Search Results (704)

Federated Learning (FL) has been widely adopted in privacy-sensitive and distributed environments. However, training stability becomes significantly challenged when differential privacy (DP) noise and Byzantine client behaviors coexist, as these heterogeneous perturbations jointly introduce time-varying distortions to model updates. Existing approaches typically address privacy and robustness in isolation. Under DP constraints, noise injection increases gradient variance and obscures the distinction between benign and adversarial updates, causing many robust aggregation methods to misclassify normal clients or fail to detect malicious ones. As a result, their effectiveness degrades substantially in practical IoT environments where noise and attacks interact. In this work, we propose a dual-factor adaptive and robust aggregation framework (DARA) to improve the stability of FL under such combined disturbances. DARA adjusts the differential privacy noise scale by jointly considering local update magnitudes and training-round dynamics, aiming to mitigate noise-induced bias under a fixed privacy budget. Meanwhile, a direction-aware weighted aggregation scheme assigns continuous trust weights based on cosine similarity between updates, thereby suppressing the influence of potentially anomalous or adversarial clients. We conduct extensive experiments on multiple benchmark datasets to evaluate DARA under differential privacy constraints and Byzantine attack scenarios. The results indicate that DARA achieves favorable robustness and convergence behavior compared with representative aggregation baselines, while maintaining competitive model accuracy. Full article

Resource-constrained IoT systems face a fundamental conflict between cryptographic security and energy efficiency, particularly in critical infrastructure monitoring requiring long-term autonomous operation. This study presents a hybrid framework integrating signal-adaptive compression with hardware-accelerated authenticated encryption to resolve this trade-off. The Dynamic Payload Compression with Selective Encryption framework classifies sensor data into three SNR regimes and applies adaptive compression strategies: 24.15-fold compression for low-SNR backgrounds, 1.77-fold for transitional states, and no compression for high-SNR leak detection events. Experimental validation using 2714 acoustic sensor samples demonstrates 5.91-fold average payload reduction with 100% detection accuracy. The integration with STM32L5 hardware AES acceleration reduces power–data correlation from 0.820 to 0.041, increasing differential power analysis attack complexity from 500 to over 221,000 required traces. Compression-induced timing variance provides additional side-channel masking, burying cryptographic signals beneath a 0.00009 signal-to-noise ratio. Projected on 19,200 mAh lithium thionyl chloride batteries, the system achieves 14-year operational lifetime under realistic duty cycles, exceeding industrial requirements for critical infrastructure protection while maintaining robust security against physical attacks. Full article

The convergence of Large Language Models (LLMs) and Internet of Things (IoT) systems has created a new class of intelligent applications across healthcare, industrial automation, smart cities, and connected homes. However, this integration introduces a complex and largely underexplored security landscape. LLMs deployed in IoT contexts face threats spanning both the AI and embedded systems domains, including prompt injection through sensor-driven inputs, model extraction from edge devices, data poisoning of IoT data streams, and privacy leakage through LLM-generated responses grounded in personal data. Simultaneously, LLMs are proving to be powerful tools for IoT security, with LLM-based intrusion detection systems achieving 95–99% accuracy on standard IoT datasets and LLM-driven threat intelligence outperforming traditional machine learning by significant margins. We systematically review 88 papers from IEEE, ACM, MDPI, and arXiv (2020–2025), providing: (1) a structured taxonomy of security threats targeting LLM-IoT systems, (2) a review of LLMs as security enablers for IoT, (3) an evaluation of privacy-preserving architectures including federated learning, differential privacy, homomorphic encryption, and trusted execution environments, (4) domain-specific security analysis across healthcare, industrial, smart home, smart grid, and vehicular IoT, and (5) a literature-based comparative analysis of LLM-based security systems. A central finding is the accuracy–efficiency–privacy trilemma: the model compression techniques needed to deploy LLMs on resource-constrained IoT devices can degrade security and even introduce new vulnerabilities. Our analysis provides researchers and practitioners with a structured understanding of both the risks and opportunities at the frontier of LLM-IoT security. Full article

