Aims and Background: Image dehazing is an essential task in computer-vision, aimed at enhancing the clarity and visibility of images degraded by fog and haze. Traditional methods often struggle with handling the complex scattering effects of haze and maintaining the natural colours of the scene. Objectives and Methodology: To address these challenges, we propose a novel approach termed Dual Colour Space Attentional Deep Network (DCSADN) for efficient image dehazing. Our method leverages the advantages of two-colour spaces: RGB and YCbCr, to boost the network's skill to capture both the luminance and chrominance information, thereby improving the dehazing performance. We employ a multi-stage training strategy to optimize the performance of the DCSADN. This multi-stage approach ensures that the network generalizes well across different types of haze conditions. Extensive experiments demonstrate the superiority of our method over state-of-the-art dehazing techniques. Results and Discussions: The results indicate that our approach not only achieves higher quantitative scores in terms of Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM) but also produces dehazed images with more natural colours and details. Conclusion: The findings reveal that the dual colour space processing and attentional modules are crucial for the enhanced performance of our method and providing a valuable tool for improving image quality in hazy conditions.
Background: The System-on-Chip (SoC) design incorporates a high-performance, distributed, dynamic protocol adaptation framework, representing a substantial advancement in embedded system architecture. This design efficiently enhances computational performance and energy efficiency. Objective: This research implements a novel SoC architecture incorporating a high-performance, distributed, dynamic protocol adaptation framework to enhance energy efficiency and performance in wireless network devices. Methods: The framework continuously evaluates a real-time monitoring system that perpetually assesses workload and performance metrics, enabling adaptive selection of the most energy-efficient communication protocol. Results: The proposed design proved superior to existing designs in terms of device utilization and performance metrics. For a 32-bit processor, the proposed design could effectively reduce resource utilization, with 3491 slice registers, 3029 slice LUTs, and 2,800 slice flip-flops, compared with 4,010, 3,180, and 3,910 in the baseline design, respectively. Additionally, it has the potential to decrease the delay from 7.3 ns to 5.25 ns and the power usage from 13.43 mW to 9.69 mW. It reduced the area footprint to 13,109 units from 17,791, while maintaining the operating frequency at 225 MHz. The throughput increased to 130 Gbps with the proposed design, higher than 88 Gbps in the previous implementation. For a 64-bit processor, the proposed design led to improved resource efficiency, utilizing 3491 slice registers, 3029 slice LUTs, and 2800 slice flip-flops, compared with 4010, 4514, and 3012 in the reference design. Furthermore, it decreased the delay from 9.45 ns to 5.35 ns and lowered power consumption from 14.43 mW to 9.24 mW. Finally, it reduced the area footprint to 15,493 units from 17,249 units, while enhancing the throughput to 112 Gbps, greater than that in the existing implementation (96 Gbps). Conclusion: The quantitative enhancements in resource utilization, performance, and energy efficiency underscore the innovation and effectiveness of the proposed SoC design. The results also indicate that the proposed design is superior to existing architectures in terms of computational capability and operational efficiency.
The Internet of Things, a rapidly growing revolutionary technology with the potential to transform human lives, promises a future of interconnected smart devices. Smart IoT applications have influenced almost all aspects of life, ranging from trivial to essential and potentially lifesaving. However, security challenges have limited its widespread adoption. Conventional cryptography, though highly secure, is unsuitable for constrained devices due to complex algorithms and high resource requirements. In contrast, lightweight cryptography can unlock the potential of smart IoT by providing an optimal level of security. This paper presents a detailed design rationale for the most popular and widely used lightweight symmetric block ciphers suitable for smart IoT applications. It provides a comprehensive overview of block cipher design principles, including core construction schemes, diffusion and confusion techniques, key scheduling strategies, hardware- and software-specific design approaches, and trade-offs among design and security metrics. Furthermore, it highlights the key aspects of lightweight block cipher design to assist designers and researchers in selecting the most appropriate cipher for specific smart IoT applications. Considering the constraints and varying requirements associated with IoT, nine lightweight block ciphers— Ascon, AES, LBLOCK, Midori, PRESENT, SIMON, SPECK, SIMECK, and SPARX—were implemented on the ARM Cortex-M3-based LPC1768 IoT hardware development platform. These lightweight primitives were evaluated and compared for various design metrics, including memory (RAM/ROM) utilization, execution time, energy, and power consumption, using ULINKpro and ULINKplus debug adapters. The results indicate that the higher evaluation metric values of the AES block cipher underscore the need for lightweight cryptographic primitives that provide optimal security for resource-limited IoT devices. Moreover, the experimental findings guide the selection of lightweight block ciphers tailored to specific IoT applications: SIMECK, SPECK, SIMON, SPARX, Ascon, Midori, and PRESENT for health systems with limited area and high-speed requirements; SIMECK, SPECK, SIMON, SPARX, and Ascon for energy-efficient transportation systems; and SIMECK, SIMON, SPECK, SPARX, and Midori for memory- and processing-constrained smart home devices.
