Internet of Things最新文献

{"title":"Corrigendum to ‘An explainable neuro-fuzzy intrusion detection system for industrial IoT’ [Internet of Things 38 (2026)101989]","authors":"Malena Pérez Sevilla, José Ángel Olivas Varela, Álvaro Herrero Cosio, Daniel Urda Muñoz, Jaime Andrés Rincón Arango","doi":"10.1016/j.iot.2026.101999","DOIUrl":"10.1016/j.iot.2026.101999","url":null,"abstract":"","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"39 ","pages":"Article 101999"},"PeriodicalIF":7.1,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148940806","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":"Corrigendum to “μTEE: Certification-ready RISC-V platform for secure IIoT lifecycle management” [Internet of Things 36 (2026) 101885]","authors":"Alessandro Cilardo, Jesús García-Rodríguez, Jose Manuel Merlos Espín, Roberto Nardone","doi":"10.1016/j.iot.2026.102018","DOIUrl":"10.1016/j.iot.2026.102018","url":null,"abstract":"","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"39 ","pages":"Article 102018"},"PeriodicalIF":7.1,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148940805","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":"A Gale–Shapley matching game for data collection waypoints using multi-UAVs in wide-area target localization applications","authors":"Mohammed Shurrab,&nbsp;Shakti Singh,&nbsp;Rabeb Mizouni,&nbsp;Hadi Otrok,&nbsp;Jamal Bentahar","doi":"10.1016/j.iot.2026.102016","DOIUrl":"10.1016/j.iot.2026.102016","url":null,"abstract":"<div><div>Unmanned aerial vehicles (UAVs) have emerged as a powerful enabler for smart environmental monitoring, particularly in applications that require rapid, flexible, and on-demand data collection such as target localization. Compared to stationary IoT sensing systems, UAVs provide mobility and adaptability, making them well suited for large-scale and hard-to-access areas. However, their limited onboard energy significantly constrains mission duration and coverage, posing a major challenge for wide-area sensing tasks. Therefore, in this work, an energy-aware, information-driven multi-UAV target localization framework is proposed. The framework integrates waypoint (WP) selection and clustering, path planning via Tabu Search, task assignment using two-sided Gale–Shapley game, a probabilistic localization algorithm based on Bayesian, and an interpolation and contouring mechanism. This mechanism introduces regions of importance (RoIs), coupled with a Potential Importance (PI) metric, to guide adaptive waypoint selection. This closed-loop design enables the system to progressively refine sensing decisions, improve localization efficiency while reducing redundant data collection. The proposed framework is assessed in comparison to different benchmarks, where the results show that the proposed approach consistently achieves better performance by up to ∼30%–60% in terms of localization time, mission completion time, energy consumption, system’s residual energy, and quality of localization. Furthermore, scalability and adaptability analyses are conducted under varying area sizes and target strengths, confirming the robustness of the proposed approach. Overall, the results demonstrate the effectiveness of the framework in achieving accurate and timely localization while preserving the energy resources of the UAV swarm.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"38 ","pages":"Article 102016"},"PeriodicalIF":7.1,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148523723","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":"Triple-branch dehazing network with grayscale distribution alignment for IoT-Driving visual perception","authors":"Feng Ling,&nbsp;Jinghua Zhang,&nbsp;Yan Zhang,&nbsp;Zhiguang Shi,&nbsp;Tsviatkou Viktar,&nbsp;Yu Zhang","doi":"10.1016/j.iot.2026.102001","DOIUrl":"10.1016/j.iot.2026.102001","url":null,"abstract":"<div><div>This study addresses the critical bottleneck of hazy image degradation in IoT-driving visual perception, where compromised image quality undermines object detection reliability by proposing a synergistic framework, Triple-branch Dehazing Network with Grayscale Distribution Alignment (TBDN-GDA). First, the core Triple-branch Dehazing Network (TBDN) resolves the long-standing trade-off between global dehazing consistency and local detail preservation in dynamic environments. Through two novel modules Local Attention Convolution (LAC) and Context-aware Feature Mixing (CAFM), it dynamically fuses CNN-derived local inductive biases and Transformer-captured global representations, outperforming conventional additive or modulation-based fusion strategies that fail to balance these two key requirements. Second, the Grayscale Distribution Alignment (GDA) mechanism effectively mitigates synthetic-to-real domain shift by aligning latent feature statistics with real-world hazy image distributions, thus opening up new possibilities for large-scale real-world IoT dehazing deployments. Extensive evaluations across four benchmarks (Haze-Det, Haze-RD, RTTS, and BEDDE) demonstrate that TBDN-GDA delivers competitive results in both quantitative image quality metrics and downstream object detection accuracy. From an IoT application perspective, our method shows promise in enhancing the operational reliability of autonomous driving under hazy conditions by improving detection accuracy. Moreover, with continued progress in model compression and hardware acceleration, the proposed framework may serve as a foundation for future IoT-oriented perception systems.