This study presents ZK-EHR, a decentralized access control framework designed to enable secure and privacy-preserving sharing of encrypted electronic health records across institutional boundaries. Unlike existing blockchain-based EHR access control systems that expose user identities on-chain or lack cryptographic privacy guarantees, ZK-EHR decouples authorization from identity disclosure by integrating zk-SNARK-based proofs with blockchain smart contracts to verify policy compliance without revealing user roles, affiliations, or credentials. The framework employs three differentiated actor roles—Patient (Data Owner), Doctor (Care Provider), and Researcher (Authorized Analyst)—with distinct policy-driven access workflows, a custom Groth16 zero-knowledge circuit for role-based constraint enforcement, and a modular architecture combining on-chain verification with off-chain encrypted storage via IPFS. Concrete design proposals for access revocation and replay attack prevention are introduced to address operational security requirements. The system was evaluated under multiple operational and adversarial scenarios. Experimental results indicate consistent on-chain verification latency (approximately 390 ms), reliable rejection of tampered submissions, and per-verification gas consumption of 216,631 gas. A comparative analysis against representative baseline systems demonstrates that ZK-EHR uniquely combines identity anonymity, on-chain cryptographic policy enforcement, and auditable encrypted record retrieval. These findings establish the feasibility of zk-SNARK-based access control for decentralized, verifiable, and privacy-aware EHR management. Full article

Muscle biopsy (MB) is an important tool to help differentiate idiopathic inflammatory myopathies (IIMs) from hereditary muscular diseases (HMDs). The usefulness of immunohistochemical stains of the major histocompatibility complex class I and the membrane attack complex are controversial, as both may be identified in some HMDs. More sensitive markers of IIMs have recently been used, such as myxovirus resistance A (MxA), a type I interferon-inducible protein. We selected skeletal MB samples from 81 patients diagnosed with IIM and HMD harbouring overt inflammatory infiltrates on their MBs in the period between March 2022 and September 2024. Two groups were identified: the IIM group (46 cases) and the HMD group (35 cases). We characterized and compared the patterns of MxA protein expression among the two groups. In the IIM group, positive sarcoplasmic MxA expression was detected on the myofibres of 10 patients (24%), among whom were eight dermatomyositis patients. In the HMD group, we did not identify any sarcoplasmic positivity. However, five patients (14%) showed positive labelling restricted to the sarcolemmal membrane, including non-necrotic or regenerating fibres. Our study demonstrates the value of MxA for increasing dermatomyositis diagnostic accuracy and suggests the potential role of interferon type I in the pathophysiology of HMD. Full article

Unmanned Aerial Vehicle (UAV) swarms operating in contested environments face a critical “semantic gap” between raw, high-velocity network traffic and high-level mission security constraints, compounded by the risk of privacy leakage during collaborative learning. Existing deep learning (DL)-based Network Intrusion Detection Systems (NIDSs) suffer from opacity, prohibitive resource consumption, and vulnerability to gradient leakage attacks in federated settings, while traditional rule-based systems fail to handle encrypted payloads and evolving attack patterns. To bridge this gap, we present NeSySwarm-IDS (Neuro-Symbolic Swarm Intrusion Detection System), an end-to-end differentiable neuro-symbolic framework that simultaneously achieves high accuracy, strong privacy guarantees, and built-in interpretability under resource constraints. NeSySwarm-IDS integrates an extremely lightweight 1D convolutional neural network with a differentiable Łukasiewicz fuzzy logic reasoner incorporating attack-specific rules. By aggregating only low-dimensional logic rule weights with calibrated differential privacy noise, we drastically reduce communication overhead while providing $(ϵ,\delta )$-DP guarantees with negligible utility loss. Extensive experiments on the UAV-NIDD dataset and our self-collected dataset demonstrate that NeSySwarm-IDS achieves near-perfect detection accuracy, significantly outperforming traditional machine learning baselines despite using limited training data. A detailed case study on GPS spoofing confirms the interpretability of our approach, providing axiomatic explanations suitable for autonomous mission verification. These results establish that end-to-end neuro-symbolic learning can effectively bridge the semantic gap in UAV swarm security while ensuring privacy and interpretability, offering a practical pathway for deploying trustworthy AI in contested environments. Full article

To address the high redundancy and weak security inherent in satellite image transmission, this paper proposes an image encryption algorithm founded on a novel five-dimensional fractional-order cosine memristive Hopfield neural network (5D-FOCMHNN). The constructed hyperchaotic system exhibits long-term memory and multistability, capable of generating reconfigurable multi-scroll attractors. A multivariate bit-level scrambling strategy effectively disrupts pixel correlations using neuron state sequences. Furthermore, the system’s chaotic output dynamically governs DNA encoding rules, while a bidirectional diffusion mechanism ensures strong randomization and resistance to differential attacks. Comprehensive experiments demonstrate that the 5D-FOCMHNN-based scheme provides a key space of $2^{256}$, has an information entropy approaching the ideal value of 8, and exhibits robust resilience against cropping, noise, and statistical cryptanalysis, thereby providing a highly secure solution for satellite image transmission. Full article