Introduction: Over the past few years, PWM techniques have been widely used in DSP applications and digital systems. The vital role of this PWM is to control power utilisation for devices such as LEDs and motors. Power consumption is controlled by adjusting the pulse width of the modulating signals. In DSP applications, the PWM signal can also control brightness, motor speed, and other parameters. PWM signals are crucial in fields such as digital communications, signal processing, and power electronics, where they play a vital role in controlling and managing devices. This paper explores the implementation of a PWM (Pulse Width Modulation) generator on an FPGA, aiming to support digital applications and promote energy-efficient, “green” communication. Methodology: The implementation is performed using VIVADO ISE, and power consumption results are targeted for two FPGAs: Zynq 7000 and Zynq Ultrascale+. Verilog HDL is used to write the PWM generator code. This research emphasises a novel capacitance-aware power optimisation context that systematically analyses power dissipation patterns across varying output load capacitances from 0 pF to 20 pF. Results: The experimental results demonstrate that power consumption increases exponentially with capacitance loading, with the Zynq-7000 showing a 104.73% power increase and the Zynq UltraScale+ exhibiting a 152.13% increase at maximum capacitance. The proposed framework contributes to sustainable digital design by providing quantitative guidelines for power-efficient PWM implementations in green communication systems. Discussion: For both FPGAs, as the output load capacitance increases from 0 pF to 20 pF, the power also increases. Additionally, for both FPGAs, the DP increases with increasing capacitance, while the SP declines. There is a 104.73% increase in TP for Zynq 7000 as the capacitance rises to 20 pf. The increase in Zynq Ultrascale+ is 152.13% as the capacitance increases to 20 pf. Conclusion: This work highlights how simple adjustments in components like capacitance can significantly reduce power usage, supporting greener technology and more efficient power management for modern digital applications.
Introduction: Security is an essential problem in Wireless Sensor Networks (WSNs), especially when it is utilized for managing critical healthcare information. Although BlockchainEnabled Wireless Sensor Networks (BEWSNs) offer increased transparency, data integrity, and trust, it is also prone to various security risks. The aim of this study is to identify, classify, and evaluate various emerging risks associated with blockchain-based patient health records systems and their effect on healthcare operations. Methods: A quantitative methodology was employed to conduct this study. A literature review was conducted to identify various risks, which were classified into five categories. The collected data were analyzed to identify various risk patterns through univariate analysis. Results: The results revealed that technological, operational, and security risks are at a higher risk compared to regulatory and privacy risks. The major risks identified were incorrect data entry, legal liability, and high resource utilization for blockchain system maintenance. Additionally, there is a positive correlation between organizational size and perceived security risk. Discussion: The study has demonstrated the challenges associated with blockchain technology in the healthcare industry, despite the numerous benefits and advantages associated with the technology. For instance, the study has pointed out the need to address access control, encryption in smart contracts, the use of standardized data formats, and the enhancement of legal compliance to address the challenges associated with the technology. conclusion: This study makes several noteworthy contributions to Blockchain-Enabled Wireless Sensor Network (BEWSN)risk management practices. A diversified approach is applied to improve the data management and security of Blockchain Wireless Sensor Network (BWSN). Firstly, dependable access control systems and smart contracts should ensure data quality and efficient record transfer, effectively reducing technologya hazards. Second, enhancing security measures within smart contracts with encryption methods is critical to mitigating security risks connected with illegal access and data breaches. Third, operational risks can be avoided by rewarding off- peak transactions, which helps to distribute system load and uniformly minimize operational disturbances. Fourth, the danger to privacy can be reduced by encouraging the use of standardized data formats and working with industry associations to establish standard privacy precautions. Finally, careful adherence to data protection rules and joint efforts with legal professionals is critical in a data breach, assuring the organization's resilience in the face of potential obstacles. The study paves the way for exciting research studies in this fledgling area. Perceived risks and organizational size were shown to be related in our study. Subsequent investigations may explore this correlation in further detail to comprehend how their size influences healthcare organizations' perceptions of the risks associated with implementing Conclusion: This has been achieved by the identification of the challenges associated with the technology, which will enable the development of a more robust blockchain technology in the healthcare industry.
