Future Internet

16 pages, 1267 KB

Open AccessArticle

Differentially Private Federated Learning with Adaptive Clipping Thresholds

by Jianhua Liu, Yanglin Zeng, Zhongmei Wang, Weiqing Zhang and Yao Tong

Future Internet 2026, 18(3), 148; https://doi.org/10.3390/fi18030148 (registering DOI) - 14 Mar 2026

Under non-independent and identically distributed (Non-IID) conditions, significant variations exist in local model updates across clients and training phases during the collaborative modeling process of differential privacy federated learning (DP-FL). Fixed clipping thresholds and noise scales struggle to accommodate these diverse update differences, [...] Read more.

Under non-independent and identically distributed (Non-IID) conditions, significant variations exist in local model updates across clients and training phases during the collaborative modeling process of differential privacy federated learning (DP-FL). Fixed clipping thresholds and noise scales struggle to accommodate these diverse update differences, leading to mismatches between local update intensity and noise perturbations. This imbalance results in data privacy leaks and suboptimal model accuracy. To address this, we propose a differential privacy federated learning method based on adaptive clipping thresholds. During each communication round, the server adaptively estimates the global clipping threshold for that round using a quantile strategy based on the statistical distribution of client update norms. Simultaneously, clients adaptively adjust their noise scales according to the clipping threshold magnitude, enabling dynamic matching of clipping intensity and noise perturbation across training phases and clients. The novelty of this work lies in a quantile-driven, round-wise global clipping adaptation that synchronizes sensitivity bounding and noise calibration across heterogeneous clients, enabling improved privacy–utility behavior under a fixed privacy accountant. Using experimental results on the rail damage datasets, our proposed method slightly reduces the attacker’s MIA ROC-AUC by 0.0033 and 0.0080 compared with Fed-DPA and DP-FedAvg, respectively, indicating stronger privacy protection, while improving average accuracy by 1.55% and 3.35% and achieving faster, more stable convergence. We further validate its effectiveness on CIFAR-10 under non-IID partitions. Full article

► Show Figures

Figure 1

26 pages, 1470 KB

Open AccessArticle

ANRF: An Adaptive Network Reconstruction Framework for Community Detection in Bipartite Networks

by Furong Chang, Songxian Wu, Yue Zhao and Farhan Ullah

Future Internet 2026, 18(3), 147; https://doi.org/10.3390/fi18030147 - 13 Mar 2026

Abstract

Bipartite network community detection is of significant importance for understanding the underlying structure and functional organization of real-world complex systems. Although many mature community detection algorithms exist for unipartite networks, they cannot be directly applied to bipartite networks due to their unique topological [...] Read more.

Bipartite network community detection is of significant importance for understanding the underlying structure and functional organization of real-world complex systems. Although many mature community detection algorithms exist for unipartite networks, they cannot be directly applied to bipartite networks due to their unique topological structure, characterized by heterogeneous node types and cross-layer connections. Furthermore, some existing bipartite network community detection methods still rely heavily on manual experience to set key parameters, which limits their applicability and scalability in practical scenarios. To address these issues, this paper proposes an enhanced framework—the Adaptive Network Reconstruction Framework (ANRF)—by introducing an adaptive parameter optimization mechanism based on the existing Network Reconstruction Framework (NRF). This framework can be effectively integrated with traditional unipartite network community detection algorithms to achieve automatic community detection with reduced dependence on manual parameter tuning. The core procedure of the method consists of four main steps. First, we calculate the interaction forces between node pairs. Second, through comprehensive analysis of the network topological features, we adaptively determine the threshold parameter

θ

and related parameters for the interaction forces. Third, based on these thresholds and parameters, we perform edge filtering on the bipartite network to construct a reconstructed network. Finally, we apply unipartite community detection algorithms directly to the reconstructed network to obtain the community structure. To validate the effectiveness of ANRF, we combined it with the Louvain method and the Greedy modularity method, and conducted experimental evaluations on multiple synthetic and real-world network datasets. A systematic comparison with current state-of-the-art algorithms was made. The experimental results on multiple synthetic and real-world datasets within our evaluated scope demonstrate that ANRF achieves competitive performance in terms of community modularity and community density compared to state-of-the-art algorithms, while significantly reducing reliance on manual parameter tuning and enhancing robustness under the tested conditions. Full article

(This article belongs to the Section Big Data and Augmented Intelligence)

► Show Figures

Graphical abstract

31 pages, 974 KB

Open AccessArticle

Model Procurement for Industrial Cyber-Physical Systems Using Cryptographic Performance Attestation

by Jay Bojič Burgos, Urban Sedlar and Matevž Pustišek

Future Internet 2026, 18(3), 146; https://doi.org/10.3390/fi18030146 - 13 Mar 2026

Abstract

Integrating third-party Machine Learning (ML) models into industrial Operational Technology (OT) creates a procurement deadlock: operators cannot verify vendor performance claims without sharing representative evaluation data with vendors, while vendors refuse to reveal proprietary model weights before purchase, rendering traditional safeguards such as [...] Read more.

