For the past few years, centralized decision-making is being used for malicious node identification in wireless sensor networks (WSNs). Generally, WSN is the primary technology used to support operations, and security issues are becoming progressively worse. In order to detect malicious nodes in WSN, a blockchain-routing- and trust-model-based jellyfish search optimizer (BCR-TM-JSO) is created. Additionally, it provides the complete trust-model architecture before creating the blockchain data structure that is used to identify malicious nodes. For further analysis, sensor nodes in a WSN collect environmental data and communicate them to the cluster heads (CHs). JSO is created to address this issue by replacing CHs with regular nodes based on the maximum remaining energy, degree, and closeness to base station. Moreover, the Rivest–Shamir–Adleman (RSA) mechanism provides an asymmetric key, which is exploited for securing data transmission. The simulation outcomes show that the proposed BCR-TM-JSO model is capable of identifying malicious nodes in WSNs. Furthermore, the proposed BCR-TM-JSO method outperformed the conventional blockchain-based secure routing and trust management (BSRTM) and distance degree residual-energy-based low-energy adaptive clustering hierarchy (DDR-LEACH), in terms of throughput (5.89 Mbps), residual energy (0.079 J), and packet-delivery ratio (89.29%).

CyberShieldDL: A Hybrid Deep Learning Architecture for Robust Intrusion Detection and Cyber Threat Classification

Early and accurate detection of crop diseases is critical for global food security and efficient agricultural management. Recent advances in deep learning, particularly convolutional neural networks (CNNs) and vision transformers (ViT), have demonstrated exceptional ability to recognize disease symptoms from leaf images. In this article, we present a comprehensive framework for plant disease detection that integrates state-of-the-art deep learning models into smart agriculture systems. We review publicly available datasets (e.g. the PlantVillage dataset with 54,306 leaf images across 14 crop species and 26 disease classes), and discuss data preprocessing and augmentation techniques. We then detail various model architectures: traditional CNNs (e.g. ResNet, MobileNet), efficient CNN variants, ViT-based models, and hybrid CNN–ViT architectures (e.g. FOTCA, AppViT). Our proposed models leverage transfer learning and attention mechanisms to improve accuracy. We describe an experimental setup using multiple leaf-image datasets (tomato, potato, apple, cassava, wheat) and report hypothetical results: for example, our hybrid model achieves ≈99.7% accuracy on PlantVillage and 98–99% on tomato/potato datasets. We include precision, recall, F1 metrics and confusion matrices to analyze performance. Integration into smart farming is discussed: IoT sensors and mobile devices capture leaf images, which are processed by on-device or cloud CNN/ViT models to alert farmers in real time, the depthwise separable convolution block, and the ViT encoding block, respectively. We compare results across models and examine the trade-offs between model complexity and accuracy. Our findings confirm that hybrid CNN–ViT architectures yield the best performance, while lightweight models (e.g. MobileViT, AppViT) enable on-device inference.

Communication in a heterogeneous, dynamic, low-power, and lossy network is dependable and seamless thanks to Mobile Ad-hoc Networks (MANETs). Low power and Lossy Networks (LLN) Routing Protocol (RPL) has been designed to make MANET routing more efficient. For different types of traffic, RPL routing can experience problems with packet transmission rates and latency. RPL is an optimal routing protocol for low power lossy networks (LLN) having the capacity to establish a path between resource constraints nodes by using standard objective functions: OF0 and MRHOF. The standard objective functions lead to a decrease in the network lifetime due to increasing the computations for establishing routing between nodes in the heterogeneous network (LLN) due to poor decision problems. Currently, conventional Mobile Ad-hoc Network (MANET) is subjected to different security issues. Weathering those storms would help if you struck a good speed-memory-storage equilibrium. This article presents a security algorithm for MANET networks that employ the Rapid Packet Loss (RPL) routing protocol. The constructed network uses optimization-based deep learning reinforcement learning for MANET route creation. An improved network security algorithm is applied after a route has been set up using (ClonQlearn). The suggested method relies on a lightweight encryption scheme that can be used for both encryption and decryption. The suggested security method uses Elliptic-curve cryptography (ClonQlearn+ECC) for a random key generation based on reinforcement learning (ClonQlearn). The simulation study showed that the proposed ClonQlearn+ECC method improved network performance over the status quo. Secure data transmission is demonstrated by the proposed ClonQlearn + ECC, which also improves network speed. The proposed ClonQlearn + ECC increased network efficiency by 8-10% in terms of packet delivery ratio, 7-13% in terms of throughput, 5-10% in terms of end-to-end delay, and 3-7% in terms of power usage variation.

Today's society has seen a sharp rise in the number of accidents caused by drivers failing to pay attention to traffic signals and regulations. Road accidents are increasing daily as the number of automobiles rises. By using synthesis data for training, which are produced from photos of road traffic signs, we are able to overcome the challenges of traffic sign identification and decrease violations of traffic laws by identifying triple-riding, no-helmet, and accidents, which vary for different nations and locations. This technique is used to create a database of synthetic images that may be used in conjunction with a convolution neural network (CNN) to identify traffic signs, triple riding, no helmet use, and accidents in a variety of view lighting situations. As a result, there will be fewer accidents, and the vehicle operator will be able to concentrate more on continuing to drive but instead of checking each individual road sign. Also, simplifies the process to recognize triple driving, accidents, but also incidents when a helmet was not used.

The integration of deep learning into precision agriculture has transformed land monitoring by enabling automated, accurate mapping of farmland from satellite imagery. In this study, we develop and evaluate convolutional neural network (CNN) models for semantic segmentation of farmland areas, utilizing multispectral data from Sentinel-2 and Landsat-8. We describe a pipeline combining data preparation, deep convolutional networks, and post-processing to classify each image pixel into farmland vs. non-farmland (and other classes). We experiment with state-of-the-art architectures including U-Net, ResNet-based models, and attention-enhanced CNNs. Our experiments use public datasets (e.g. CORINE, DeepGlobe) and achieve high performance: the best models reach over 90% overall accuracy (OA) and F1-scores above 0.90, outperforming traditional machine learning baselines. For instance, a transfer-learning ResUNet achieved an IoU of 0.81 on DeepGlobe data. Comparative results are reported in tables. Figures illustrate model architectures and example segmentation results. This work demonstrates that modern deep networks can robustly extract farmland from remote sensing data, supporting precision agriculture applications such as crop monitoring and land-use planning.