Internet of Things最新文献

{"title":"An energy-balanced and load-aware routing algorithm based on molecular diffusion theory for energy harvesting assisted WSN","authors":"Sheng Hao ,&nbsp;Junwei Gao ,&nbsp;Jianqun Cui ,&nbsp;Yinyi Chen ,&nbsp;Xiying Fan ,&nbsp;Zhen Li","doi":"10.1016/j.iot.2025.101691","DOIUrl":"10.1016/j.iot.2025.101691","url":null,"abstract":"<div><div>Energy Harvesting-Wireless Sensor Networks (EH-WSNs) play a crucial role in the development of Green Internet of Things (GIoT). While the energy-harvesting process alleviates the constraints of energy supply in WSNs, most current routing protocols for EH-WSNs inadequately account for the heterogeneity in energy states and traffic loads among sensor nodes, which may impair the energy efficiency and transmission performance of networks. To address the above issues, we utilize molecular diffusion theory to design an energy-balanced and load-aware routing algorithm (EBLARA-MD for short) for EH-WSNs. Initially, we construct a dual EH prediction model based on the clustering Markov chain (MC) method, to accurately forecast the amount of solar and wind power generation. Subsequently, an energy-rank model is established to assess the energy levels of nodes. Building on this, we propose a cross-layer adjustment scheme to avoid energy depletion and wastage. Namely, at the Media Access Control (MAC) layer, the backoff time is optimized dynamically to affect the channel access probability of each node; at the physical layer, the transmission power is determined adaptively by considering the wireless fading property. In addition, we construct a load-aware model to reflect the congestion degree of data buffer. Finally, we leverage molecular diffusion theory to allocate the routing probabilities for suitable paths. Simulation results demonstrate that the proposed routing algorithm achieves superior performance in terms of energy efficiency, end-to-end delay variance, and packet delivery ratio.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101691"},"PeriodicalIF":6.0,"publicationDate":"2025-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144579209","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":"Fast data aware neural architecture search via supernet accelerated evaluation","authors":"Emil Njor ,&nbsp;Colby Banbury ,&nbsp;Xenofon Fafoutis","doi":"10.1016/j.iot.2025.101688","DOIUrl":"10.1016/j.iot.2025.101688","url":null,"abstract":"<div><div>Tiny machine learning (TinyML) promises to revolutionize fields such as healthcare, environmental monitoring, and industrial maintenance by running machine learning models on low-power embedded systems. However, the complex optimizations required for successful TinyML deployment continue to impede its widespread adoption.</div><div>A promising route to simplifying TinyML is through automatic machine learning (AutoML), which can distill elaborate optimization workflows into accessible key decisions. Notably, Hardware Aware Neural Architecture Searches — where a computer searches for an optimal TinyML model based on predictive performance and hardware metrics — have gained significant traction, producing some of today’s most widely used TinyML models.</div><div>TinyML systems operate under extremely tight resource constraints, such as a few kB of memory and an energy consumption in the mW range. In this tight design space, the choice of input data configuration offers an attractive accuracy-latency tradeoff. Achieving truly optimal TinyML systems thus requires jointly tuning both input data and model architecture.</div><div>Despite its importance, this “Data Aware Neural Architecture Search” remains underexplored. To address this gap, we propose a new state-of-the-art Data Aware Neural Architecture Search technique and demonstrate its effectiveness on the novel TinyML “Wake Vision” dataset. Our experiments show that across varying time and hardware constraints, Data Aware Neural Architecture Search consistently discovers superior TinyML systems compared to purely architecture-focused methods, underscoring the critical role of data-aware optimization in advancing TinyML.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101688"},"PeriodicalIF":6.0,"publicationDate":"2025-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144696711","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":"Privacy and security of federated learning in resource-constrained Internet of Things environment: Systematic literature review","authors":"Walla Khalaifat,&nbsp;Wael Elmedany,&nbsp;Haroun Alryalat","doi":"10.1016/j.iot.2025.101679","DOIUrl":"10.1016/j.iot.2025.101679","url":null,"abstract":"<div><div>Federated Learning (FL) has become a suggested method due to the growing adoption of Internet of Things (IoT) devices, tied with the growing need for collaborative learning and data analysis. FL allows for distributed learning, while ensuring the privacy. Nevertheless, the potential for security vulnerabilities and privacy breaches became more and more complex because of the increasing number of devices that are connected. Moreover, FL in Resource-Constrained IoT environments introduces additional challenges due to the nature of these environments, as these environments have devices with limited resources. Additionally, since FL enables collaborative learning across a decentralized IoT environment, ensuring strong security and privacy becomes crucial to protect information and maintain the trust of participants. This research aims to present comprehensive Systematic Literature Review (SLR) of academic articles on security and privacy of Resource-Constrained environments in FL published from 2016 to 2024. The study intends to maximize the knowledge in FL, FL in Resource-Constrained IoT environments, identification of privacy and security concerns, and techniques designed to enhance them in FL.