Internet of Things最新文献

{"title":"Fault detection of industrial air separation stations based on metaheuristic optimization and bidirectional long short-term memory technique","authors":"Thanh-Phuong Nguyen ,&nbsp;Chung-Chieh Lin ,&nbsp;Ming-Yuan Cho","doi":"10.1016/j.iot.2025.101850","DOIUrl":"10.1016/j.iot.2025.101850","url":null,"abstract":"<div><div>Industrial air separation stations play a critical role in numerous sectors, necessitating robust fault detection frameworks to ensure reliability and operational safety. This study presents a novel approach combining Bidirectional Long Short-Term Memory (BLSTM) networks with Enhanced Particle Swarm Optimization (EPSO) for fault detection in industrial air separation stations. The BLSTM model, renowned for its ability to capture temporal dependencies in sequential data, is optimized using EPSO to fine-tune its hyperparameters, enhancing its fault detection performance. The proposed EPSO-BLSTM framework is rigorously evaluated against conventional techniques, including Recurrent Neural Networks (RNN), Bidirectional RNN, Gated Recurrent Units (GRU), LSTM, Convolutional Neural Networks (CNN), and standard BLSTM model with the most notable improvements of error-based 58.32 % Loss, 63.37 % Val Loss, 1.24 % CP, 0.6 % Val CP, 60.11 % MAE, 63.37 % Val MAE, 63.46 % MSE, and 81.75 % Val MSE, and with accurate-based 4.68 % Pre, 5.55 % Val Pre, 5.72 % Rec, and 4.39 % Val Rec. Comparative analysis highlights the superior fault classification accuracy and generalization capability of the EPSO-BLSTM model under diverse operational scenarios. This research underscores the potential of integrating metaheuristic optimization with advanced deep learning architectures to address complex fault detection challenges, offering a scalable and efficient solution for industrial air separation stations.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101850"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738841","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":"Group-based link modeling for wireless digital twins: Towards accurate network performance prediction","authors":"Samir Si-Mohammed ,&nbsp;Fabrice Théoleyre","doi":"10.1016/j.iot.2026.101875","DOIUrl":"10.1016/j.iot.2026.101875","url":null,"abstract":"<div><div>Wireless networks are increasingly deployed in diverse domains, from best-effort object tracking to real-time control in smart factories. Yet, their performance strongly depends on configuration choices, especially at the MAC level. Thus, a single homogeneous configuration is often suboptimal due to the heterogeneous nature of individual links. We argue that Digital Twins (DTs) are a promising enabler for autonomous networks, capable of adapting configurations dynamically to prevailing conditions. However, global modeling approaches in DTs make it difficult to capture link-level variability and to accurately model the impact of configuration changes on performance. In this work, we propose a link-oriented prediction model designed to serve as a cornerstone for future wireless Digital Twins. Our model estimates the Packet Reception Rate (Packet Reception Rate (PRR)) under different MAC configurations, capturing the unique characteristics of each communication link. To improve scalability in large deployments, we explore a clustering-based approach, where predictive models are trained per group of similar links rather than per individual link. Our experimental evaluation shows that these data-driven methods effectively capture link heterogeneity while offering robust prediction accuracy and enhanced generalization capabilities.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101875"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146077992","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":"PT-TrafficAnalyzer: A weighted ensemble prediction tree for IoT attack detection","authors":"Recep Sinan Arslan","doi":"10.1016/j.iot.2026.101874","DOIUrl":"10.1016/j.iot.2026.101874","url":null,"abstract":"<div><div>The Internet of Things has a network structure that is vulnerable to cyberattacks and susceptible to attackers. Ensuring the privacy of organizations and individuals is a crucial issue in IoT networks, where sensitive data is transmitted and various types of attacks are prevalent. The importance of Intrusion Detection Systems (IDS) in detecting attacks and infiltration attempts on IoT networks is increasing daily. In this way, it will contribute to the resistance against attackers and the spread of this modern technology. In this study, Prediction Tree Traffic Analyzer (PT-TAnalyzer), an IDS system capable of detecting and classifying attacks on IoT networks, is proposed. PT-TAnalyzer features an ensemble model structure, where weighting is determined by the validation scores of machine learning models, and a prediction tree comprising eight ensemble models trained on the CIC-IoT-2023 dataset. This proposed model detects 34 attack types (including 33 malicious and one benign) with high success rates, due to its unique attack-detection approach, and does so efficiently and cost-effectively. Unlike traditional studies, it achieves this by using eight trained models rather than classifying all attacks with a single model and a single prediction structure within the tree architecture. In the tests performed, PT-TAnalyzer achieved 99.76 % accuracy in the binary classification experiment (Benign vs. Malicious) and 98.70 % accuracy in the 34-class experiment, yielding a similar F1 Score. The test time per sample is less than 0.1 ms. Compared with previous frameworks using the same dataset, PT-TAnalyzer shows a 2 % improvement in overall accuracy and a lower processing time. In practice, the proposed model can be deployed on IoT gateways or edge devices to provide real-time, low-cost, and scalable intrusion detection capabilities. The model outperforms previous studies using the same dataset, while also addressing the limitations.