Concurrency and Computation-Practice & Experience最新文献

The rapid proliferation of Internet of Things (IoT) systems has underscored the critical need for robust security measures to safeguard interconnected devices and data. This study presents an extensive bibliometric analysis of research advancements in anomaly detection in an IoT environment, leveraging data from the Web of Science repository to comprehend key trends, influential contributors, and evolving research themes. The analysis identifies the most prolific organizations, authors, and countries contributing to IoT anomaly detection literature, highlighting global scientific production and collaboration networks. The study traces publication trends, revealing the temporal distribution of article production and the impact of locally and globally cited sources. It also examines the most relevant authors in the field, their scholarly influence, and the dynamics of their research output over time. The co-occurrence of authors' keywords provides insights into emerging themes and the evolution of research focus areas. At the same time, a detailed review of the most globally cited articles elucidates foundational contributions to the field. Additionally, the study analyzes the frequency and evolution of key terms, identifying trending topics that shape current and future research. The authors' and countries' collaboration networks illustrate the extent of international cooperation, highlighting key partnerships driving innovation. The application areas, challenges, and future research directions are also discussed, offering valuable guidance for further research. This bibliometric analysis offers a valuable resource for researchers and practitioners seeking to understand this domain's development, current state, and future research trajectory.

Delay-tolerant networks (DTNs) are increasingly used in environments characterized by intermittent connectivity, long delays, and limited resources, such as disaster recovery, vehicular networks, and remote sensing. However, ensuring strong security and privacy in DTNs remains challenging due to unreliable communication links and the lack of continuous trust relationships among nodes. Although blockchain technology offers decentralization, immutability, and enhanced trust, its direct integration into DTNs is hindered by scalability issues, consensus latency, and computational overhead. The motivation of this study is to design a secure and privacy-preserving DTN framework that effectively balances blockchain-induced overhead with network performance. To address these challenges, this work proposes a blockchain-enhanced DTN architecture incorporating multiple optimization and security mechanisms. Adaptive Bilateral Kernel Filtering (ABKF) is employed to eliminate noise while preserving critical data features. Network Function Virtualization Orchestration with Transport Layer Security (NFVO-TLS) ensures authenticated and secure data transmission. Secure Multi-Party Computation (SMPC) enables collaborative computation without revealing private data, while the Hypergraph Partitioning Algorithm (HGPA) improves scalable and efficient data distribution. Furthermore, a Multiplex Adaptive Modality Fusion Graph Attention Network (MAMF-GAN) is integrated to optimize latency and throughput through intelligent data prediction and routing. Simulation results demonstrate that the proposed framework achieves a Packet Delivery Ratio (PDR) of 96%, privacy preservation of 99%, and approximately 95% resistance to impersonation and data manipulation attacks, outperforming existing DTN models. The framework is implemented using Python-based simulations. Future work will focus on real-time threat detection using advanced machine learning models, energy-aware consensus optimization, and validation through real-world DTN testbeds to enhance scalability and practical deployment.

This study addresses the industrial challenge that the temperature inside regenerative aluminum smelting furnaces cannot be directly or accurately measured. To overcome this issue, a TGAL hybrid model combining a Temporal Convolutional Network (TCN), Graph Convolutional Network (GCN), Multi-Head Attention mechanism, and Long Short-Term Memory (LSTM) network is proposed for multi-step accurate prediction of furnace temperature. The method first applies wavelet denoising to suppress noise in the industrial data collected by the SCADA system and employs the Pearson correlation coefficient to select highly correlated features, thereby improving the quality of the input data. The proposed TGAL model exploits the synergy of TCN in capturing long-term temporal dependencies, GCN in uncovering spatial correlations among variables, the attention mechanism in dynamically weighting features, and LSTM in temporal dynamic modeling. Validation on 44,640 one-minute data samples from actual production shows that, compared with traditional models, the proposed model achieves maximum improvements of 7.44% in RMSE, 24.85% in MAE, and 25.27% in MAPE for 2-step prediction, respectively. For 10-step prediction, the improvement rates remain at least 4.23% in RMSE, 6.91% in MAE, and 6.31% in MAPE. Moreover, Diebold–Mariano statistical tests confirm that the TGAL model's predictive accuracy is significantly superior to that of the comparison models. Nevertheless, the model performance under extreme operating conditions remains limited by data noise and nonlinear dynamics, and the physical mechanisms of the smelting process have yet to be incorporated. To address these limitations, future work will focus on dynamic coupling modeling and the embedding of physical information to further enhance the model's generalization capability and physical consistency.

