abstract":"<div><div>Cyberattacks on power systems have increased, posing serious threats to control systems and protective relays. Line differential relays (LDRs) are widely used to protect critical transmission lines due to their fast, selective, and sensitive operation. However, despite these advantages, LDRs remain vulnerable to cyberattacks as they rely on communications to exchange measurements, which can be compromised. This paper proposes a new deep learning-based cyberattack detection method to detect false-tripping and missed-tripping/fault-masking cyberattacks targeting LDRs. The proposed scheme relies solely on LDR's local measurements, enhancing its security compared to previous solutions, as local measurements are more difficult for hackers to manipulate. The proposed method is based on a deep learning neural network (DLNN), providing a robust model to protect LDRs from cyberthreats. The DLNN model is trained offline on a wide multi-state dataset that includes possible conditions of normal operation, internal faults, and nearby external faults. Additionally, the hyperparameters of the DLNN model are optimized using Bayesian optimization. To reduce complexity, a rule-based system is integrated to identify the type of potential cyberattack instead of incorporating all cyberattack scenarios into the DLNN training phase as done in previous studies. The performance of the proposed method is evaluated under various scenarios, including normal operation, faults, and cyberattacks. The results demonstrate the superiority and efficacy of the proposed scheme in detecting cyberattacks. The proposed scheme outperforms recent literature by achieving nearly 100% classification accuracy on the test dataset. Even under the worst-case scenario of measurement noise, the classification accuracy drops slightly to 99.3667%.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101574"},"PeriodicalIF":6.0,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143628384","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":"AI-driven insights into B5G/6G MAC mechanisms: A comprehensive analysis","authors":"Djamila Talbi, Zoltan Gal","doi":"10.1016/j.iot.2025.101571","DOIUrl":"10.1016/j.iot.2025.101571","url":null,"abstract":"<div><div>In the 6G wireless communication domain, optimizing the medium access control mechanism is crucial for enhancing the performance of high-speed, over Terabit/sec transmission rate networks. This paper evaluates the adaptive directional antenna protocol for terahertz frequencies technology using the ns-3 simulator, employing different techniques like Shannon entropy, wavelet transform, supervised, unsupervised machine learning, and classical processing methods focusing on the impact of the two used parameters: overlapping ratio and the rotation step. Our approach is to highlight the importance of the <em>Paleo-AI Classical Processing</em> method, which is about incorporating the classical mathematical processing method, with the AI models for better results. The proposed method includes applying some analytical tools like the entropy metrics to understand the dynamic behavior of the radio control frames, particularly in distinguishing between the stable and unstable phases of the communication process and includes the adoption of fractal wavelet analysis for better learning. Additionally, the RNN classification of MAC event sequences into categories supported by transfer learning enhanced the model's efficiency, where we introduced the weighted accuracy to time ratio, a novel approach to assess the competency of various deep learning models. Moreover, different generative AI methods were used to produce synthetic data where the similarity levels were quantified by using six distinct metrics. The overall results of this paper demonstrated the necessity of adapting the MAC protocols to specific environmental conditions, thereby contributing to the development of more resilient B5G/6G communication networks.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101571"},"PeriodicalIF":6.0,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143643269","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sungpil Woo , Muhammad Zubair , Sunhwan Lim , Daeyoung Kim
{"title":"Deep multimodal emotion recognition using modality-aware attention and proxy-based multimodal loss","authors":"Sungpil Woo , Muhammad Zubair , Sunhwan Lim , Daeyoung Kim","doi":"10.1016/j.iot.2025.101562","DOIUrl":"10.1016/j.iot.2025.101562","url":null,"abstract":"<div><div>Emotion recognition based on physiological signals has garnered significant attention across various fields, including affective computing, health, virtual reality, robotics, and content rating. Recent advancements in technology have led to the development of multi-modal bio-sensing systems that enhanced the data collection efficiency by simultaneously recording and tracking multiple bio-signals. However, integrating multiple physiological signals for emotion recognition presents significant challenges due to the fusion of diverse data types. Differences in signal characteristics and noise levels significantly deteriorate the classification performance of a multi-modal system and therefore require effective feature extraction and fusion techniques to combine the most informative features from each modality without causing feature conflict. To this end, this study introduces a novel multi-modal emotion recognition method that addresses these challenges by leveraging electroencephalogram and electrocardiogram data to classify different levels of arousal and valence. The proposed deep multimodal architecture exploits a novel modality-aware attention mechanism to highlight mutually important and emotion-specific features. Additionally, a novel proxy-based multimodal loss function is employed for supervision during training to ensure the constructive contribution of each modality while preserving their unique characteristics. By addressing the critical issues of multi-modal signal fusion and emotion-specific feature extraction, the proposed multimodal architecture learns a constructive and complementary representation of multiple physiological signals and thus significantly improves the performance of emotion recognition systems. Through a series of experiments and visualizations conducted on the AMIGOS dataset, we demonstrate the efficacy of our proposed methodology for emotion classification.