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101874"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146023230","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":"Analysis of microservices-based IoT systems: deployment challenges, industry practices, and performance insights","authors":"Yahia El Fellah ,&nbsp;Jean Baptiste Minani ,&nbsp;Naouel Moha ,&nbsp;Julien Gascon-Samson ,&nbsp;Yann-Gaël Guéhéneuc","doi":"10.1016/j.iot.2025.101867","DOIUrl":"10.1016/j.iot.2025.101867","url":null,"abstract":"<div><div>As the adoption of microservices in Internet of Things (IoT) systems grows, deploying them on the Edge remains a significant challenge for practitioners. While Edge Computing offers improved latency and resource efficiency by processing data near the source, it also introduces complexity in managing microservices. Despite increasing academic interest, few comprehensive studies have investigated the specific challenges and effective software engineering (SE) practices for deploying microservices-based IoT systems on the Edge. Therefore, we conducted a multi-method study to bridge this gap. We used three methods: (1) a systematic literature review (SLR) to identify known challenges and approaches, (2) a gray literature review (GLR) to extract SE practices used in the field, and (3) an empirical evaluation using two versions of a case study, one with and one without selected SE practices. The findings show that (1) the most reported challenges relate to resource utilization and performance optimization, (2) containerized microservices, API gateways, and database-per-service are among the most commonly recommended practices, and (3) implementing these practices led to a 132% throughput improvement, 49% reduction in latency, and memory savings of up to 13% in Edge-based IoT systems. However, increased architectural complexity also led to higher CPU usage. This study offers a catalog of best practices and empirical evidence to support IoT developers aiming to optimize microservices-based deployments on the Edge, particularly in resource-constrained environments.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101867"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978121","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":"Sustainable, QoS, and cost-aware placement of microservices on the continuum: a use case on the Internet of Medical Things","authors":"Alejandro Moya ,&nbsp;Juan Luis Herrera ,&nbsp;Juan López-Rodenas ,&nbsp;Javier Berrocal ,&nbsp;Juan Manuel Murillo ,&nbsp;Elena Navarro","doi":"10.1016/j.iot.2026.101894","DOIUrl":"10.1016/j.iot.2026.101894","url":null,"abstract":"<div><div>Domains such as healthcare, which are intensive in contrast to user-grade domains, are increasingly interested to develop Internet of Things (IoT) applications to automate their critical processes, leading to the Internet of Medical Things (IoMT). IoMT applications leverage the Computing Continuum infrastructure. However, for intensive IoT and IoMT applications to be feasible, their strict Quality of Service (QoS) requirements must be met, and the economic cost of their deployment must be low to ensure their business viability. Furthermore, sustainability and the reduction of the carbon footprint are currently a priority due to industrial and governmental initiatives and requirements, such as Green IoT or the Sustainable Development Goals. The efforts in sustainability are aimed not only at achieving energy-efficient applications, but carbon-aware ones, aligning carbon footprint reduction with high QoS and low economic cost. While all three objectives can be achieved by strategically placing the microservices of these IoT applications, navigating the trade-offs across the three is a complex issue, calling for automated solutions that provide IoT application developers with a manageable number of Pareto-optimal microservice placements. This work presents the Many-Objective Genetic Algorithm for Microservice Placement (MOGAMP), which leverages evolutionary computing to assist IoT application developers in navigating the QoS, cost, and sustainability trade-off in microservice placement. In an evaluation with an IoMT use case, MOGAMP is shown to be scalable, up to 459.82 times faster and with a memory footprint of up to 0.37% compared to alternatives, enabling IoT application developers to explore wide, yet manageable, Pareto fronts.