Introduction: The aspects of safety and reliability are inherent in autonomous driving systems, particularly in predicting and evaluating the risks of failures before they lead to unsafe conditions. The objective of the study is to enhance failure prediction by diversifying data learning to enhance the steering angle prediction. The main goal of the solution is to develop a scalable model. Methods: The proposed framework uses a convolutional neural network trained with a pilot-based approach, where saliency maps are generated via Visual Backpropagation. The model is initially trained on the Udacity Jungle simulated dataset and subsequently extended with the Udacity Lake dataset and a real-world driving dataset to increase the environmental variety. The steering angles of all the datasets are normalised through min-max scaling to make them have the same numerical range and learning behaviour. results: The saliency-enhanced CNN demonstrated improved interpretability and prediction reliability by achieving higher pixel accuracy and lane boundary precision than the baseline CNN. Enhanced feature localization was also confirmed by the integration of VisualBackProp, which resulted in enhanced SSIM scores. Results and Discussion: The results are significant in maintaining the value of various datasets, and the uniform preprocessing improves the model's robustness under various driving conditions. Salient map learning provides interpretable indications of steering behaviour, enabling the prediction of failures in alternative ways. Conclusion: The proposed framework provides better failure prediction and enhanced generalisation, which is why it is applicable to real-world scenarios such as autonomous driving.
Abstract: The increasing integration of Wireless Sensor Networks (WSNs) and Internet of Things (IoT) technologies across domains such as smart cities, healthcare, and industrial automation demands sensor platforms that are not only efficient and scalable but also adaptable to intelligent and secure applications. While many earlier reviews focused on general specifications and often included outdated hardware, this paper presents an up-to-date and comprehensive review of Commercial Off-The-Shelf (COTS) platforms that are actively used in current research and industry. By systematically excluding nodes that are no longer produced and incorporating newly available sensor nodes, this study addresses critical gaps in previous surveys and ensures practical relevance. COTS platforms are categorized into Single Board Microcontrollers (SBMs) and Single Board Computers (SBCs), providing a clear and structured framework that considers performance, energy consumption, cost, and AI readiness. The review further highlights emerging trends such as the adoption of secure edge computing, real-time behavioral analytics, and on-device learning, which are increasingly vital for modern deployments. Moreover, the growing interest in Reduced Instruction Set Computer Five (RISC-V) architectures is discussed, offering open-source flexibility and enhanced performance as a promising alternative for future WSN implementations, although its supporting software ecosystem is still maturing. Finally, the paper introduces a data-informed decision- support framework to help practitioners select appropriate platforms based on applicationspecific constraints and performance trade-offs. By bridging the gap between traditional hardware classifications and evolving computational demands, this work provides a valuable and practical guide for researchers and developers navigating the rapidly changing WSN and IoT landscape.
