Split Learning (SL) enables collaborative model training across resource-constrained Internet of Things (IoT) devices while preserving data privacy, making it ideal for applications like smart homes and intrusion detection. However, SL systems are vulnerable to poisoning attacks, where malicious clients manipulate activations to compromise model integrity. This paper proposes a novel detection framework for poisoning attacks in quantized SL systems, leveraging pruned weights to enhance efficiency and an autoencoder-based method to identify malicious clients. Our approach uses intraclass-distance inflated loss to detect anomalies in quantized activations, tailored for IoT environments with device churn and data heterogeneity. Experimental results on the Bot-IoT dataset demonstrate high detection accuracy, with the autoencoder method achieving up to 80% true positive rates in random partitioning scenarios and maintaining model accuracy above 85% under attack. By integrating pruned weights and quantization, our framework reduces computational overhead by approximately 40% and communication costs by 80%, making it suitable for resource-constrained IoT devices. The approach aligns with Trustworthy AI principles by ensuring robustness, preserving privacy, and promoting fairness, offering a scalable solution for secure and ethical AI in IoT ecosystems.

The rapid advancement of Internet of Medical Things (IoMT) technologies and the growing demand for ubiquitous healthcare services have created an urgent need for intelligent networking solutions that can seamlessly integrate space, air, and ground communication infrastructures. This paper presents a novel intent-based networking framework designed for IoMT communications within space-air-ground integrated systems. The proposed framework addresses the complex requirements of healthcare applications by introducing an intelligent intent understanding and translation mechanism that can automatically configure network resources based on high-level medical service requirements. Our approach incorporates four core components: medical intent construction for IoMT scenarios, intelligent intent classification using enhanced BERT-CNN models, sophisticated intent parsing through GlobalPointer-based entity extraction, and dynamic intent translation for real-time network policy generation. Experimental results demonstrate that our framework achieves 94.37% accuracy in intent classification and 83.76% F1-score in entity extraction. Experimental validation across space-air-ground network simulations demonstrates substantial improvements in resource utilization efficiency (23.4% increase), bandwidth allocation optimization, and latency reduction (18.6% improvement), directly enhancing patient care capabilities and clinical decision-making reliability in distributed healthcare environments.

The proliferation of large-scale Internet-of-things (IoT) deployments and the emergence of 6G wireless technologies have created a pressing need for intelligent, energy-aware, and low-latency communication frameworks. In this work, we propose a novel two-phase reinforcement learning (RL)-based architecture designed to minimize the age of information (AoI) in 6G-enabled IoT networks. Our approach integrates (i) a deep deterministic policy gradient (DDPG)-driven backscatter-assisted cognitive radio non-orthogonal multiple access (CR-NOMA) scheme in the uplink, and (ii) a lightweight Q-learning-based power-domain NOMA (PD-NOMA) strategy for the downlink. In the uplink, energy harvesting (EH) sensors employ deep RL to jointly optimize backscatter reflection coefficients and transmission scheduling over shared spectrum using CR-NOMA. This enables energy-efficient communication and reduced AoI under dynamic energy and channel conditions. In the downlink, the edge node serves multiple IoT users simultaneously using PD-NOMA, where a Q-learning agent intelligently decides whether to transmit fresh or cached data to each user based on battery levels, channel quality, and information freshness. Both phases are modeled as Markov decision processes (MDPs), allowing agents to learn independently and converge toward optimal policies that balance information freshness, spectral efficiency (SE), and energy constraints. Extensive simulations demonstrate that the proposed framework effectively reduces AoI across both phases, with consistent convergence even under varying sensor densities and EH conditions. Moreover, by relying on explainable and verifiable learning mechanisms, our model addresses emerging concerns around reliability and trustworthiness in artificial intelligence (AI)-driven 6G-IoT systems. This framework represents a step toward scalable, adaptive, and responsible AI integration for future mission-critical IoT applications.

Sixth Generation (6G) wireless networks with ultra-low latency, high reliability, and massive connectivity require intelligent and privacy-concerned infrastructure optimization. This work presents a federated autoencoder platform combined with Explainable AI (XAI) for performance optimization of 6G-IoT systems. The method integrates traditional machine learning algorithms (Decision Tree, Random Forest, Logistic Regression, AdaBoost, Gradient Boosting) with a Variational Autoencoder (VAE) for dimensionality reduction and feature extraction. Federated Learning (FL) is utilized to maintain data privacy among distributed edge nodes, and SHAP and LIME explainers are utilized for explaining model decisions at the local and global levels. The framework points out key QoS parameters like latency and throughput as major optimization levers. Experimental outcomes on the 6G-IoT dataset indicate that Random Forest with highest accuracy for 80:20 split and Gradient Boosting has a 99.8% accuracy in a 10-fold validation, and FL gets a ROC-AUC value of 0.999 with robust privacy guarantees. XAI enhances transparency and regulatory compliance by making attribution of predictions to contributing features. As a whole, the proposed approach provides an interpretable, privacy-conscientious, and scalable tool for intelligent 6G-IoT infrastructure management.