Emotion recognition through multi-modal techniques uses multiple formats of input data in a significant area of research. However, existing studies face challenges related to the extraction of high-level emotional features and increased model complexity. Thus, this paper introduces Multi-modal data. Initially, text, video, and audio data are collected from BAUM 1 and Enterface05 datasets to classify emotions. Audio files undergo preprocessing with the Pass Gaussian Filter (PGF), while video clips are transformed into key frames using Spherical Interpolation based Q-learning (SIQ). The text data are preprocessed through tokenization and stemming. The features are extracted using a casual neural network for audio data, geometric feature calculations (GFC) for video data, and an improved term frequency-inverse document frequency (ITF-IDF) model for text data. The features are selected by the enhanced genetic gray lag goose optimization (EGG-LGO) method for audio data, the parrot optimization algorithm (POA) for video data, and the Adapted Firefly Optimization Algorithm (AFOA) for text data. The densely connected recurrent network with dual attention (D-RNA) model classifies emotions from audio data. Text emotions are classified using the self-attention based capsule-bi-directional long short term memory (SA-CBiLSTM) model, and emotions from visual data are classified using the gated attention enclosed residual context aware transformer (GRCAT). Finally, text, visual, and audio modality outputs are fused by a decision-level strategy to obtain the final output. The BAUM-1 dataset achieves accuracies of 99% for video, 99.3% for audio, and 98.4% for text data. The Enterface05 attains 98.18% accuracy for video, 98.75% for audio, and 97.7% for text.

Integrating deep learning with interpretable machine learning methods offers significant benefits for predicting mental health risks. This research introduces a hybrid architecture MH-XAI, aimed at precisely predicting mental health while providing transparent, feature-level explanations. MH-XAI integrates a multi-scale 1D Convolutional Neural Network (CNN), channel-wise attention mechanisms, an ensemble of (XGBoost) classifiers, and SHAP (SHapley Additive exPlanations) to achieve both local and global interpretability. The model is trained on a comprehensive dataset derived from Open Sourcing Mental Illness (OSMI) Mental Health in Tech surveys conducted between 2016 and 2023. This dataset encompasses a variety of demographic, psychological, and occupational characteristics. The input features are restructured into a format that facilitates deep convolutional learning. The CNN feature extractor utilizes parallel convolutional layers with three different kernel sizes, allowing the model to capture both short-range and long-range dependencies in tabular data. The multi-scale representations are subsequently enhanced using a channel-wise attention process. The acquired features are transmitted to an ensemble of XGBoost classifiers for enhanced prediction accuracy. MH-XAI attains a test accuracy of 91.54%, F1-score of 92%, precision of 92%, and recall of 91%, surpassing standalone CNN and XGBoost. SHAP elucidates the model's predictions by quantifying the contributions of each feature. Results underscore critical factors such as past mental health history, workplace culture, current mental health, and family history that affect outcomes. MH-XAI provides a precise, scalable, and comprehensible solution for the early detection of mental health in workplace environments.