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"38 ","pages":"Article 102001"},"PeriodicalIF":7.1,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148499377","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 security solutions for IoMT devices: A survey","authors":"Islam Hentous,&nbsp;Valentino Merlino,&nbsp;Tayeb Kenaza,&nbsp;Saïd Mahmoudi,&nbsp;Dario Allegra","doi":"10.1016/j.iot.2026.101991","DOIUrl":"10.1016/j.iot.2026.101991","url":null,"abstract":"<div><div>The Internet of Medical Things (IoMT) marks a transformative leap in modern healthcare by enabling intelligent, connected medical devices to support real-time monitoring, diagnostics, and treatment. However, this rapid digitalization also expands the attack surface of healthcare systems, exposing them to a wide array of sophisticated cyber threats. The unique characteristics of IoMT, including device heterogeneity, resource limitations, and operation across diverse network environments, render traditional cybersecurity approaches insufficient. In response, the integration of Artificial Intelligence (AI) into security architectures is emerging as a powerful paradigm shift. This survey systematically explores the fusion of AI and cybersecurity in IoMT, covering threat landscapes, attack surfaces, and existing defense mechanisms across medical environments. By identifying IoMT-specific security requirements and challenges, investigating AI-driven security mechanisms, and analyzing how such techniques can be adapted to IoMT constraints, this work fills a critical gap in existing research on secure medical IoT systems.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"38 ","pages":"Article 101991"},"PeriodicalIF":7.1,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148523264","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":"A terrain-aware Fresnel zone simulator for LoRa link feasibility in non-Urban areas","authors":"Vicente Torres-Sanz,&nbsp;Pablo Donate,&nbsp;Julio A. Sanguesa,&nbsp;Piedad Garrido,&nbsp;Francisco J. Martinez","doi":"10.1016/j.iot.2026.102012","DOIUrl":"10.1016/j.iot.2026.102012","url":null,"abstract":"<div><div>Planning reliable LoRaWAN deployments in non-urban environments demands tools that move beyond distance-based models to account for the complex interplay between terrain and radio wave propagation. This work addresses this need by introducing the Fresnel Zone Simulator (FresSim), an open-source web tool for terrain-aware LoRa link feasibility assessment in LoRaWAN deployments. FresSim is not intended to model the complete LoRaWAN PHY/MAC protocol stack or packet-level network dynamics. Instead, FresSim focuses on LoRa link feasibility by combining high-resolution Digital Elevation Models (DEMs) with first Fresnel Zone clearance analysis, providing a physically grounded metric to assess coverage viability for each node-gateway link. A key innovation of our tool is its dual operational mode: it supports both pre-deployment planning, generating extensive coverage maps (up to 50 km²), and post-deployment validation by seamlessly importing and analyzing real-world data from major LoRaWAN network servers (The Things Network and ChirpStack). To evaluate its predictive capability, we conducted an empirical validation using real deployment traces across diverse terrains. The results obtained show that FresSim correctly predicted all terrain-aware link feasibility outcomes in the evaluated scenarios, achieving higher agreement with field observations than LoRaSim, which does not account for terrain or Fresnel Zone effects. These results indicate that incorporating detailed topographic information and Fresnel Zone clearance can be critical for accurate LoRaWAN deployment planning in topographically complex, non-urban environments.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"38 ","pages":"Article 102012"},"PeriodicalIF":7.1,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148499266","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":"Element-wise weights aggregation for contribution-aware personalized federated learning with adaptive initialization","authors":"Zhiming Huang,&nbsp;Xiaohan Wang,&nbsp;Jiankang Dai","doi":"10.1016/j.iot.2026.102000","DOIUrl":"10.1016/j.iot.2026.102000","url":null,"abstract":"<div><div>Personalized Federated Learning (PFL) is designed to generate customized models for each client, so as to better adapt to the Non-Independent and Identically Distributed (Non-IID) client data. However, existing PFL methods adopt a model decoupling mechanism that separates the shared backbone network from the personalized classifier: clients participate in collaborative learning by updating the shared backbone network while fixing the personalized classifier. This gives rise to insufficient coordination between shared and personalized parameters, which in turn leads to inconsistencies in knowledge representation and optimization objectives. Moreover, existing PFL methods fail to leverage local objectives to update personalized models, resulting in the redundant integration of global model information into personalized models. This ultimately degrades the performance of local models and increases computational overhead. Therefore, this paper proposes a novel method named FedACE. First, through Adaptive Local Initialization, clients can adaptively extract valid information from the downloaded global model and guide the initialization of local models before each training iteration, aligning the models with the specific optimization objectives of each client. In addition, the server can efficiently select eligible clients for global training, reduce communication overhead via parameter sparsification, and allocate aggregation weights dynamically and rationally to enhance training stability. Extensive experiments are conducted on two benchmark datasets in computer vision to verify the effectiveness of FedACE, covering scenarios with varying degrees of data heterogeneity.