Wake-up Radio Medium Access Control (WuR-MAC) protocols represent a vital advancement in Wireless Sensor Networks (WSNs), which aim to reconcile energy efficiency with network responsiveness. This study presents a detailed classification and analysis of WuR-MAC protocols that function without duty cycling. These are systematically categorised into Always Active WuR-MAC and Radio Triggered WuR-MAC Protocols. The Always Active WuR-MAC protocols, among the earliest developed, ensure high responsiveness by maintaining a continuously active, low-power radio, separately for transmission and reception of wake-up frames, albeit at the cost of increased energy consumption and a limited wake-up range. In contrast, Radio Triggered WuR-MACs are further divided into Passive WuRx and Ultra-low Power (ULP) Active WuRxbased protocols. Passive WuRx achieves exceptional energy savings by harvesting energy from incoming signals; however, it suffers from a constrained activation range and wake-up time, limiting the amount of power available to switch on the wake-up circuit. Conversely, ULP-Active WuRx protocols strike a more favourable balance between energy consumption and wake-up range, making them suitable for applications demanding extended coverage and energy efficiency. The classification highlights how evolving trade-offs in power consumption and wake-up capabilities have shaped protocol development, corroborating findings from earlier studies. This review aims to guide future protocol design by underscoring the technological nuances and performance implications of WuR-MAC strategies without duty cycling.
The article has been withdrawn at the request of the editor of International Journal of Sensors, Wireless Communications and Control. The publisher apologizes to the readers of the journal for any inconvenience this may have caused. The Bentham Editorial Policy on Article Withdrawal can be found at https://benthamscience.com/editorial-policies-main.php. BENTHAM SCIENCE DISCLAIMER: It is a condition of publishers that manuscripts submitted to this journal should not be simultaneously submitted or published elsewhere. Furthermore, any data, illustration, structure, or table that has been pub- lished elsewhere must be reported, and copyright permission for reproduction must be obtained. Plagia- rism is strictly forbidden, and while submitting the article for publication, the authors agree that the pub- lishers have the legal right to take appropriate action against the authors if plagiarism or fabricated infor- mation is discovered. By submitting a manuscript, the authors agree that the copyright of their article is transferred to the publishers if and when the article is accepted for publication.
In a recent study, a generalized implicit transmission (GIT) technique that can transmit multiple implicit sequences while transmitting a single explicit sequence over a channel has been introduced. Instead of treating all sequences as independent, as in the previous study, this study considers the explicit sequence and all implicit sequences of a GIT collectively as a single code, referred to as a GIT coding scheme. The overall code rate Roverall and the inherent coding gain achieved by a GIT coding scheme, due to the transmission of information, is discussed implicitly. A GIT coding scheme constructed from a rate R code to function as a rate Roverall code, on average, transmits Roverall/R number of codewords of a rate R code for every single codeword transmitted over the channel by transmitting (Roverall − R)/R number of codewords over all implicit sequences. A simple way to convert existing practical codes into GIT coding schemes is also discussed. The numerical results presented with the LDPC codes employed in the WiFi and the 5G standards demonstrate that GIT coding schemes can achieve very high coding gains over conventional codes while functioning as high-rate codes. Due to its ability to transmit the majority of information implicitly, which does not require any additional bandwidth or transmitted power, it is demonstrated here that GIT coding schemes can operate in the so-called unreachable region relative to the Shannon-Hartley bound.