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101894"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173164","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":"Intelligent drone pickup scheduling via deep reinforcement learning in low altitude economy networks","authors":"K M Rumman,&nbsp;Dimitrios Kaltsogiannis,&nbsp;Eirini Eleni Tsiropoulou","doi":"10.1016/j.iot.2026.101882","DOIUrl":"10.1016/j.iot.2026.101882","url":null,"abstract":"<div><div>The advent of Low Altitude Economy (LAE) has attracted the interest of the research community to analyze the autonomous drone operations, as they become the enablers for on-demand services. In this paper, we introduce an intelligent drone pickup scheduling framework to support LAE networks based on a Deep Reinforcement Learning (DRL) approach that exploits the Maskable Proximal Policy Optimization (M-PPO). A detailed aerodynamic energy model is introduced to capture the nonlinear power requirements related to the takeoff and landing phases under varying payload masses. An aging-aware mechanism is proposed to deal with the service fairness and responsiveness by prioritizing the parcels based on their waiting time, and ultimately mitigating their backlog within the service field. A feasibility masking technique is designed to support the M-PPO framework in order to guarantee the operational constraints related to the drone’s payload capacity and its battery limitations, thus, ultimately preventing infeasible routing decisions during the policy training. A multi-objective reward function is formulated accounting for the drone’s energy efficiency optimization, backlog reduction, and service urgency and allowing the drone to adaptively balance the trade-offs among these factors. Detailed results demonstrate the superiority of the M-PPO framework in terms of its adaptability, fairness, and energy efficiency compared to traditional heuristics and non-adaptive learning mechanisms.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101882"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173270","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":"Multi-hypervisor-based authorization and DoS attack mitigation framework using LC-WTRNN technique","authors":"Kalyan Gattupalli ,&nbsp;Poovendran Alagarsundaram ,&nbsp;Harikumar Nagarajan ,&nbsp;Venkata Surya Bhavana Harish Gollavilli ,&nbsp;Surendar Rama Sitaraman ,&nbsp;Pushpakumar R","doi":"10.1016/j.iot.2025.101843","DOIUrl":"10.1016/j.iot.2025.101843","url":null,"abstract":"<div><div>Hypervisors allow the management of several Virtual Machines (VMs) on a single device but are highly susceptible to DoS attacks, which deprive resources and disrupt cloud services. Techniques currently in use fail to establish proper authorization between multi-hypervisors, thereby exposing VMs to security threats. To ameliorate this situation, we developed a LeCun Wave Tanh Recurrent Neural Network (LC-WTRNN)-based multi-hypervisor authorization framework integrated with Hamming Code Quantum Cryptography (HC-QC), Kullback-Leibler De-Swinging K-Anonymity (KLDS-KAnonymity), and the Hell Bhatt Tiger Hashing Algorithm (HB-THA). Thereby, the system efficiently detects DoS attacks, secures VM registration, and ensures data integrity. With experimental results on the CICDDoS2019 dataset, it is seen that its method achieves an accuracy of 98.62 %, a recall value of 98.45 %, and a specificity of 98.65 % on average, outperforming traditional RNN, DBN, RBM, and DNN methods by 5.3 %. Additionally, the newly proposed framework contributes to a 56.1 % reduction in the time needed for anonymization while providing 8.5 % better encryption security and 44.5 % less tree generation time against the traditional methods. These results thus validate LC-WTRNN as a scalable and secure solution to mitigating DoS attacks in cloud environments.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101843"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145791863","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":"ASEADOS-SDN-IoT: A novel SDN-IoT network intrusion detection dataset and framework","authors":"Tharindu Lakshan Yasarathna,&nbsp;Nhien-An Le-Khac","doi":"10.1016/j.iot.2026.101891","DOIUrl":"10.1016/j.iot.2026.101891","url":null,"abstract":"<div><div>The convergence of Software-Defined Networking (SDN) and the Internet of Things (IoT) offers a transformative architecture for intelligent, programmable, and scalable network management. However, this integration also exposes new attack surfaces and security challenges that require robust Intrusion Detection Systems (IDS). Progress in IDS research for SDN-IoT environments has been limited by the absence of realistic datasets that capture both IoT traffic diversity and SDN control-plane dynamics. To address this gap, this paper presents ASEADOS-SDN-IoT, a novel, publicly available intrusion detection dataset supported by a fully documented hybrid testbed framework. The proposed framework integrates physical IoT devices and virtual IoT nodes within an ONOS-controlled OpenFlow infrastructure, enabling synchronised collection of packet-flow data and SDN controller telemetry under benign and adversarial conditions. The dataset covers four representative attack categories-Denial-of-Service (DoS), Distributed Denial-of-Service (DDoS), Probe, and Botnet-and captures genuine control-data plane interactions observed during live network operation. ASEADOS-SDN-IoT comprises 457,044 labelled flow instances with 83 statistical features and is publicly released to support reproducible research. The experimental evaluation using machine learning (ML) and deep learning (DL) models demonstrates clear separability between benign and malicious traffic. It confirms the dataset’s suitability for training and benchmarking intrusion detection systems. By combining realistic traffic generation, cross-layer visibility, and reproducible design, ASEADOS-SDN-IoT provides a robust benchmark for advancing secure and adaptive SDN-IoT infrastructures.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101891"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173162","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":"TrustEdge: A quantum-safe, self-healing framework for federated TinyML in critical ICU monitoring","authors":"Umar Hayat Khan ,&nbsp;Affaq Qamar ,&nbsp;Rahim Khan ,&nbsp;Mohamad A. Alawad ,&nbsp;Fahad Alturise ,&nbsp;Maqsood Hayat","doi":"10.1016/j.iot.2025.101855","DOIUrl":"10.1016/j.iot.2025.101855","url":null,"abstract":"<div><div>As adversarial threats to AI-extended medicine increases, there is evidence to suggest that existing IoMT architectures lack autonomous cyber-resilience; the ability of a system to detect (observe and record) and then self-heal from attack events without any human effort. Against this backdrop, we develop TrustEdge: an end-to-end IoMT system which seamlessly integrates four critical functions at the edge namely (i) TinyML-driven IDS-on-device (ii) PQC driven secure federated learning using NIST standard ML-KEM-512 algorithm (iii) lightweight blockchain audit trail and (iv) smart contract based self-healing. To that end and specifically designed for resource-constrained ESP32-class low-resource devices, TrustEdge, adapted to the tailored requirements of non-IID clinical environments, enables secure sub-second detection of both sensor spoofing and model poisoning attacks with 98.2 % accuracy while incurring no more than 1 % false positives; under mistaken diagnosis indications it automatically responds by triggering vetted recovery actions (e.g., node isolation or model rollback) without having to depend on cloud back-ends. The framework was tested on clinician-verified synthetic ICU data spanned in 3 simulated hospitals and within a high-fidelity OMNeT++ network simulation with realistic ICU WiFi, at 54Mbps (bit error rate of <span><math><msup><mn>10</mn><mrow><mo>−</mo><mn>5</mn></mrow></msup></math></span>, co-channel interference). Experimental results show that TrustEdge is 99.4 %(accuracy) of the health anomaly detection with negligible overhead (0.09ms(PQC encryption) and 1.2ms(per transaction of blockchain)). Our co-simulation ensures ethical safety and eliminates adversarial testing in living clinical sites; we provide full reproducibility, sharing code, data and simulation scripts. TrustEdge raises the bar for self-defending, autonomous critical care systems impermeable to present and quantum-age threats without the presence of any false positives in normal operation and is fully consistent with quantum-safe edge native design principles.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101855"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146173157","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":"An enhanced group matching method with transformed fingerprints, similarity filtering, and adaptive selection for indoor localization","authors":"Huang Lin ,&nbsp;Shuo Li ,&nbsp;Wei Peng ,&nbsp;Arthur Peng","doi":"10.1016/j.iot.2025.101866","DOIUrl":"10.1016/j.iot.2025.101866","url":null,"abstract":"<div><div>WiFi fingerprinting-based indoor localization faces persistent challenges from signal instability and environmental interference. To address these issues, this study introduces an enhanced Group Matching Method with Transformed Fingerprints, Similarity Filtering, and Adaptive Selection, (GMM-TSFA), that improves localization accuracy, interpretability, and robustness. GMM-TSFA integrates three core mechanisms: instance-wise transformation to normalize signal variations and strengthen similarity metrics; similarity filtering to dynamically refine the candidate pool based on signal relevance; and adaptive selection to flexibly determine the most informative reference points. Together, these components improve both pattern recognition and spatial precision. Experiments on four benchmark datasets demonstrate the effectiveness of the proposed method, with a mean localization error (MLE) of 7.30 m on UJIIndoorLoc and 77.86 % of errors under 10 m, consistently outperforming baseline methods including the original GMM. Unlike deep learning-based models, GMM-TSFA offers a transparent and adaptable framework that supports explainable decision-making. These advantages make it a practical and scalable solution for real-world indoor localization systems deployed in complex environments.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"36 ","pages":"Article 101866"},"PeriodicalIF":7.6,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978120","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}