Integrating third-party Machine Learning (ML) models into industrial Operational Technology (OT) creates a procurement deadlock: operators cannot verify vendor performance claims without sharing representative evaluation data with vendors, while vendors refuse to reveal proprietary model weights before purchase, rendering traditional safeguards such as Non-Disclosure Agreements technically unenforceable. This paper introduces a framework combining Zero-Knowledge Proofs (ZKPs) with smart contracts to enable trust-minimized, cryptographically verifiable competitive model procurement in Industrial Cyber-Physical Systems (ICPS). Vendors cryptographically prove that their model outperforms a legacy baseline without disclosing proprietary weights, a process we term cryptographic performance attestation, while the on-chain workflow automates escrow, proof verification, and best-vendor selection with arbiter-based dispute resolution. ZKP privacy is scoped to vendor model weights; operator-side evaluation-data confidentiality is managed separately via synthetic, de-identified, or public benchmark data. We analyze three ZKP workflow variations and evaluate them on consumer-grade hardware, achieving proving times of approximately three seconds and sub-dollar on-chain verification costs under Layer-2 fee assumptions for the recommended single-proof variation, while identifying computational trade-offs of recursive proof aggregation. The entire verification phase operates offline with no impact on real-time OT control paths, bridging the IT/OT pre-transaction trust gap while deferring artifact deployment to existing OT tooling. Full article

(This article belongs to the Special Issue Cyber-Physical Systems in Industrial Communication Systems)

► Show Figures

Figure 1

26 pages, 755 KB

Open AccessArticle

A Stage-Wise Framework Using Class-Incremental Learning for Unknown DoS Attack Detection

by Juncheng Ge, Yaokai Feng and Kouichi Sakurai

Future Internet 2026, 18(3), 145; https://doi.org/10.3390/fi18030145 - 12 Mar 2026

Abstract

Denial-of-Service (DoS) attacks remain one of the most dangerous threats in modern Internet environments. They aim to overwhelm networks, servers, or online services with massive volumes of traffic, and maintaining service availability is a core pillar of cybersecurity. More importantly, DoS attack techniques [...] Read more.

Denial-of-Service (DoS) attacks remain one of the most dangerous threats in modern Internet environments. They aim to overwhelm networks, servers, or online services with massive volumes of traffic, and maintaining service availability is a core pillar of cybersecurity. More importantly, DoS attack techniques continue to evolve. However, traditional intrusion detection systems (IDS) trained on fixed attack categories struggle to identify previously unknown DoS attack types and cannot dynamically incorporate newly emerging classes. To address this challenge, this study proposes a stage-wise network intrusion detection framework that integrates unknown attack detection, attack discovery, and class-incremental learning into a unified pipeline. The framework consists of three stages. First, an autoencoder-based anomaly detection approach is used to separate potential unknown DoS attack samples from known classes. Second, a clustering-and-merging strategy is applied to the detected unknown DoS samples to discover emerging attack clusters with similar structural characteristics. Third, the classifier architecture is expanded for each newly discovered cluster through a class-incremental learning mechanism, enabling the continual incorporation of new attack classes while maintaining stable detection performance on previously learned classes. Experimental results on the DoS category of the NSL-KDD dataset demonstrate that the proposed stage-wise framework can effectively isolate samples of unknown DoS attacks, accurately aggregate emerging attack clusters, and incrementally integrate newly discovered attack classes without significantly degrading recognition performance on previously learned classes. These results confirm the capability of the proposed framework to handle progressively emerging unknown DoS attacks. Full article

► Show Figures

Figure 1

26 pages, 5380 KB

Open AccessArticle

Analyzing Characteristics of Public Transport Complex Networks Based on Multi-Source Big Data Fusion: A Case Study of Cangzhou, China

by Linfang Zhou, Yongsheng Chen, Dongpu Ren and Qing Lan

Future Internet 2026, 18(3), 144; https://doi.org/10.3390/fi18030144 - 11 Mar 2026

Abstract

Quantitative evaluation of public transit networks (PTNs) with complex-network models informs route optimization and operational adjustments. Prior studies emphasize large cities and pay limited attention to small-sized urban systems. This study examines the bus network of Cangzhou City, Hebei Province, China, to broaden [...] Read more.