The digitization of healthcare insurance claims faces persistent challenges including data breaches, fraudulent submissions, and inefficiencies in verification and settlement. This paper presents a Zero-Knowledge Succinct Non-Interactive Argument of Knowledge (Zk-SNARK) enabled blockchain framework deployed on the Polygon Proof of Stake (PoS) network for secure and privacy-preserving health insurance processing. The proposed architecture integrates Attribute-Based Encryption (ABE) for data confidentiality and the Elliptic Curve Digital Signature Algorithm (ECDSA) for authentication, ensuring end-to-end data integrity and access control. Experimental evaluation on the Polygon PoS testbed demonstrates a transaction cost of approximately $0.002, which is over 99% lower than Ethereum's 3–10 $ per transaction, while maintaining 100% resistance to data tampering, replay attacks, and transaction manipulation. Under the Polygon real network, the proposed framework supports a network-level transaction capacity of up to 7000 transactions per second (TPS) under nominal operating conditions, with an approximately 9.3% reduction in effective capacity under stress scenarios, while maintaining 100% verification accuracy for all Zk-SNARK proofs. The average on-chain verification and settlement latency was measured at 4.7 s, confirming the system's suitability for real-time healthcare claim settlement. These results validate that the proposed Zk-SNARK enabled Polygon PoS framework offers a scalable, cost-efficient, and cryptographically robust solution for healthcare insurance automation, outperforming existing blockchain implementations across security, efficiency, and economic performance metrics.

In Quantum chemical computation, numerical schemes such as the Hartree–Fock (HF) and density functional theory (DFT) are widely used to solve the Schrödinger equation numerically, to realize experiment-free prediction and analysis of key molecular properties such as structure and energy. Computing one-electron integrals, such as kinetic energy integrals and nuclear attraction integrals, is essential in both HF and DFT to characterize the molecular electronic states. However, as molecules of practical interest grow in size and angular momentum, computing one-electron orbitals becomes computationally expensive in most cases. Although computing kinetic energy integrals on CPUs is straightforward, bottlenecks in CPU-GPU data transfer have often been overlooked. In this study, we propose an efficient method to compute both the kinetic-energy and nuclear-attractive integrals on GPUs. First, we explicitly and symbolically expand recurrence relations based on the Obara–Saika and McMurchie–Davidson methods to eliminate redundant operations, thus improving computational efficiency. Second, we implemented a hybrid method that selects the best/fastest of both methods depending on the integration task. Third, we achieved further speedups by using CUDA streams to parallelize the execution of multiple kernels and efficiently utilize multiprocessor resources on the GPU. Computational experiments using NVIDIA A100 GPUs and Intel Xeon Gold 6338 CPU on relevant molecules of interest demonstrated the superiority of our one-electron integral GPU implementations, achieving a speedup of 20.2 times over PySCF, and a speedup of 132.6 times over GPU4PySCF.

The methods and procedures used to prevent the disclosure of private or sensitive information during data collection, processing, dissemination, or analysis are referred to as privacy preservation. K-prototypes have become a widely adopted clustering technique for mixed-type data in large-scale data mining tasks due to their efficiency and simplicity. However, as sensitive user information is frequently included in the underlying data, this approach presents privacy issues. Conventional privacy-preserving clustering techniques usually rely on a reliable third party to perform data preprocessing; however, it is frequently impractical to place total trust in such organizations. By implementing local differential privacy measures directly on user data, this study presents a revolutionary Two-tier Perturbation-based Local Differential Privacy Preservation K-prototyping (TP-LDPK) architecture that does away with the need for a trusted third party. Both numerical and categorical data are preprocessed using the Optimized Unary Encoding (OUE) algorithm and the Enhanced Min-Max Normalization (EMN) technique, respectively. The Improved chaotic map and the Obfuscation approach are used to disturb the normalized numerical data and encoded categorical data, respectively, in the first-tier perturbation. The Generalized Random Response (GRR) algorithm is used in the second tier to determine the new centroid set based on the information about the disturbed cluster. This technique protects sensitive data at every stage while enabling clustering through direct contact between the user and the server. Our approach produces high-quality clustering results under stringent local privacy restrictions, as demonstrated by both theoretical and empirical assessments that support its practical effectiveness and privacy guarantees. The TP-LDPK method yielded the lowest recall rate (0.494), accuracy rate (0.291), and F-measure (0.387).