Releasing structured microdata requires balancing utility and privacy under group-based disclosure risks. We propose AE-LRHMA, a hybrid anonymization framework that performs Mondrian-style hierarchical partitioning in an autoencoder-learned latent space and integrates local (k,e)-microaggregation. To explicitly control sensitive-value concentration and diversity within each equivalence class, we introduce a tunable constraint set consisting of k, a maximum sensitive proportion threshold, and an optional sensitive-entropy threshold (used as a hard gate when enabled and otherwise as a soft term in split scoring). The anonymized output is generated via standard interval/set generalization in the original space. Experiments on Adult and Bank Marketing demonstrate that AE-LRHMA yields lower information loss and more stable group structures than representative baselines under comparable settings. We further report linkage-attack-oriented risk metrics to empirically characterize relative disclosure trends without claiming formal guarantees, such as differential privacy. Full article

Low Earth Orbit (LEO) mega-constellations are becoming the backbone of global communications, yet their cybersecurity remains critically under-addressed. Intrusion detection systems (IDSs) for such constellations face a unique trilemma of accuracy, efficiency, and interpretability under extreme SWaP-C (size, weight, power, and cost) constraints. We present FedLTN-CubeSat (FedLTN refers to Federated Logic Tensor Networks), a neuro-symbolic federated learning framework for intrusion detection in LEO CubeSat constellations. The framework first employs a lightweight spatio-temporal separable perception encoder to efficiently extract features from telemetry and IQ data, designed to operate within the computational budgets of resource-constrained on-board processors. These features feed into a differentiable first-order logic layer based on Logic Tensor Networks, which incorporates domain knowledge as logical axioms to guide learning and enhance interpretability. To enable collaborative learning across a constellation, FedLTN-CubeSat introduces an intra-orbit symbolic federated learning mechanism that aggregates only the logic-layer parameters via inter-satellite links, drastically reducing communication overhead while preserving data privacy. Furthermore, an orbit-adaptive predicate migration module transfers learned rules across different orbital configurations with minimal supervision, facilitating rapid deployment. We evaluate on two benchmarks: the CuCD-ID dataset (NASA NOS3 telemetry) and the STIN dataset (satellite-terrestrial integrated networks). FedLTN-CubeSat achieves 0.98 F1-score on CuCD-ID and 0.96 accuracy on STIN—significantly outperforming prior federated learning baselines (7% improvement) while incurring a minimal daily communication load per satellite. The framework also outputs interpretable decision traces grounded in logical axioms, enabling operators to understand and validate detections. Logical constraints improve detection of unseen attack variants by 25% over pure neural baselines. Full article

This paper proposes an impulsive SIRQ model for the analysis of computer network resilience against malware propagation and distributed denial-of-service (DDoS) attacks. The model extends classical epidemic frameworks by combining the continuous-time dynamics of malicious object spreading with discrete control actions corresponding to mass updates, node isolation, and access control policies. A qualitative analysis of the resulting system of impulsive differential equations is performed. The basic reproduction number ${\mathcal{R}}_0$, identified as a threshold parameter characterizing the intensity of attack propagation, and sufficient conditions for the global asymptotic stability of the infection-free state are established. It is shown that, under periodic impulsive control, the infection-free state can be stabilized with respect to the target population coordinates even when ${\mathcal{R}}_0>1$. An exponential decay estimate for the total active threat is derived, guaranteeing the asymptotic extinction of the infected and quarantined node populations. The proposed approach provides quantitative criteria for the effectiveness of impulsive cyber defense strategies and offers a theoretical foundation for the design of adaptive multi-layer protection systems for critical information infrastructures. Practical interpretation of the results illustrates the dependence of the critical impulsive control period on the model parameters and demonstrates the applicability of the approach to cybersecurity strategy design. Full article

Aggressive behavior in Muscovy ducks (Cairna moschata) has become a predominant concern in intensive farming systems, leading to reduced animal welfare and production losses. To unravel the molecular mechanisms underlying this behavior, transcriptomic profiling was performed on the hypothalamus, a key regulatory hub for aggressive responses. A total of 120 healthy 60-day-old female Muscovy ducks were continuously monitored for 24 h/day over one month using Media Recorder 2.0 software. Based on instantaneous and continuous behavioral observations, the ducks were categorized into three groups: aggressor (Experimental group I, actively attacking conspecifics), victim (Experimental group II, receiving aggression), and non-aggressive (Control group, no aggressive interactions). Hypothalamic tissues were collected from each group (n = 4 per group) for Illumina HiSeq 2000 high-throughput transcriptome sequencing. Functional annotation and enrichment analysis of differentially expressed genes (DEGs) were performed using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases, followed by quantitative real-time PCR (qRT-PCR) validation. GO analysis identified 626 DEGs in the aggressor group and 649 DEGs in the victim group compared to the control group, with 26 DEGs directly involved in aggressive behavior regulation. Integration of GO and KEGG annotations revealed 69 candidate genes associated with aggressive behavior, enriched in two GO terms (behavior [GO:0007610] and sensory perception of pain [GO:0019233]) and the ERBB signaling pathway (map04012). qRT-PCR validation of 14 randomly selected candidate genes (e.g., NPY, ERBB4, MAPK9, PRDM12) confirmed that their expression patterns were consistent with transcriptomic data, verifying the reliability of the sequencing results. These findings provide novel insights into the molecular genetic basis of aggressive behavior in Muscovy ducks and lay a foundation for developing targeted strategies to mitigate aggression in intensive farming systems. Full article