Quantitative evaluation of public transit networks (PTNs) with complex-network models informs route optimization and operational adjustments. Prior studies emphasize large cities and pay limited attention to small-sized urban systems. This study examines the bus network of Cangzhou City, Hebei Province, China, to broaden the empirical scope and characterize PTNs in smaller cities. The dataset for this study comprises route and stop records, passenger boarding logs, and bus GPS traces. We develop a general workflow for bus data cleaning and completion. To characterize the dynamic bus network and compare it with the static network, we construct a static network and Directed Weighted Dynamic Network I (DWDN I) using the L-space method, and we construct Directed Weighted Dynamic Network II (DWDN II) using the P-space method. We calculated network metrics including degree, weighted degree, clustering coefficient, path length, network diameter, network efficiency, and small-world coefficient. The principal results show that: (1) at the macroscopic level, the dynamic PTN tracks passenger demand, as the average degree, weighted average degree, and clustering coefficient fluctuate in concert with passenger flows; (2) key stations concentrate in the urban core, and stations with high weighted degree display pronounced spatial autocorrelation; (3) the exponential form of the weighted-degree distribution indicates that the examined bus network is not scale-free, while the dynamic network’s small-world coefficient exceeds that of the static network across time periods, reflecting stronger small-world characteristics. This study integrates network and spatial attributes of the PTN to offer an exploratory case for investigating public transit networks in third-tier cities. The findings can inform comparable studies and offer practical guidance for bus operators. Full article

(This article belongs to the Section Big Data and Augmented Intelligence)

► Show Figures

Figure 1

19 pages, 6254 KB

Open AccessArticle

Earthquake Magnitude Detection Utilizing a Novel Hybrid Earth–Transformer–LSTM Architecture

by Amir A. Ghavifekr, Elman Ghazaei, Mohsen Mirzajani and Paolo Visconti

Future Internet 2026, 18(3), 143; https://doi.org/10.3390/fi18030143 - 11 Mar 2026

Abstract

One of the complicated and demanding tasks in seismology is the reliable detection of earthquakes. The key challenge is that the detection models must be applied to a specific region, and models trained on one region may not perform as well in others. [...] Read more.

One of the complicated and demanding tasks in seismology is the reliable detection of earthquakes. The key challenge is that the detection models must be applied to a specific region, and models trained on one region may not perform as well in others. The limitations of datasets for most regions of the world pose another task. Comprehensive, high-quality datasets are essential for developing robust earthquake detection algorithms. Despite these challenges, developing effective earthquake detection systems is critically important. This paper proposes a novel deep network, Earth–Transformer–LSTM (ETL), to estimate earthquake magnitude with high precision. The proposed method uses Transformer encoders as its first layer to extract profound features from the dataset. To obtain highly accurate results, the extracted data is used as the input to the Long Short-Term Memory (LSTM) neural network. Additionally, one-dimensional convolution is replaced by Multi-Layer Perceptron (MLP), which performs better in Transformer encoders’ feed-forward networks. The Turkey earthquake dataset 2000–2018 was used in this research because significant earthquakes have occurred in this region in recent years. According to the obtained results, the proposed method’s Root Mean Squared Error (RMSE) is 0.7, representing a noticeable improvement over advanced conventional models. Full article

(This article belongs to the Special Issue Artificial Intelligence: Innovation, Applications and Transformative Experiences)

► Show Figures

Figure 1

26 pages, 1910 KB

Open AccessArticle

A Framework to Measure Maturity of Industrial IoT Technology for Agricultural Regulatory Compliance Activities and Decentralization

by Jinying Li, Ananda Maiti and Stephen Cahoon

Future Internet 2026, 18(3), 142; https://doi.org/10.3390/fi18030142 - 11 Mar 2026

Abstract

Compliance checks are a critical aspect of any agricultural production and supply chain. These comprise a chain of activities and responsibilities shared by producers and regulators. Currently, the responsibility for collecting and processing data primarily lies with the regulators. Regulators are the primary [...] Read more.

