The routing protocol for low-power and lossy networks (RPL) is a specialized routing protocol designed for optimized data routing, specifically for resource-constrained Internet of Things (IoT) networks with unreliable links and high packet loss. However, RPL is highly vulnerable to significant security challenges, particularly the decrease rank attack (DRA), in which malicious nodes attract child nodes by falsely advertising lower ranks, leading to routing inefficiencies, unnecessary retransmissions, and increased energy consumption. To address this problem, we propose a novel intent-based networking-driven centralized real-time reinforced detection scheme (CRRDS), which translates high-level security intents into policy-driven automated control strategies for DRA detection. In the proposed CRRDS, a resource-rich root node acts as a deep reinforcement learning agent that collects critical information from the child nodes, including the node ID, end-to-end delay, received signal strength indicator, and hop count, to detect suspicious behavior accurately and intelligently. Initially, we implemented a deep Q-network (DQN)-assisted CRRDS in detecting DRA. Subsequently, we utilized double DQN (DDQN) and dueling DDQN due to their enhanced capabilities in value estimation and policy learning. The dueling DDQN performed optimally because of its deeper architecture. Simulation results demonstrate that the proposed dueling DDQN-assisted CRRDS achieves the highest detection accuracy of 98% with notable gains in true positive and false positive rates, even in complex scenarios with up to 30% malicious nodes.

Large-scale deep learning models rely on wireless networks for distributed training approaches, which are essential to meet the immense computational and data demands. However, the stochastic nature of wireless environments introduces significant challenges such as variable delays, noise interference, and packet loss, which lead to degraded gradient synchronization and hinder model convergence. In this work, we propose a novel communication-aware distributed training (CADT) framework that integrates real-time channel state information (CSI) feedback into the gradient aggregation process. Unlike conventional methods that assume static or ideal communication conditions, CADT dynamically reweights gradients from each node based on instantaneous channel quality, enabling robust aggregation under adverse wireless conditions. By dynamically adjusting the contribution of each node based on instantaneous channel conditions, CADT effectively compensates for wireless impairments, thereby ensuring more reliable gradient aggregation and significantly improving both convergence speed and final model accuracy. Extensive experiments on CIFAR-10, CIFAR-100, ImageNet, and SVHN using Vision Transformer and ResNet-50 demonstrate that CADT outperforms baseline methods in terms of convergence, accuracy, and communication efficiency. In addition, we provide a rigorous theoretical analysis that establishes convergence guarantees under realistic wireless conditions, thereby advancing the theoretical foundation of distributed optimization in non-ideal communication environments.Our framework offers a practical solution for real-world scenarios such as edge computing, where communication constraints and environmental variability are dominant factors.

Social networks, particularly Twitter, significantly influence public emotions during health crises, often amplifying distress and misinformation. Effective sentiment analysis is crucial for mitigating social unrest and enabling timely interventions. This study introduces a novel two-stage framework, the fuzzy-integrated case-based and adaptive deep-belief neural network (2-SCBAADBNN) model, for real-time sentiment classification in public health emergencies. The framework integrates neural architecture search (NAS) and large machine learning models (LMMs) such as bidirectional encoder representations from transformers (BERT) to optimize fuzzy logic components and enhance feature extraction, improving sentiment detection accuracy. The two-stage classification first distinguishes between personal and news-related tweets using a fuzzy clue-based method, followed by sentiment classification of personal tweets into positive or negative categories. By combining fuzzy logic with deep learning, this multimodal approach aligns with advancements in emotion AI, offering greater scalability, adaptability, and precision in sentiment analysis. Comparative evaluations show that 2-SCBAADBNN outperforms existing models, providing a robust solution to monitor emotional distress and combat misinformation during crises. This research advances emotion AI by integrating NAS and LMM, allowing more context-sensitive real-time sentiment analysis. It contributes to developing AI-driven empathy-based systems capable of understanding and responding to public sentiment more effectively in critical social scenarios.

The Internet of Things (IoT) has grown rapidly, and nowadays, it is exploited by cyber attacks on IoT devices. An accurate system to identify malicious attacks on the IoT environment has become very important for minimizing security risks on IoT devices. Botnet attacks are among the most serious and widespread attacks, and they threaten IoT devices. Motionless IoT devices have a security weakness due to lack of sufficient memory and computation results for a security platform. In addition, numerous existing systems present themselves for finding unknown patterns from IoT networks to improve security. In this study, hybrid deep learning, a convolutional neural network and long short-term memory (CNN-LSTM) algorithm, was proposed to detect botnet attacks, namely, BASHLITE and Mirai, on nine commercial IoT devices. Extensive empirical research was performed by employing a real N-BaIoT dataset extracted from a real system, including benign and malicious patterns. The experimental results exposed the superiority of the CNN-LSTM model with accuracies of 90.88% and 88.61% in detecting botnet attacks from doorbells (Danminin and Ennio brands), whereas the proposed system achieved good accuracy (88.53%) in identifying botnet attacks from thermostat devices. The accuracies of the proposed system in detecting botnet attacks from security cameras were 87.19%, 89.23%, 87.76%, and 89.64%, with respect to accuracy metrics. Overall, the CNN-LSTM model was successful in detecting botnet attacks from various IoT devices with optimal accuracy.