Scientific Reports最新文献

Tactical Edge Networks (TENs) serve as critical infrastructure for disseminating time-sensitive intelligence under resource-constrained and hostile conditions such as Network-Centric Warfare (NCW) and the Internet of Battlefield Things (IoBT), where secure and efficient data sharing is a core requirement. To ensure security and privacy in such environments, strict adherence to the "need-to-know" principle is imperative, requiring that sensitive mission data are accessible only to entities with specific authorization attributes. Ciphertext-Policy Attribute-Based Encryption (CP-ABE) binds fine-grained access policies to ciphertexts and permits decryption only for attribute-satisfying users, rendering it inherently suitable for need-to-know control in these settings. However, the prohibitive computational overhead of bilinear pairings in CP-ABE is often impractical for lightweight frontline terminals in tactical edge networks. While outsourcing decryption to Tactical Cloud Nodes (TCNs) can alleviate this burden, it brings critical vulnerabilities in zero-trust deployments, including key exposure upon node capture, incorrect computation results, and the leakage of query intent to an honest-but-curious Command Center (CC). To address these issues, we present a novel resilient and verifiable outsourced attribute-based non-interactive oblivious transfer protocol. The proposed framework balances system efficiency with security and privacy, as well as addresses the inherent computational asymmetry between resource-constrained tactical edge devices and powerful cloud nodes. We integrate a receiver-privacy-only, NIOT-style index-hiding mechanism (RP-NIOT) into an offline/online encryption pipeline to conceal the user's query index from an honest-but-curious command center (CC). We do not claim classical OT sender privacy; unauthorized-record confidentiality is enforced by the CP-ABE/KEM-DEM layer. In addition, we incorporate a user-held blinding factor into the transformation keys to decouple the outsourcing capability from final decryption to ensure resilience against TCN compromise. A novel lightweight hash-based verification mechanism is designed to guarantee the correctness of outsourced computations. Detailed security and efficiency analysis show that the proposed protocol achieves resilience and data confidentiality as well as other security objectives at a cost of constant (policy-size independent) online terminal overhead-dominated by two exponentiations (plus one [Formula: see text] inversion), one MAC verification, two hash/KDF evaluations, and one symmetric decryption per query-making it suitable for latency-sensitive tactical applications.

The integration of NIST-standardized lattice-based cryptographic algorithms (ML-KEM and ML-DSA) into TLS 1.3 enables post-quantum secure authentication for Industrial IoT (IIoT) environments. This work implements these algorithms within the X.509 certificate infrastructure and evaluates their performance on resource-constrained IIoT hardware (Raspberry Pi 4). Experimental measurements covering key generation, encapsulation, decapsulation, and signature operations were obtained using a liboqs-enabled TLS 1.3 stack. The results show that PQ TLS 1.3 achieves comparable handshake latency to conventional TLS on IIoT-class gateways, with certificate size identified as the dominant overhead. These findings confirm the practicality of post-quantum authentication in IIoT systems while acknowledging limitations related to the hardware scope and simulated networking. Future work will extend validation to field deployments and additional PQC candidates.

Preeclampsia (PE) is a pregnancy-specific hypertensive disorder characterized by new-onset hypertension and proteinuria after 20 weeks of gestation. Endothelial dysfunction and abnormal placental vascular remodeling are central to its pathogenesis. Despite its significant impact on maternal and fetal health, current therapeutic options remain limited and largely symptomatic. This study aimed to investigate the protective effects and underlying mechanisms of ginsenoside Rb1 (Rb1) on endothelial function and vascular remodeling in PE, with a particular focus on the phosphoinositide 3-kinase/protein kinase B/endothelial nitric oxide synthase (PI3K/Akt/eNOS) signaling pathway. Plasma and placental samples from PE patients and normotensive pregnant women were analyzed for vascular markers and histological changes. A PE-like rat model was induced using NG-Nitro-L-arginine Methyl Ester, Hydrochloride (L-NAME) and treated with low, medium, or high doses of Rb1. Blood pressure, urinary protein excretion, nitric oxide (NO) and endothelin-1 (ET-1) levels, placental pathology, and related protein expression were evaluated. In vitro, an angiotensin II (Ang II)-induced human umbilical vein endothelial cell (HUVEC) injury model was used to assess the involvement of the PI3K/Akt/eNOS pathway using pharmacological inhibitors and activators. PE patients exhibited reduced NO levels, increased ET-1 levels, placental vascular damage, suppressed vascular endothelial growth factor (VEGF) expression, and elevated alpha-smooth muscle actin (α-SMA) expression. In the L-NAME-induced PE-like rat model, Rb1 treatment significantly reduced systolic blood pressure and urinary protein excretion, restored endothelial function, and alleviated placental structural damage. Rb1 also reversed inhibition of the PI3K/Akt/eNOS pathway observed in PE. In vitro, Rb1 improved HUVEC viability, angiogenesis, oxidative stress, and apoptosis, effects that were abolished by PI3K inhibition, while PI3K activation mimicked the protective effects of Rb1. Rb1 exerts significant protective effects against endothelial dysfunction and placental vascular remodeling in PE, likely through activation of the PI3K/Akt/eNOS signaling pathway. These findings suggest that Rb1 may represent a promising therapeutic candidate for the management of PE.