The data collected in Internet of Things (IoT) applications consist of unreliable and erroneous data due to their deployment in harsh or unattended environments. Such data is considered an anomaly as it deviates from the regular data. These anomalies need to be identified correctly to enhance decision-making. For this purpose, machine learning techniques have gained significant attention due to their ability to classify the data into normal and abnormal (or anomaly). Methods: This work proposes novel adaptations to supervised and semi-supervised machine learning algorithms by integrating the Mahalanobis Distance (MD) metric. These adapted algorithms are named as Mahalanobis Binary Classification (M-BC) and Mahalanobis One Class Classification (MOCC). The performance of these proposed algorithms was evaluated on well-known IoT sensor datasets using performance metrics such as balanced accuracy, F1-Score, and AUC-ROC score. The results show that the M-BC algorithm exhibits significant improvements over conventional machine learning methods across several datasets considered in this study, including SHM4, MHM1, Occupancy, and Timeseries. The M-BC achieved an average improvement of 13.03% in balanced accuracy, 10.29% in F1-Score, and 13.16% in AUC score. Similarly, the M-OCC algorithm demonstrated substantial gains in one-class classification, with an average improvement of 21.07% in balanced accuracy, 26.49% in F1-Score, and 26% in AUC score across datasets such as AnomIoT, IBRL, SHM4, MHM1, Occupancy, and Timeseries compared to OCSVM. The results confirm that the proposed MD-based approaches are found to be simple, effective, and more accurate for detecting anomalies in IoT sensor data compared to their base methods. The integration of the MD metric significantly enhanced the ability of the algorithms to identify anomalous data points across various IoT domains. The work presented successfully demonstrated the incorporation of the Mahalanobis distance into binary and one-class classification algorithms to improve anomaly detection performance. These M-BC and M-OCC algorithms show a robust and efficient solution to ensure data reliability in IoT sensor networks.
Internet of Things-Low Power and Lossy Networks (IoT-LLN) is a technology that allows devices with limited power and bandwidth to communicate, making it ideal for IoT applications. In IoT-LLN, two key protocols are responsible for network formation and maintenance: IPv6 over the IEEE 802.15.4e TSCH mode (6TiSCH) and the Routing Protocol for Low-Power and Lossy Networks (RPL). These protocols exchange many control messages between devices to facilitate prompt network formation and maintenance. The 6TiSCH protocol allows these control messages to be transmitted or received only during the minimal cell. In IoT-LLN, only one minimal cell is allotted for each slot frame. All nodes in the network send their control messages at the minimum cell rate, which becomes overcrowded, leading to message collisions and control message queuing delays. This causes delays in joining new nodes to the network and frequent parent switches in existing nodes. To study this issue, simulations were run using the Contiki-NG COOJA simulator. Simulation results show that, even in scenarios with static nodes, most nodes switch parents often. This reduces network reliability and increases power consumption, a major concern for low-power IoT devices.
Unmanned Aerial Vehicle (UAV) communication networks can move in three-dimensional (3D) space, which can lead to a higher Line of Sight (LoS) probability for wireless communication channels between network nodes. Non-Orthogonal Multiple Access (NOMA) techniques are one of the proposed methods to overcome emerging challenges and meet requirements in the fifth (5G) and sixth (6G) generations of cellular systems. Due to the utilisation of an imperfect Successive Interference Cancellation (SIC) receiver, the performance of the NOMA scheme is vulnerable to error propagation. This paper aims to minimise the power consumption of the NOMA-UAV uplink network, which has limited energy resources. In addition, the impact of Channel State Information (CSI) on resource allocation methods is analysed. The hybrid multiple access schemes are introduced to minimise the sum power in the NOMA-UAV uplink network, subject to minimum data rate and available resource constraints. The frequency sub-bands and transmission power are jointly optimised to decrease the impact of error propagation in the SIC receiver via three hybrid multiple access models, including (i) a hybrid NOMA and OMA scheme with user grouping on multi-NOMA sub-bands, (ii) a hybrid NOMA and OMA scheme with one NOMA sub-band, and (iii) a multi-NOMA sub-bands scheme. The iterative Sequential Quadratic Programming (SQP) algorithm is employed to solve the optimisation problems. The numerical results show the effectiveness of the proposed hybrid schemes in terms of power consumption compared to the conventional NOMA scheme as benchmarks, where the improvement ratio is more than 20%. The hybrid schemes outperform the NOMA schemes reported in other papers, where the performance was compared relative to the OMA scheme. The high diversity of available resources in the hybrid multiple access schemes contributed to achieving the best performance in terms of sum power, which negatively affects the time cost performance. Considering an imperfect SIC receiver, available resources in the NOMA-UAV uplink network are optimised at the ground base station to decrease the transmission power of the UAV communication system, which enhances the energy efficiency of UAVs.