Compliance checks are a critical aspect of any agricultural production and supply chain. These comprise a chain of activities and responsibilities shared by producers and regulators. Currently, the responsibility for collecting and processing data primarily lies with the regulators. Regulators are the primary users of regulatory technologies, while producers, such as farmers, have limited capabilities. This has been due to a lack of reliable hardware and software technology that can be deployed at the producer’s sites. However, recent advancements in Artificial Intelligence (AI), the Industrial Internet of Things (IIoT), and cloud computing have significantly increased reliability and reduced deployment costs. This paper reviews the regulatory landscape of agriculture, regulatory technologies, IIoT, and their feasibility for producers, using a new Technology Readiness Framework for Compliance, specifically designed for Regulatory Technologies (RegTech) and regulation compliance. It classifies technologies into three categories: Early Stage, Emerging, and Established. It concludes that most essential agricultural issues already have mature technological solutions within the scope of IIoT that producers can use directly without supervision, while still maintaining the integrity and validity of the data. The paper discusses and measures the maturity of IIoT and other electronic and digital technologies for integration into RegTech. This categorization offers a new perspective on technology readiness and lays the foundation for future RegTech platform design, e.g., by decentralizing and empowering producers with reliable technologies. Full article

(This article belongs to the Special Issue AI and Industry 4.0)

► Show Figures

Graphical abstract

4 pages, 1020 KB

Open AccessEditorial

Machine Learning and IoT as Enablers of Intelligent Industrial Transformation

by Paulina Ayala and Marcelo V. Garcia

Future Internet 2026, 18(3), 141; https://doi.org/10.3390/fi18030141 - 11 Mar 2026

Abstract

The rapid expansion of cyber-physical systems, sensor networks, and data-driven automation platforms has profoundly reconfigured the operational logic of contemporary industrial environments [...] Full article

(This article belongs to the Special Issue Machine Learning and Internet of Things in Industry 4.0)

27 pages, 656 KB

Open AccessArticle

Towards a Protocol-Aware Intrusion Detection System for LoRaWAN Networks

by Zsolt Bringye, Rita Fleiner and Eszter Kail

Future Internet 2026, 18(3), 140; https://doi.org/10.3390/fi18030140 - 9 Mar 2026

Abstract

The increasing reliance of Internet of Things (IoT) applications on low-power wide-area network technologies, particularly Long Range Wide Area Network (LoRaWAN), has amplified the need for security monitoring approaches that go beyond attack-specific signatures and generic traffic anomalies. Existing solutions are often tailored [...] Read more.

The increasing reliance of Internet of Things (IoT) applications on low-power wide-area network technologies, particularly Long Range Wide Area Network (LoRaWAN), has amplified the need for security monitoring approaches that go beyond attack-specific signatures and generic traffic anomalies. Existing solutions are often tailored to individual threat scenarios or rely on statistical indicators, which limits their ability to systematically capture protocol-level misuse in an interpretable manner. This paper addresses this gap by proposing a protocol-aware validation methodology based on a Digital Twin abstraction of LoRaWAN communication behavior. The Over-The-Air Activation (OTAA) procedure is modeled as a finite-state machine that encodes expected message sequences, timing constraints, and specification-driven state transitions. Observed network events are continuously evaluated against this formal state model, enabling the identification of protocol-level deviations indicative of anomalous or non-conformant behavior. Illustrative examples include replay behavior, timing inconsistencies, and integrity-related anomalies, although the framework is not limited to predefined attack categories. The results demonstrate that state machine-based Digital Twin provides a structured and extensible foundation for protocol-aware security validation and Security Operation Center (SOC)-oriented telemetry enrichment. In this sense, the presented approach represents a concrete step toward protocol-aware intrusion detection for LoRaWAN networks by establishing a state-synchronized semantic validation layer upon which higher-level detection mechanisms can be built. Full article

(This article belongs to the Special Issue Anomaly and Intrusion Detection in Networks)

► Show Figures

Graphical abstract

16 pages, 1301 KB

Open AccessArticle

Implementation and Evaluation of an Open-Source Chatbot for Patient Information Leaflets

by Lisa Heiler, Katharina Kirchsteiger, Sten Hanke and Markus Bödenler

Future Internet 2026, 18(3), 139; https://doi.org/10.3390/fi18030139 - 9 Mar 2026

Abstract

Accessing and understanding medication information can be challenging for many people, especially when patient information leaflets (PILs) are long, complex, and printed in small font. This study presents MediChat, an open-source, locally executable chatbot designed to provide reliable, easy-to-read answers to medication-related questions [...] Read more.