The evolution toward 6G wireless networks necessitates innovative solutions to support the massive Internet of Things (IoT) deployments with unprecedented computational and communication requirements, motivating this A comprehensive framework that integrates Reconfigurable Intelligent Surface (RIS) technology with hierarchical aerial computing networks by combining RIS-equipped Unmanned Aerial Vehicles (UAVs) operating as mobile edge computing nodes with High-Altitude Platforms (HAPs) to create a three-tier computing hierarchy addressing the limitations of conventional terrestrial infrastructure. The system model encompasses RIS-equipped UAVs serving terrestrial IoT devices with a single HAP providing high-capacity computational resources, where the RIS phase optimization is formulated as a Riemannian conjugate gradient problem on complex circle manifolds to maximize total system throughput while naturally handling unit modulus constraints through a three-stage sequential decomposition approach. Extensive Monte Carlo simulations demonstrate significant performance improvements over the comparable algorithm without RIS enhancement, with the RIS-enhanced system achieving 18% throughput improvement, near-linear scalability serving approximately 100% of available IoT devices compared to the algorithm In the comparable algorithm at 100 devices, a 95% task completion rate was maintained across all network loads versus 79-80% for the algorithm compared to the comparable algorithm. The results validate the potential of RIS-enabled aerial networks as a transformative solution for scalable and efficient 6G IoT services, with enhanced channel quality from intelligent phase configuration, enabling superior resource utilization and service provisioning in hierarchical computing architectures, establishing key contributions including novel RIS-aerial computing integration, advanced Riemannian manifold optimization with superior convergence properties, unified resource allocation combining stable matching theory with externality elimination, comprehensive performance analysis demonstrating practical viability, and real-world implementation considerations for future multi-UAV scenarios and energy-efficient designs.

Triple-negative breast cancer (TNBC) remains a significant clinical challenge due to its resistance to conventional treatments and the well-established role of epidermal growth factor receptor (EGFR) as a critical molecular target. This study explores the potential of Origanum majorana, a Mediterranean herb, as a source of natural EGFR inhibitors. The essential oil from Saudi Arabian O. majorana aerial parts was extracted and profiled using Gas Chromatography-Mass Spectrometry (GC-MS), identifying the major constituents. Predominant compounds were virtually screened against the EGFR tyrosine kinase domain via molecular docking and docking poses were validated. Subsequent molecular dynamics simulations together with MM-GBSA free energy analysis indicated that limonene formed the most stable complex with EGFR, exhibiting minimal structural deviation and consistent hydrophobic contacts. Density Functional Theory (DFT) analysis characterized limonene's electronic stability and hydrophobic nature, which underpin its binding mechanism. ADME predictions indicated limonene's favorable drug-likeness and blood-brain barrier permeability. This integrated approach identifies limonene as a promising natural scaffold for EGFR inhibition, warranting further experimental validation for its potential advancement as a treatment option for breast cancer.

The effectiveness of air-entraining agents in enhancing the frost resistance of concrete is significantly reduced under low-pressure conditions, such as those found in plateau and alpine regions, leading to severe freeze-thaw damage. To address this challenge, this study investigates the use of rubber powder as a compensatory material for air-entraining agents, introducing "solid pores" to replace traditional air voids. The combined effect of rubber powder and nano-silica was evaluated through macroscopic performance tests and microstructural analyses, focusing on the evolution of pore structure parameters and frost resistance during freeze-thaw cycles. The results show that rubber powder increases the air content and optimizes the pore structure, with "solid pores" accounting for an increasing proportion of total air content as the dosage rises. The addition of nano-silica further refines the pore size distribution by reducing the proportion of larger pores and stabilizing the bubble spacing coefficient. Concrete incorporating both rubber powder and nano-silica exhibits significantly improved frost resistance, with only a slight reduction in compressive strength compared to ordinary concrete. These findings demonstrate that the synergistic use of rubber powder and nano-silica effectively compensates for the diminished performance of air-entraining agents under low-pressure conditions, offering a practical approach to enhancing the freeze-thaw durability of concrete in cold, high-altitude environments.

Cardiovascular diseases are the leading cause of death worldwide. With electrocardiogram (ECG) machines becoming more accessible, passive monitoring for arrhythmia detection is now possible. This work highlights the importance of self-supervised learning in detecting arrhythmias by leveraging large-scale unlabelled ECG data to improve performance and reduce overfitting to class imbalance and noise. We propose Masked Patch Modelling (MPM) and use 8.2 million unlabelled ECGs for self-supervised pre-training, introducing PatchECG, a 1D Transformer model that can be fine-tuned for various ECG tasks. PatchECG achieves state-of-the-art results on standard datasets, including PTB-XL multi-label classification, and sets new benchmarks on the largest and highest-quality multi-label dataset to date. Compared to existing methods, PatchECG is five times more computationally efficient while increasing model capacity by a factor of 14. We also compare the 1D PatchECG model to a state-of-the-art 2D vision Transformer, HeartBEiT, and observe significantly higher performance. Finally, ablation studies reveal a 2% performance improvement in handling class imbalance, label noise, and over-parameterization. These findings demonstrate the potential of self-supervised learning in advancing automated arrhythmia detection.