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101562"},"PeriodicalIF":6.0,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611267","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Artificial intelligence-driven security framework for internet of things-enhanced digital twin networks","authors":"Samuel D. Okegbile , Ishaya P. Gambo","doi":"10.1016/j.iot.2025.101564","DOIUrl":"10.1016/j.iot.2025.101564","url":null,"abstract":"<div><div>In the evolving area of internet of things (IoT)-enabled digital twin networks (DTNs), ensuring robust security and data privacy is a necessity. This paper presents an AI-driven security framework designed to address security and privacy requirements in DTNs by integrating advanced machine learning techniques. We propose a novel approach that combines long short-term memory (LSTM) networks with transfer learning and differential privacy (DP) to enhance threat detection and preserve sensitive data. The LSTM networks are employed to model sequential data patterns, crucial for identifying and mitigating security threats in such dynamic environments. In addition, transfer learning is utilized to leverage pre-trained models, improving accuracy and reducing training time while DP is incorporated to protect user privacy by introducing the Gaussian noise into the training process, thereby ensuring confidential data handling. We formulate the proposed AI-driven security solution as a multi-layer framework and investigate its ability to achieve significant improvements, in terms of detection accuracy and privacy preservation, compared to conventional methods. We then obtain simulation results to demonstrate the effectiveness of the solution in adapting to evolving threats while maintaining high-performance standards. It is believed that the proposed solution will open new research directions towards improving security in emerging cyber–physical systems such as DTNs.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101564"},"PeriodicalIF":6.0,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143611266","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":"Optimizing cutting parameters with Industrial IoT system for automated continuous dry milling","authors":"Chin-Shan Chen, Shao-Chien Hsu","doi":"10.1016/j.iot.2025.101569","DOIUrl":"10.1016/j.iot.2025.101569","url":null,"abstract":"<div><div>Applying the combination of the Industrial Internet of Things (IIoT) and noise technology to automated milling operation systems, this study aims to understand the relationship between cutting noise, vibration, the surface roughness of the workpieces, and tool wear in the automatic processing process. Taguchi method is first used for acquiring the optimal cutting parameters; IIoT, accelerometer, and noise meter are integrated into automated cutting systems, under different cutting parameters, to acquire vibration and noise data in the cutting process; and the surface roughness of workpiece and end mill wear is measured in the experimental process. Finally, the data analysis results explain the causation among cutting noise, vibration, surface roughness of the workpiece, and tool wear. It is proved by experiments that indicate a strong linear and positive correlation between vibration and noise generated during continuous milling that a single noise meter can replace the use of three vibration sensors, and a cutting noise value of 82.812db indicates the best timing to replace the tool in this case. Such results could provide optimized tool change timing and reduce processing costs, and they are expected to offer valuable opinions to automated cutting processing.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101569"},"PeriodicalIF":6.0,"publicationDate":"2025-03-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143636757","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"IoT wearables in child health: A comprehensive scoping review and exploration of ubiquitous computing","authors":"Kajal Mistry, Georgios Dafoulas","doi":"10.1016/j.iot.2025.101556","DOIUrl":"10.1016/j.iot.2025.101556","url":null,"abstract":"<div><div>The increasing adoption of health technologies, especially the Internet of Things (IoT), for monitoring children's independent mobility is a growing trend. Despite this, concerns from both children and parents persist, particularly regarding the tracking of health vitals and the child's whereabouts. The integration of these technologies introduces a unique challenge, considering the diverse landscape of available options for monitoring and tracking a child's location. In this context, this review aims to provide an up-to-date analysis of emerging concerns and novel technologies related to vital signs and location tracking, exploring the implications of ubiquitous computing in the aspect of child health monitoring. By addressing these concerns, the research seeks to contribute to a better understanding of the practical considerations and potential solutions in the adoption of health technologies for monitoring and tracking children's well-being.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101556"},"PeriodicalIF":6.0,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143628387","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Blockchain based handle system to secure Predictive maintenance analysis system in Industrial IoT using L2S–GRU","authors":"Mahamat Ali Hisseine , Deji Chen , Xiao Yang","doi":"10.1016/j.iot.2025.101549","DOIUrl":"10.1016/j.iot.2025.101549","url":null,"abstract":"<div><div>The handle system that assigns persistent identifiers to the information resources is utilized by the Predictive Maintenance analysis. The persistent identifiers are referred to as the handles that assist as a unique and enduring reference for locating, accessing, and utilizing the resources. Nevertheless, Industrial Internet of Things devices were affected by insider attacks in most of the prevailing works. Thus, an effective Blockchain-enabled secure handle system using Adversarial Stylometry-KAnonymity and <span><math><mrow><mi>L</mi><mi>o</mi><msup><mrow><mi>g</mi></mrow><mrow><mn>2</mn></mrow></msup></mrow></math></span> Sigmoid–Gated Recurrent Unit is presented in this paper. It is proposed to enhance the Handle System’s scalability and support real-time data processing. Primarily, the Industrial Internet of Things device and user are registered with the Blockchain. Then, based on Adversarial Stylometry-KAnonymity, the details are privacy preserved. Meanwhile, for the Industrial Internet of Things devices and users, the keys, Quick Response code, and hashcode are generated. Later, the Industrial Internet of Things device setup grounded on Message Authentication Code creation and data sensing is conducted. After that, by using Triangular and S-shaped-fuzzy, a controlled access-based smart contract is created. The Fuzzy system is used to enhance the smart contract by enabling flexible and dynamic decision-making in the context of undefined or inaccurate data. Extensive experiments and analysis proved the effectiveness of the proposed framework for predictive maintenance in Industrial Internet of Things. The outcomes proved that a higher accuracy of 99.12%, Precision of 99.24% and of Recall 99.33% was attained by the proposed model, thereby outperforming similar existing models.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"31 ","pages":"Article 101549"},"PeriodicalIF":6.0,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143593307","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}