Post-quantum cryptography (PQC) algorithms have been developed in recent decades, substituting standard Post-Quantum (PQ) algorithms to withstand quantum attacks. The Advanced Encryption Standard (AES) is a prevalent symmetric encryption technique utilized for data security and efficiency. It guarantees the secrecy and integrity of data encryption. Crystal-Kyber is a key encapsulation mechanism (KEM) utilizing lattice-based cryptography, engineered to withstand both conventional and quantum attacks. The AES and Crystal-Kyber algorithms exemplify distinct methodologies in encryption and key management. A research gap exists in the amalgamation of AES with Crystal-Kyber to implement a hybrid encryption system that ensures security against both conventional and quantum threats. Current cutting-edge research examines quantum computing (QC) for safe advanced RISC-V SoCs, leveraging the flexibility and scalability of Post-Quantum cryptosystems. QCA (Quantum Dot Cellular Automata) is a technology that operates on quantum mechanics at high frequencies (Terahertz), offering a transistor-less architecture to minimize circuit complexity. Moreover, QCA consumes less power, operates at elevated frequencies, and exhibits greater density in comparison to traditional CMOS (Complementary Metal Oxide Semiconductor) circuits. All PQC methods, in conjunction with classical quantum cryptography algorithms, are presented to tackle the prevailing challenges. Additionally, other studies examine the hardware and software implementations of post-quantum cryptography within the RISC-V architecture. A thorough research effort on QCA-based cryptographic circuits is examined, along with an innovative method for the next generation of secure nano-communication.
The error was noted in the article titled “Implementing Wireless Sensor Network Through Machine Learning Techniques” published in International Journal of Sensors, Wireless Communications and Control, 2025, 15(3), 268-280 [1]. In the abstract of the article, the conclusion was incomplete, which has now been completed. Details of the error and a correction are provided here. Original: Conclusion: The training data accuracies are as follows: Linear Regression (46.64%), Random Forest (99.67%), XG Boost (99.99%), SVM (32.10%), K Nearest Neighbor (98.56%), Naïve Bayes Classifier (96.45%), Principle Component Analysis (97.88%) and NN (34.86%). The testing data accuracies are as follows: Linear Regression (22.00%), Random Forest (97.18%), XG Boost (96.94%), SVM (21.75%), K Nearest Neighbor (97.66%), Naïve Bayes Classifier(64.49%), PCA(99.38%) and NN(28.94%). Linear Regression, Random Forest, SVM, Naive Bayes Classifier, PCA, K Nearest Neighbour, and XG Boost were all examined. Results revealed that Corrected: Conclusion: The training data accuracies are as follows: Linear Regression (46.64%), Random Forest (99.67%), XG Boost (99.99%), SVM (32.10%), K Nearest Neighbor (98.56%), Naïve Bayes Classifier (96.45%), Principle Component Analysis (97.88%) and NN (34.86%). The testing data accuracies are as follows: Linear Regression (22.00%), Random Forest (97.18%), XG Boost (96.94%), SVM (21.75%), K Nearest Neighbor (97.66%), Naïve Bayes Classifier(64.49%), PCA(99.38%) and NN(28.94%). Linear Regression, Random Forest, SVM, Naive Bayes Classifier, PCA, K Nearest Neighbour, and XG Boost were all examined. Results revealed that Random Forest and XGBoost achieved the highest accuracy, making them the most effective models for WSN prediction. We regret the error and apologize to the readers. The original article can be found online at: https://www.benthamscience.com/article/143403
Background and Objective: In this paper, we address potential solution for downlink wireless communications by integrating visible light communications (VLC) with the broadband radio frequency (RF) networks operating at 60 GHz-millimeter wave (mmWave) band. For this hybrid design, the outdoor 60 GHz-mmWave based RF link is utilized to ensure backhaul connectivity for VLC indoor system, while the VLC exploiting the lighting infrastructure that essentially used LEDs as optical source to provide low-cost and high-speed data access. Methods: The hybrid 60 GHz-mmWave/VLC system is analyzed by using 16-QAM OFDM signal, and its performance is investigated by evaluating the signal to noise ratio (SNR) and the received signal power distributions, for regular placement of LEDs and random location of receivers within the room. The impact of the transmitter-receiver parameters, and their orientations and directivities are also considered. Results and Discussion: Numerical results show the efficiency of the proposed system, which can retransmit RF signals at 60 GHz-mmWave band using visible optical carriers in the indoor environment. Conclusion: The results suggest that the hybrid 60 GHz-mmWave/VLC system is able to provide reliable wireless data transmission, making it an attractive solution for downlink indoor communications.