Accessing and understanding medication information can be challenging for many people, especially when patient information leaflets (PILs) are long, complex, and printed in small font. This study presents MediChat, an open-source, locally executable chatbot designed to provide reliable, easy-to-read answers to medication-related questions based exclusively on official PILs. MediChat follows a retrieval-augmented generation (RAG) architecture: PILs from the Austrian Medicinal Product Index are received via API, converted to text, split into overlapping chunks, embedded, and stored in a Chroma vector database. From there the top-k relevant chunks are retrieved, and Llama 3.1 generates German responses based on this evidence. The system was evaluated using a hybrid framework. Quantitatively, 200 yes/no questions across ten drugs were answered with 80% accuracy, overall precision 0.977, recall 0.686, F1-score 0.806, and a mean response time of 727 ms. Qualitatively, two personas were used in eight simulated dialogues. Response times were around 1.1–1.3 s, and task completion exceeded 85% with high ratings for relevance and quantity. These results indicate that an open-source RAG chatbot can deliver leaflet-grounded, user-friendly medication information and provide a reproducible template for future healthcare chatbot evaluations. Full article

(This article belongs to the Special Issue Artificial Intelligence (AI) in Healthcare: Technologies, Applications, Challenges)

► Show Figures

Figure 1

35 pages, 5289 KB

Open AccessArticle

Sentiment Classification of Amazon Product Reviews Based on Machine and Deep Learning Techniques: A Comparative Study

by Eman Daraghmi and Noora Zyadeh

Future Internet 2026, 18(3), 138; https://doi.org/10.3390/fi18030138 - 7 Mar 2026

Abstract

Sentiment classification plays a crucial role in analyzing customer feedback to identify market trends, enhance product recommendations, and improve customer satisfaction. This study focuses on sentiment analysis of Amazon reviews using two major datasets—Fine Food Reviews and Unlocked Mobile Reviews—which exhibit label imbalance. [...] Read more.

Sentiment classification plays a crucial role in analyzing customer feedback to identify market trends, enhance product recommendations, and improve customer satisfaction. This study focuses on sentiment analysis of Amazon reviews using two major datasets—Fine Food Reviews and Unlocked Mobile Reviews—which exhibit label imbalance. To address this challenge, both oversampling and undersampling techniques were applied to balance the datasets. Various machine learning (ML) algorithms, including Random Forest (RF), Logistic Regression (LR), Support Vector Machine (SVM), Naïve Bayes (NB), and Gradient Boosting Machine (GBM), as well as deep learning (DL) models such as Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and transformer-based models like RoBERTa, were implemented. After data cleaning and preprocessing, models were trained, and performance was evaluated. The results indicate that oversampling significantly enhances classification accuracy, particularly for the Fine Food dataset. Among ML models, Random Forest achieved the highest accuracy due to its ensemble approach and robustness in handling high-dimensional data. DL models, particularly RoBERTa, also demonstrated superior performance owing to their capacity to capture contextual dependencies. The findings emphasize the importance of data balancing for optimal sentiment analysis and contribute valuable insights toward advancing automated opinion classification in e-commerce applications. Full article

(This article belongs to the Section Big Data and Augmented Intelligence)

► Show Figures

Figure 1

27 pages, 566 KB

Open AccessArticle

Digital Twins at the Edge: A High-Availability Framework for Resilient Data Processing in IoT Sensor Networks

by Madalin Neagu, Codruta Maria Serban, Anca Hangan and Gheorghe Sebestyen

Future Internet 2026, 18(3), 137; https://doi.org/10.3390/fi18030137 - 6 Mar 2026

Abstract

The expansion of Internet-of-Things deployments at the network edge challenges service continuity, as single points of failure can interrupt critical data-processing pipelines. This paper introduces the Operational Digital Twin (ODT) —a live, state-synchronized standby system designed for node-level failover in resource-constrained edge environments. [...] Read more.