Architectural Distortion (AD) is a common abnormality in digital mammograms, alongside masses and microcalcifications. Detecting AD in dense breast tissue is particularly challenging due to its heterogeneous asymmetries and subtle presentation. Factors such as location, size, shape, texture, and variability in patterns contribute to reduced sensitivity. To address these challenges, we propose a novel feature fusion-based Vision Transformer (ViT) attention network, combined with VGG-16, to improve accuracy and efficiency in AD detection. Our approach mitigates issues related to texture fixation, background boundaries, and deep neural network limitations, enhancing the robustness of AD classification in mammograms. Experimental results demonstrate that the proposed model achieves state-of-the-art performance, outperforming eight existing deep learning models. On the PINUM dataset, it attains 0.97 sensitivity, 0.92 F1-score, 0.93 precision, 0.94 specificity, and 0.96 accuracy. On the DDSM dataset, it records 0.93 sensitivity, 0.91 F1-score, 0.94 precision, 0.92 specificity, and 0.95 accuracy. These results highlight the potential of our method for computer-aided breast cancer diagnosis, particularly in low-resource settings where access to high-end imaging technology is limited. By enabling more accurate and timely AD detection, our approach could significantly improve breast cancer screening and early intervention worldwide.

Smart transportation networks face increasing demands for efficiency and sustainability. This study presents a reinforcement learning approach that optimizes eco-driving strategies for connected and automated vehicles (CAVs) in urban environments, with a particular application to healthcare logistics. Specifically, we propose a novel approach using reinforcement learning, specifically a twin delayed deep deterministic policy gradient (TD3) algorithm, to dynamically optimize CAV trajectories at signalized intersections. The proposed healthcare eco-driving trajectory optimization (TD3-HETO) model incorporates real-time traffic conditions, signal timing information, and healthcare urgency levels to generate optimal acceleration profiles. The reward function is designed to balance energy efficiency, traffic flow, safety, comfort, and healthcare delivery timeliness. Additionally, the model introduces a dynamic exploration strategy that adapts to healthcare task urgency, enabling efficient balancing between energy consumption and delivery timelines. Experimental results show that TD3-HETO reduces energy consumption by up to 28.7% compared to baseline methods while improving average speeds by 3.7% for urgent healthcare deliveries. The model achieves superior safety performance with 98.7% of time steps showing zero conflicts, compared to 95.3% for the best baseline. TD3-HETO also demonstrates remarkable adaptability to varying traffic demands and signal timings, maintaining consistent performance even at high traffic volumes. This research contributes to developing intelligent transportation systems to enhance environmental sustainability and healthcare accessibility in smart cities, potentially improving patient outcomes and operational efficiency in urban healthcare logistics.

Next-generation communication networks are revolutionizing digital publishing through intelligent content distribution and collaborative optimization capabilities. However, existing federated learning approaches face fundamental limitations, including trusted third-party dependencies, excessive communication overhead, and vulnerability to collusion attacks between servers and participants. This paper introduces VHFL-DP, a verifiable homomorphic federated learning framework for digital publishing environments operating within 6G network infrastructures. The framework addresses critical privacy and scalability challenges through four key innovations: a distributed cryptographic key generation protocol that eliminates trusted third-party requirements, Chinese remainder theorem-based dimensionality reduction, auxiliary validation nodes that enable independent verification with constant-time complexity, and an intelligent incentive mechanism that rewards digital publishing platforms based on objective contribution quality metrics. Experimental evaluation on MNIST and Amazon reviews datasets across six baseline methods demonstrates that VHFL-DP achieves superior performance with accuracy improvements of 4.2% over the best baseline method. The framework maintains constant verification time ranging from 2.73 to 2.91 seconds regardless of platform count, increasing from ten to fifty, or dropout rates reaching thirty percent. Security evaluation reveals strong resilience with only 2.4 percentage point accuracy degradation under poisoning attacks compared to 6.7-7.0 points for baseline method, inference attack success near random guessing at 51.3%, and 92.4% successful aggregation under Byzantine adversaries.

The optimization of modern healthcare facilities presents unique challenges at the intersection of medical service quality, energy efficiency, and environmental impact. By integrating carbon-intelligent Industrial Internet of Things (IIoT) technologies with healthcare operations, our approach enables real-time monitoring and optimization of carbon emissions while maintaining medical service quality. Specifically, this paper presents a novel deep reinforcement learning-based multi-objective optimization algorithm (HC-DMOPSO) for IIoT-enabled healthcare building management. By integrating healthcare-specific constraints with an enhanced swarm intelligence framework, our approach optimizes building operations while considering medical device power demands, patient comfort, and environmental requirements. The proposed algorithm combines dual-distance metrics -population average distance and crowding distance -with deep Q-networks to effectively explore the complex solution space. Experimental results demonstrate HC-DMOPSO’s superior performance across multiple metrics: The integration of carbon intelligent IIoT sensors and actuators enables HC-DMOPSO to achieve 24.8% reduction in energy consumption while maintaining 99.92% medical power reliability, 32.5% decrease in peak load with only 0.38∘C average temperature deviation, and 28.7% improvement in carbon reduction compared to baseline methods.

Intelligent Transportation security requires cooperative credentials for sharing navigation and communication data between the vehicles. However due to the dynamic environment, communication is interrupted by the adversaries, resulting in non-privacy issues. This article introduces an Agreement-induced Data Verification Model (ADVM) for securing vehicular communication against adversaries. The connected vehicles in a grid communicate with each other based on direct and indirect recommendation. This recommendation is based on mutual identity sharing between the vehicles for masked information exchange. Non-replicated and recommendation based verifications are performed using the vector classification learning. In this learning process, the credential validity and communication tolerance amid adversaries are augmented. The constraint-failing vehicles are disconnected from the communication grid, preventing its insecure impact over the communication. The proposed model’s performance is verified using false rate, success ratio, processing time, complexity, and recommendation ratio. For the different vehicles, the proposed model achieves 9.69% less false rate, 10.3% success ratio, 10.49% less processing time, 10.3% less complexity, and 12.87% high recommendation ratio.