Myocardial ischemia/reperfusion (MI/R) injury affects heart attack outcomes. Endothelial cells dysfunction immediately after MI/R, but the key molecules and how to block them remain unclear. We combined single-cell atlas analysis, AI simulation, and experimental single-cell RNA sequencing data from mouse MI/R; we did quality control, cell annotation, hdWGCNA, and differential gene screening to identify endothelial genes. We constructed a protein network with STRING, predicted structure with AlphaFold3, and used AutoDock for molecular docking to find potential drugs. Virtual knockout simulations were used to check gene deletion effects. The compound andrographolide (AG) was tested in in vitro and in vivo MI/R models by measuring cell viability, inflammation, pathway activity, infarct size, and cardiac function. Single-cell analysis showed that S100 calcium binding protein A8 (S100A8) is an important element in vascular inflammation. It promotes inflammation by interacting indirectly with Cluster of differentiation 14 (CD14). Molecular docking showed that AG binds stably to S100A8. In vitro, AG reduced endothelial injury and blocked the IL-17 pathway. In vivo, AG reduced infarct size, improved cardiac function, and lowered S100A8 and IL-17 pathway proteins. Using single-cell analysis, AI, and experiments, we showed that S100A8 is related to MI/R injury. Andrographolide protects microvasculature via the S100A8 pathway, offering a promising treatment approach and new insights into heart injury mechanisms. Full article

Low-rate Denial-of-Service (LDoS) attacks exploit periodic traffic pulses to trigger congestion while maintaining a low average rate, making them highly stealthy and difficult to distinguish from legitimate bursty traffic using threshold-based or simple statistical detectors. To address this challenge, this paper proposes DELP-Net, an end-to-end Differentiable Entropy Layer Pyramid Network for window-level online LDoS detection directly from raw traffic. DELP-Net combines a multi-scale one-dimensional convolutional pyramid with a differentiable Rényi-entropy-driven attention mechanism to capture distributional regularity and weak repetitive patterns characteristic of LDoS traffic. In addition, an entropy-conditioned temporal convolutional network is employed to model cross-window periodic dependencies in a lightweight manner, together with an entropy-regularized hybrid loss to enhance robustness under complex background traffic. Experiments on the low-rate DoS dataset show that DELP-Net achieves an average F1 score of 0.9877 across six LDoS attack types, with a detection rate of 98.69% and a false-positive rate of 1.15%, demonstrating its effectiveness and suitability for practical online intrusion detection deployments. Full article

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

In contemporary communication systems, digital images occupy an irreplaceable role; however, the privacy-related risks attendant to their prevalent application have grown increasingly salient. This paper presents an image encryption scheme integrating a novel two-dimensional Ackley-Sine chaotic map (2D-ASM) with dynamic DNA operations. First, a two-dimensional Ackley-Sine chaotic map, constructed based on the Ackley function and sine function, is designed and validated through a series of chaotic indicators. Results demonstrate that 2D-ASM exhibits superior chaotic properties compared to several existing state-of-the-art chaotic maps, with its maximum Lyapunov exponent (LE) exceeding 23, Permutation Entropy (PE) close to 1 in the full parameter range, and correlation dimension (CD) significantly higher than comparative chaotic systems. The proposed 2D-ASM-based image encryption scheme leverages the SHA-256 hash value of the plaintext image and four external keys to jointly generate the initial conditions and parameters of the 2D-ASM chaotic system, thereby ensuring a sufficiently large key space of 2²⁵⁶. Subsequently, chaotic sequences generated by 2D-ASM are employed to permute and diffuse the plaintext image, followed by dynamic DNA coding, operations, and decoding to obtain the encrypted image. Security analyses and comparisons with several existing representative algorithms confirm that the proposed encryption scheme achieves excellent encryption performance: the Number of Pixels Change Rate (NPCR) is above 99.6%, the Unified Average Changing Intensity (UACI) approaches 33.4%, and the information entropy of ciphertext images reaches 7.999 or higher. The scheme can effectively resist various potential attacks, including statistical and differential attacks, and outperforms representative algorithms in pixel correlation reduction and anti-interference performance. Full article