The expansion of Internet-of-Things deployments at the network edge challenges service continuity, as single points of failure can interrupt critical data-processing pipelines. This paper introduces the Operational Digital Twin (ODT) —a live, state-synchronized standby system designed for node-level failover in resource-constrained edge environments. In contrast to Digital Twins designed for modeling and analysis, an ODT is designed for operational continuity, standing ready to assume control when the primary node fails. We instantiate this concept through a self-configuring, high-availability architecture that implements the ODT for node-level redundancy. To ground this new conceptual category empirically, we define and validate four measurable criteria for ODT fidelity—state fidelity, synchronization timeliness, behavioral mirroring, and failover validation—establishing a framework that extends beyond passive replication. The design adopts a primary–secondary model with automated node discovery, configuration mirroring, and Virtual IP-based failover. Fault-injection experiments demonstrate low failover latency, prompt service restoration, limited message loss during transitions, and minimal resource overhead. These findings demonstrate that the proposed Operational Digital Twin mechanism reduces single points of failure and provides a lightweight, cost-efficient approach to sustaining reliable data processing in distributed edge environments. Full article

(This article belongs to the Special Issue IoT Architecture Supported by Digital Twin: Challenges and Solutions)

► Show Figures

Graphical abstract

24 pages, 6188 KB

Open AccessArticle

Multi-Modal Artificial Intelligence for Smart Cities: Experimental Integration of Textual and Sensor Data

by Nouf Alkhater

Future Internet 2026, 18(3), 136; https://doi.org/10.3390/fi18030136 - 5 Mar 2026

Abstract

Smart city decision-making increasingly relies on heterogeneous urban data sources. Dense traffic sensor streams provide continuous quantitative measurements, while citizen-generated textual reports offer event-driven contextual information. However, integrating these modalities remains challenging due to temporal misalignment, textual sparsity, and semantic noise. This paper [...] Read more.

Smart city decision-making increasingly relies on heterogeneous urban data sources. Dense traffic sensor streams provide continuous quantitative measurements, while citizen-generated textual reports offer event-driven contextual information. However, integrating these modalities remains challenging due to temporal misalignment, textual sparsity, and semantic noise. This paper investigates multi-modal learning for traffic congestion severity prediction through an experimental integration of open traffic sensor data (METR-LA: Los Angeles, USA) and citizen-generated textual reports (NYC 311: New York City, USA). Congestion severity is formulated as a four-class classification task derived from traffic speed measurements. We propose an end-to-end framework that combines: (i) sensor time-series encoding using a GRU-based temporal encoder, (ii) textual representation learning using a BERT-based encoder, (iii) a symmetric time-window alignment strategy (±Δ) to associate irregular reports with sensor time steps, and (iv) multiple fusion architectures, including early fusion, late fusion, and a cross-attention module for cross-modal interaction modeling. Experiments on publicly available datasets show that multi-modal early fusion achieves the best overall performance (Accuracy = 0.8283, Macro-F1 = 0.8231) compared to uni-modal baselines. In the studied cross-city setting with sparse and weakly aligned textual signals, the proposed cross-attention fusion does not outperform the strong sensor-only baseline, suggesting that the sensor modality dominates when cross-modal signal strength is limited. These results highlight both the potential and the practical constraints of multi-modal fusion in heterogeneous smart-city environments, emphasizing the importance of alignment design, modality relevance, and transparent experimental validation. Full article

► Show Figures

Graphical abstract

32 pages, 599 KB

Open AccessArticle

MAPE-ZT: A Multi-Layer Access Policy Encryption System for Zero Trust Architectures

by Ashutosh Soni, Surendra Kumar Nanda, Jayanti Rout, Mrutyunjaya Sathua, Ganapati Panda and Manob Jyoti Saikia

Future Internet 2026, 18(3), 135; https://doi.org/10.3390/fi18030135 - 5 Mar 2026

Abstract

Organizations usually rely on stringent access control mechanisms where access policies are an important asset. Their storage or transmission in plaintext can compromise sensitive access rules. It is important in dynamic environments where access decisions are made in real time such as Zero [...] Read more.

Organizations usually rely on stringent access control mechanisms where access policies are an important asset. Their storage or transmission in plaintext can compromise sensitive access rules. It is important in dynamic environments where access decisions are made in real time such as Zero Trust (ZT). Existing ZT approaches were found to oversee the aspect of securing these policies. This investigation presents a Multi-layer Access Policy Encryption System for ZT systems (MAPE-ZT). The first stage uses the trapdoor index to generate a secure index to find the applicable access policies. Advanced Encryption Standard-256 is used in counter mode for the encryption of the policies. They are re-encrypted using the Ciphertext-Policy Attribute-Based Encryption (CP-ABE) to allow decryption based on a matching set of attributes. Various experiments using quantitative metrics, including comparison with baseline access control systems simulation, scalability evaluation, storage overhead, etc., highlight the efficacy of the MAPE-ZT and establish new benchmarks. The result count entropy for the policies ranged 3.84–4.21 for different scales of policies. The evaluation in different scales of systems shows that the MAPE-ZT reduces various observable patterns even if the deployment size grows. Its unique design of securing policies makes this approach scalable for multi-domain integration. Full article