Introduction: Early and precise landslide prediction remains a critical challenge for mitigating their devastating impacts. Traditional methods often struggle to integrate both spatial and temporal data effectively, leading to limited prediction accuracy. This study aims to develop a deep learning model that combines Convolutional Autoencoders (CAEs) for spatial feature extraction with Recurrent Neural Networks (RNNs) to capture temporal dynamics. Methods: The proposed model leverages CAEs to learn robust spatial representations from the 14- band Landslide4Sense dataset, while the RNN component captures the temporal patterns crucial for landslide detection. This integrated approach enhances the prediction capability by considering both spatial and temporal factors. Results: The approach demonstrates an impressive landslide prediction accuracy of 0.988, with performance metrics of precision: 0.987, recall: 0.972, and F1-score: 0.982, highlighting its effectiveness in landslide prediction. Discussion: The model successfully integrates spatial and temporal dimensions, outperforming traditional prediction methods. Its deep learning design enhances robustness and adaptability across geospatial terrains. Conclusion: This work paves the way for the application of advanced deep learning models in realworld landslide prediction. By integrating spatial and temporal data, the model offers a promising solution for mitigating landslide-related risks, potentially saving lives and infrastructure.
Introduction: An automobile is a software-defined machine on top of the wheels including more than 100 electronic control units (ECU) with million lines of code. The integration of ECU in the automobile ensures the customer’s needs by providing safety, security, entertainment, and comfort features. Methods: The firmware integrated into the ECU should be updated to avoid latency in operation and bugs, and to add new features. The traditional update process of ECU holds loopholes like more waiting time, unavailability of service centers, and security threats. To overcome this, over-the-air (OTA) updates are introduced in the vehicle, but security is the major concern while transmitting firmware over the air. Results and Discussion: The proposed system ensures the wireless firmware update with the uptane framework with background Timestamp Update Framework (TUF) ensures the security. The timestamp generated is valid for 86400 seconds to validate the freshness. In addition, security assessment and reverse engineering are performed on the designed system to check for security breaches. Conclusion: The system secures the firmware over arbitrary and replay attacks on the Original Equipment Manufacturer (OEM) server.
Introduction: Timely detection of catastrophic natural disasters, such as forest fires, is critical to minimizing losses and ensuring rapid response. Artificial intelligence is increasingly being recognized as a valuable tool in enhancing various stages of disaster management. Methods: This paper presents the development of a smart framework utilizing machine learning techniques for real-time detection and monitoring of natural disasters, specifically forest fires. The proposed approach employs a 10-layer convolutional neural network (CNN) that classifies aerial images into Fire, Non-Fire, and Smoke categories with high precision and speed. In addition to this, a CNN-based feature extraction process is performed and integrated with various ML classifiers, including support vector machine, k-nearest neighbor, decision tree, random forest, and extra trees. Results and Discussion: Extensive performance analysis reveals that the proposed 10-layer CNN model outperforms other classifiers, achieving an accuracy of 97.64% in the binary classification of fire vs. nonfire and 95.61% in the three-class classification of Fire, Non-Fire and Smoke classes. Furthermore, a comparative study with existing state-of-the-art methods demonstrates the proposed model's superior performance in both accuracy and computational complexity. Conclusion: These results demonstrate the potential of the proposed CNN-based framework to serve as a reliable and effective tool for real-time disaster management across various applications, providing valuable support to emergency response teams in mitigating the impact of natural disasters.