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"38 ","pages":"Article 102000"},"PeriodicalIF":7.1,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148499375","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":"Digital paradigms in medication support: A comprehensive review of technologies and approaches","authors":"Mohammed Ghazal,&nbsp;Hadil Salman,&nbsp;Arwa Sheibani,&nbsp;Ayesha Amin,&nbsp;Marah Alhalabi,&nbsp;Abdalla Gad,&nbsp;Maha Yaghi,&nbsp;Mohamed Ramadan,&nbsp;Taimur Hassan,&nbsp;Mohammad Alkhedher","doi":"10.1016/j.iot.2026.102005","DOIUrl":"10.1016/j.iot.2026.102005","url":null,"abstract":"<div><div>Smart technological systems have emerged in the past years to aid in medication management among individuals, allowing patients to store and track their daily medication and aiding those prone to forgetfulness of dosage requirements. This paper reviews three areas concerning medication support systems, namely Internet-of-Things-based pill dispenser systems, medical paper digitization systems, and intelligent medication support systems. This paper’s main focus lies in reviewing the techniques adopted within each of the reviewed medication support areas to highlight their effectiveness, and identify common trends. For the Internet-of-Things-based pill dispenser systems, three major areas are reviewed: the pill dispensing mechanism, pill scheduling method, and the employed alarm systems. Regarding medical paper digitization systems, the reviewed methodologies are categorized into three key aspects, including those that rely on visual information processing, text and sequence modeling, and decision support systems. As for intelligent medication support systems, techniques such as dialog-based, reminder-based, and treatment supervision systems are reviewed.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"38 ","pages":"Article 102005"},"PeriodicalIF":7.1,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148499379","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":"An efficient and secure channel-free ring signature for VANETs","authors":"Shah Tahir Ali,&nbsp;Shuguang Li,&nbsp;Fahad Algarni,&nbsp;Wajahat Akbar,&nbsp;Insaf Ullah,&nbsp;Weiwei Jiang","doi":"10.1016/j.iot.2026.101992","DOIUrl":"10.1016/j.iot.2026.101992","url":null,"abstract":"<div><div>Vehicles and roads are becoming increasingly reliant on real-time communication between vehicles and infrastructure. The authenticity and integrity of these exchanges are highly important to ensure safety and operational reliability. Since these systems rely on continuous wireless communication for sending and receiving messages, security must be a major concern: any forged or altered messages can pose a direct risk to road safety. However, most of the available solutions introduce significant computation and communication costs. Public key infrastructures based on certificates are expensive to maintain, as they require certificates and verifying them, whereas identity-based solutions raise the issue of key escrows. Furthermore, some schemes are based on bilinear pairing and elliptic curve cryptography that can lead to high computational cost and communication overhead. Such overhead can limit scalability and useful deployability in highly dynamic vehicular environments where latency is hard and onboard resources are limited. Therefore, we propose an efficient and secure channel-free ring signature for VANETs that removes the role of certificate management and does not raise concerns related to key escrow. Our scheme guarantees the confidentiality of both senders and recipients. The method is both cost-effective and proven secure, allowing users to send their identities to the KGC over a public network. Moreover, the partial private key is securely transmitted over a public channel, eliminating the need for a dedicated secure channel. Our technique improves privacy and security in VANET environments by using strong cryptographic mechanisms to protect vehicle data from cyber attacks. We demonstrate the scheme’s security within the Random Oracle Model (ROM), establishing its unforgeability against both Type 1 and Type 2 attackers, reliant upon the difficulty of the hyperelliptic curve discrete logarithm problem. In particular, our scheme provides a high level of authentication with a lower computational and communication cost when compared with other existing schemes.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"38 ","pages":"Article 101992"},"PeriodicalIF":7.1,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148523721","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":"Pipelined UAV-Enabled federated learning for IoT networks","authors":"Shinyoung Cho,&nbsp;Chanmin Lee,&nbsp;Anna Cho,&nbsp;Sukyoung Lee","doi":"10.1016/j.iot.2026.102014","DOIUrl":"10.1016/j.iot.2026.102014","url":null,"abstract":"<div><div>With the rapid proliferation of IoT devices in the 6G era, federated learning (FL) has been introduced for distributed on-device training, improving network efficiency while preserving privacy. However, in regions with limited communication infrastructure, such as remote areas or post-disaster environments, IoT devices struggle to upload their locally trained models to a centralized cloud server. In such environments, UAV-enabled federated learning (UAV-FL) frameworks are an alternative, in which a UAV acts as a flying parameter server to collect and aggregate local models. Despite their potential, UAV-FL frameworks suffer from slow convergence and degraded global model accuracy due to device heterogeneity, which causes the Non-IID data problem and forces faster devices to wait for slower ones during each training round. To address these issues, we propose a Pipelined UAV-enabled Federated Learning (PUFL) mechanism. Specifically, the proposed mechanism uses pipelined uplink scheduling to allow more devices to participate in each FL round and selects devices whose updates are most likely to improve the global model, thereby mitigating the Non-IID data problem. Simulation results show that PUFL improves both model convergence speed and test accuracy compared to existing baselines.</div></div>","PeriodicalId":29968,"journal":{"name":"Internet of Things","volume":"38 ","pages":"Article 102014"},"PeriodicalIF":7.1,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148499384","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}