(This article belongs to the Collection Information Systems Security)

► Show Figures

Graphical abstract

18 pages, 787 KB

Open AccessArticle

Multi-Criteria Selection of Network Security Configuration Using NSGA-II

by Bagdat Yagaliyeva, Valery Lakhno, Myroslav Lakhno, Boris Gusev, Kaiyrbek Makulov and Tomiris Sundet

Future Internet 2026, 18(3), 134; https://doi.org/10.3390/fi18030134 - 5 Mar 2026

Abstract

The problem of multi-criteria selection of network security configurations (NSC) under resource constraints and the necessity to comply with information security (IS) policies is addressed in this study. A formal mathematical model of the problem has been developed, encompassing the definition of a [...] Read more.

The problem of multi-criteria selection of network security configurations (NSC) under resource constraints and the necessity to comply with information security (IS) policies is addressed in this study. A formal mathematical model of the problem has been developed, encompassing the definition of a set of possible security mechanism configurations, the formalization of objective functions reflecting security levels, throughput, and deployment costs, and the introduction of constraints on feasible solutions. The NSGA-II (Non-dominated Sorting Genetic Algorithm II) optimization algorithm is employed to generate a set of Pareto-optimal solutions, ensuring uniform coverage of compromise configurations. A software package implemented in Python 3 incorporates modules for population generation, fitness evaluation, selection, crossover, mutation operators, and result visualization. Computational experiments (CE) were conducted to validate the effectiveness of the proposed approach. The evolution dynamics of the Pareto hypervolume were analyzed, the uniformity of solution distribution in the objective space was studied, and the impact of algorithm parameters on convergence to the optimal solution was examined. The results demonstrate that the proposed methodology enables the formation of NSC sets that achieve a balanced trade-off between security, throughput, and IS system deployment costs. Full article

(This article belongs to the Special Issue IoT Networks Security)

► Show Figures

Graphical abstract

25 pages, 633 KB

Open AccessArticle

Lightweight LSTM-Based Homogeneous Transfer Learning for Efficient On-Device IoT Intrusion Detection

by Amjad Gamlo, Sanaa Sharaf and Rania Molla

Future Internet 2026, 18(3), 133; https://doi.org/10.3390/fi18030133 - 4 Mar 2026

Abstract

The emergence of the Internet of Things (IoT) has introduced major security challenges. Deep learning models have shown strong potential for intrusion detection. However, they often require large datasets and high computational resources. In contrast, IoT environments are resource-constrained and lack sufficient labeled [...] Read more.

The emergence of the Internet of Things (IoT) has introduced major security challenges. Deep learning models have shown strong potential for intrusion detection. However, they often require large datasets and high computational resources. In contrast, IoT environments are resource-constrained and lack sufficient labeled data. This paper proposes a lightweight intrusion detection approach based on Long Short-Term Memory (LSTM) networks and homogeneous transfer deep learning. The model is first trained on a subset of the BoT-IoT dataset as a source domain. It is then fine-tuned on a disjoint subset containing a rare attack type. This setup represents adaptation to unseen attack behaviors within the same environment. By freezing earlier layers and fine-tuning only the final layers, the method reduces training overhead while preserving performance. This is important to meet the IoT requirement for frequent, lightweight model updates on resource-constrained devices. The proposed model achieved 99.9% accuracy, a macro F1-score of 0.96, and a 47.8% reduction in training time compared to training from scratch. Extensive experiments confirm that it maintains balanced detection across both common and rare classes. Full article

► Show Figures

Figure 1

42 pages, 2052 KB

Open AccessArticle

GEMS: Gas-Enhanced Marine Search for Optimizing Fusion Mamba-Attention Networks for Fake Review Classification

by Sharon Roji Priya C., Deepalakshmi Perumalsamy and Rajermani Thinakaran

Future Internet 2026, 18(3), 132; https://doi.org/10.3390/fi18030132 - 2 Mar 2026

Abstract

The rise of fake reviews has become a major problem for trust in e-commerce sites. As for traditional machine learning solutions, they fail to capture the nuanced language that separates real reviews from fake reviews. In this work, we introduce a new hybrid [...] Read more.