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101679"},"PeriodicalIF":6.0,"publicationDate":"2025-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144631203","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 interpretable deep learning framework for intrusion detection in industrial Internet of Things","authors":"Jawad Ahmad ,&nbsp;Shahid Latif ,&nbsp;Imdad Ullah Khan ,&nbsp;Mohammed S. Alshehri ,&nbsp;Muhammad Shahbaz Khan ,&nbsp;Nada Alasbali ,&nbsp;Weiwei Jiang","doi":"10.1016/j.iot.2025.101681","DOIUrl":"10.1016/j.iot.2025.101681","url":null,"abstract":"<div><div>The Industrial Internet of Things (IIoT) has revolutionized smart industries by optimizing industrial operations and accelerating the decision-making process. However, its inherently distributed architecture presents complex and evolving security threats. Traditional machine learning (ML) and deep learning (DL)-based intrusion detection systems (IDSs) often lack interpretability, which undermines their trustworthiness in critical IIoT environments. To overcome these limitations, we propose XGRU-IDS, an explainable hybrid DL-based IDS that combines the strengths of the Extra Trees Classifier (ETC) for feature selection and Gated Recurrent Units (GRU) for sequential attack detection. The ETC enhances model input quality by identifying the most influential features, while the GRU processes temporal dependencies to detect sophisticated intrusion patterns. Explainability is ensured through SHapley Additive exPlanations (SHAP), which offer class-wise insights via summary plots, feature importance scores, and force plots. XGRU-IDS is evaluated on the multiclass CICIoT2023 dataset, which covers all 34 attack types. It achieves 97.56% accuracy, outperforming recent state-of-the-art DL and explainable IDS approaches. This work demonstrates that high detection accuracy can coexist with transparency, providing a robust and trustworthy IDS solution for resource-constrained IIoT networks.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101681"},"PeriodicalIF":6.0,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144535936","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":"Editorial brief for special issue \"Mass connectivity and/or communication paradigms for the internet of things\"","authors":"Elif Haytaoglu ,&nbsp;Suayb S. Arslan ,&nbsp;Orhan Dagdeviren ,&nbsp;Huseyin Ugur Yildiz ,&nbsp;Yusuf Ozturk","doi":"10.1016/j.iot.2025.101625","DOIUrl":"10.1016/j.iot.2025.101625","url":null,"abstract":"<div><div>The Internet of Things (IoT) refers to the collective network of connected devices and technologies facilitating the communication between these devices themselves and between the devices and the cloud at a mass scale. These types of networks can be exploited in many applications that can have timeless importance; for instance, they can be utilized in detecting and preventing both natural and human-made disasters such as earthquakes, fires, floods, etc. Due to ongoing global warming and human-caused pollution, these types of disasters will continue to occur. Therefore, it’s crucial to develop IoT systems that can prevent them as a whole or lessen their long-term impact to the most part. While in operation, the failure of any node or communication link can lead to data loss and network connectivity disruption, resulting in significant burdens. Therefore, maintaining network connectivity and effectively recovering lost information using network resources remain as an elusive open problem.</div><div>Connected smart devices collect a massive amount of information and continuously transmit it for storage and subsequent analysis. Such operation style leads to a heavy load on network communication links, connectivity, and computational resources and subsequently increases the demand for required storage space. As a result, fundamental techniques need to be devised to reduce the demand for such resources. For instance, to reduce the infrastructure cost of the topology and increase fault tolerance, recent research focused on minimizing the number of devices guaranteeing maximal area coverage for a given network topology. Besides reducing the cost of the deployed infrastructure, minimization of the number of devices takes the communication cost into consideration typically expressed in terms of consumed bandwidth Preprint submitted to Elsevier IoT Journal January 24, 2025 or the time elapsed during the device-to-device transfer as the parameters of interest. Accordingly, the focus on reducing such communication overhead using deterministic or/with heuristic solutions is vital for optimal IoT ecosystem design.</div><div>In this special issue, the focus has been on the new and broader technical problems which are related to the connectivity, communication costs, resource sharing and providing resilience for IoT core networks, devices, and applications.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"32 ","pages":"Article 101625"},"PeriodicalIF":6.0,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144634475","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":"Novel deep learning-based IoT network attack detection using magnet loss optimization","authors":"Chi Duc Luu ,&nbsp;Viet Hung Nguyen ,&nbsp;Van Quan Nguyen ,&nbsp;Ngoc-Son Vu","doi":"10.1016/j.iot.2025.101680","DOIUrl":"10.1016/j.iot.2025.101680","url":null,"abstract":"<div><div>The increasing prevalence of Internet of Things (IoT) devices across various industries has raised critical security concerns due to their inherent vulnerabilities and high interconnectivity. While traditional security mechanisms have shown limitations in effectively securing large IoT networks, machine learning (ML) and deep learning (DL) methods have been explored to tackle the attack detection problem in this domain. However, existing approaches still lack optimal regularization and have limited comprehensiveness in validation across different IoT-centric datasets. To address these challenges, this research proposes the extension of the Deep Magnet Autoencoder (DMAE) and introduces a novel approach, the Cascade Deep Magnet Autoencoder (CDMAE), leveraging the Magnet Loss optimization as regularization for better class distinction through local separation in latent space. This enhanced class clustering strengthens attack detection by maximizing inter-class separation while compactly grouping data points of the same class, leading to more precise identification of benign and malicious traffic. Extensive experiments conducted on three contemporary IoT datasets, CIC-BoT–IoT, CIC-ToN–IoT, and CICIoT2023, demonstrate that our proposed models are able to produce meaningful latent representations with powerful discrimination between benign and malicious IoT network data. Empirical insights for fine-tuning the model are also provided through supplementary experiments. Comprehensive results show that the proposed methods significantly boost classification across different IoT datasets with high metric scores, outperforming other approaches.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101680"},"PeriodicalIF":6.0,"publicationDate":"2025-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144510875","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":"Detection of moving small targets in infrared images for urban traffic monitoring","authors":"Juan Wang ,&nbsp;Hao Yang ,&nbsp;Zizhen Zhang ,&nbsp;Nan Zhao ,&nbsp;Jixiang Shao ,&nbsp;Minghua Wu ,&nbsp;Zhigang Ma ,&nbsp;Jialu Zhu ,&nbsp;Xu An Wang ,&nbsp;Haina Song","doi":"10.1016/j.iot.2025.101673","DOIUrl":"10.1016/j.iot.2025.101673","url":null,"abstract":"<div><div>The Internet of Vehicles (IoV) and autonomous driving technologies require increasingly robust object detection capabilities, especially for small objects. However, reliably detecting small objects in urban traffic scenarios remains technically challenging under adverse weather conditions, including low illumination, rain, and snow. To address these challenges, we propose a fused IR–visible imaging approach using an enhanced YOLOv9 architecture. The proposed method employs a dual-branch semantic enhancement architecture, which achieves dynamic inter-modal feature weighting through a channel attention mechanism. The visible branch preserves texture details, while the infrared branch extracts thermal radiation characteristics, followed by multi-scale feature-level fusion. Firstly, we present UR-YOLO designed for detecting small targets in urban traffic environments. Secondly, we propose a novel DeeperFuse module that incorporates dual-branch semantic enhancement and channel attention mechanisms for effective multimodal feature fusion. Finally, by jointly optimizing fusion and detection losses, the method preserves critical details, enhances clarity and contrast. Experimental evaluation on the M^{relax special {t4ht=³}}FD dataset demonstrates improved detection performance relative to the baseline YOLOv9 model. The results show an increase of 1.4 percentage points in mAP (from 83.3% to 84.7%) and 2.2 percentage points in <span><math><mrow><mi>A</mi><msub><mrow><mi>P</mi></mrow><mrow><mi>s</mi><mi>m</mi><mi>a</mi><mi>l</mi><mi>l</mi></mrow></msub></mrow></math></span> (from 51.6% to 53.8%). Furthermore, our method achieves real-time processing at 30 FPS, making it suitable for deployment in urban autonomous driving scenarios. Future work will focus on enhancing model performance via multimodal fusion, lightweight design, and multi-scale feature learning. We will also develop diverse datasets to advance autonomous driving perception in complex environments.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101673"},"PeriodicalIF":6.0,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144481478","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":"Federated learning for anomaly detection on Internet of Medical Things: A survey","authors":"Rui P. Pinto,&nbsp;Bruno M.C. Silva,&nbsp;Pedro R.M. Inácio","doi":"10.1016/j.iot.2025.101677","DOIUrl":"10.1016/j.iot.2025.101677","url":null,"abstract":"<div><div>The Internet of Medical Things (IoMT), an extension of the Internet of Things (IoT) paradigm where interconnected medical devices can sense and act within healthcare environments, aims to improve patient comfort, optimize outcomes and streamline medical processes. IoMT has seen significant growth in recent years, transforming healthcare with advanced monitoring, diagnostics, and data-sharing capabilities, though it also faces security and privacy challenges. The widespread attack surface of IoMT, combined with the difficulty of embedding robust security mechanisms in resource-constrained medical devices, makes IoMT systems particularly attractive targets for cyberattacks and a source of numerous security challenges. Anomaly detection systems are frequently part of the solution for IoMT cybersecurity, but they face unique integration challenges, especially in environments where patient data privacy is paramount. Federated Learning (FL) offers a promising approach to address these privacy concerns by enabling distributed training without sharing raw data. This paper provides a comprehensive literature review of FL applications in anomaly detection within IoMT ecosystems. It describes recent implementations, highlights the main open issues, and identifies future research challenges. This work elucidates the feasibility and challenges of FL-based anomaly detection systems applied to IoMT, offering insights for advancing IoMT security.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101677"},"PeriodicalIF":6.0,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144481480","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 secure energy management model for Peer-to-Peer smart grids with user-centric constraints","authors":"Waqas Amin ,&nbsp;Qi Huang ,&nbsp;Jian Li ,&nbsp;Abdullah Aman Khan ,&nbsp;Umashankar Subramaniam ,&nbsp;Sivakumar Selvam","doi":"10.1016/j.iot.2025.101678","DOIUrl":"10.1016/j.iot.2025.101678","url":null,"abstract":"<div><div>In today’s smart grid era, ensuring fair energy distribution while protecting participants’ data privacy is a critical challenge, particularly in Peer-to-Peer (P2P) energy trading environments. To address this challenge, this paper presents a privacy-preserving energy management model that ensures fair energy allocation based on participants’ reported information. By identifying the demand-to-supply ratio, the proposed model classifies the market operation mode either buyers’ mode or sellers’ mode and manages energy accordingly. The model employs a quorum-based architecture that integrates SHA-256 encryption and Shamir’s Secret Sharing scheme to safeguard participants’ private data against potential cyber-attacks such as Man-in-the-Middle (MitM) and False Data Injection Attacks (FDIA). Simulation results demonstrate that once the system operator receives the valid threshold shares, the original information can be successfully reconstructed. Furthermore, the simulation also indicates that the proposed model not only improves grid stress by up to 76.60% during peak hours but also transforms the grid’s role from an energy taker to an energy contributor.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101678"},"PeriodicalIF":6.0,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144481477","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":"XBiDeep: A novel explainable artificial intelligence based intrusion detection system for Internet of Medical Things environment","authors":"Zeynep Turgut,&nbsp;Muhammet Sinan Başarslan","doi":"10.1016/j.iot.2025.101675","DOIUrl":"10.1016/j.iot.2025.101675","url":null,"abstract":"<div><div>In this study, an IDS XBiDeep based on the use of deep learning architectures for the IoMT - IoHT environments is proposed. To evaluate the performance of the proposed technique, three different datasets collected from IoMT environments: CICIoMT2024, IoMT-TrafficData, and ECU-IoHT are used. CICIoMT2024 and ECU-IoHT possess imbalanced data structures, while IoMT-TrafficData contains a balanced structure, allowing the effectiveness of the model to be examined across both balanced and imbalanced datasets. Rather than performing a simple binary classification between attack and benign data, multi-class classification is conducted to investigate various attack types. To achieve high performance across all IoMT datasets, RNN, LSTM, GRU, BiLSTM, and BiGRU architectures are tested individually and in hybrid configurations. The best results are observed with the hybrid BiGRU-BiLSTM model, which is subsequently integrated into the proposed XBiDeep architecture. Specifically, it reached 0.9975 accuracy for 6-class classification and 0.9985 for 19-class classification on the CICIoMT2024 dataset. On the IoMT-TrafficData dataset, the model attained 0.9990 accuracy, while 0.9987 accuracy was obtained on the ECU-IoHT dataset. The outcomes of the created architecture are analyzed using XAI models: SHAP and LIME. The SHAP analysis identifies key features distinguishing different attack types from benign data, while the LIME analysis highlights the most effective features for detecting each specific attack type. Importance of features is revealed both locally and globally, based on attack types and across the entire system. Hence, this study introduces an explainable deep learning-based IDS with high accuracy across diverse IoMT datasets and attacks.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"33 ","pages":"Article 101675"},"PeriodicalIF":6.0,"publicationDate":"2025-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144481479","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}