The rise of fake reviews has become a major problem for trust in e-commerce sites. As for traditional machine learning solutions, they fail to capture the nuanced language that separates real reviews from fake reviews. In this work, we introduce a new hybrid metaheuristic algorithm that optimizes the Fusion Mamba-Attention Network (FMA-Net) for fake review detection, called GEMS (Gas-Enhanced Marine Search). GEMS is a unique combination of the exploration capabilities of the Enhanced Marine Predators Algorithm and the exploitation process of Henry Gas Solubility Optimization, offering a dual-phase optimization design for high-dimensional, asymmetric, metaheuristic-configured GEMS-optimized FMA-Net. Geometric enhancement of GEMS optimization provides GEMS-optimized FMA-Net with an accuracy of 96.8%, F1-score of 95.4%, and AUC-ROC of 97.2%, marking 3–7% improvement over the current best models for fake review detection on the Yelp, Amazon, and Google Reviews datasets. We lower the average time of hyperparameter optimization using GEMS with FMA-Net to achieve 68% reduction in overall time spent in grid search and 42% lower for complexity in comparison to genetic algorithms. The contributions of this work are the first hybrid metaheuristic for transformers, a mathematically formulated GEMS algorithm, and an extensive empirical study for proving multi-dimensional metric plausibility. Full article

► Show Figures

Graphical abstract

3 pages, 144 KB

Open AccessEditorial

Deep Learning in Recommender Systems

by María N. Moreno-García and Fernando de la Prieta

Future Internet 2026, 18(3), 131; https://doi.org/10.3390/fi18030131 - 2 Mar 2026

Abstract

Recommender systems have undertaken significant advances over the last years, evolving from collaborative filtering techniques to deep learning architectures capable of modeling complex and multimodal interactions [...] Full article

(This article belongs to the Special Issue Deep Learning in Recommender Systems)

31 pages, 5508 KB

Open AccessArticle

An Edge–Fog–Cloud IoT Framework for Real-Time Cardiac Monitoring and Rapid Clinical Alerts in Hospital Wards

by Tehseen Baig, Nauman Riaz Chaudhry, Reema Choudhary, Pankaj Yadav, Younus Ahamad Shaik and Ayesha Rashid

Future Internet 2026, 18(3), 130; https://doi.org/10.3390/fi18030130 - 2 Mar 2026

Abstract

The difficulties of continuously monitoring cardiac patients in general hospital wards are still present because of the manual charting system and the slow clinical reaction to worsening physiological state. This paper outlines an edge- and fog-based Internet of Things (IoT) healthcare system to [...] Read more.

The difficulties of continuously monitoring cardiac patients in general hospital wards are still present because of the manual charting system and the slow clinical reaction to worsening physiological state. This paper outlines an edge- and fog-based Internet of Things (IoT) healthcare system to acquire, process, and prioritize the vital signs of patients in real time to minimize the alert latency and increase the time of clinical interventions. Wearable 12-lead ECG sensors transmit physiological measurements, such as heart rate, blood pressure, and oxygen saturation, to an intelligent edge service, where preprocessing, triage by threshold, and machine learning ECG classification are performed, and selective synchronization of physiological data with a cloud backend and data delivery to the clinician are made possible by a mobile application. The proposed architecture combines a ribbon-like streaming scheme, Flask-based gateway services, and Firebase Firestore to coordinate scalable mob/cloud with the help of multi-client data dissemination. To encompass borderline clinical deterioration, which is often unnoticed by conventional threshold systems, physiological parameters are classified into normal, alarming, emergency, and a new state, average. The Pan–Tompkins++ peak detector algorithm and multiple edge-resident classifiers, such as random forest, XGBoost, decision tree, naive Bayes, K-nearest neighbor, and support vector machine, are used to analyze the ECG waveforms. Experimental analysis of PhysioNet datasets and tests in real wards prove that the ensemble models can reach the highest possible ECG classification precision of 91.96 percent and snapshot-driven mobile alerts can decrease routine patient evaluation time by several minutes, to an average of 15.23 ± 2.71 s. These results suggest that edge-centric IoT systems can be appropriate in latency-critical hospital settings and that fog-based coordination is useful in next-generation smart healthcare systems. Full article

(This article belongs to the Special Issue Edge and Fog Computing for the Internet of Things, 2nd Edition)

► Show Figures

Graphical abstract

Journal Description

Future Internet

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI