The proposed Q-learning optimized Internet of Things (IoT) integrated VLC system is highly suitable for indoor environments where electromagnetic interference (EMI), device density and real-time data transmission are critical challenges. Visible Light Communication (VLC) serves as an effective alternative to traditional RF communication by utilizing existing LED lighting infrastructure for dual-purpose illumination and data transmission. In such indoor settings the system enables real-time monitoring and secure transmission of vital information such as environmental data to an IoT-based cloud platform. The use of VLC in the visible light spectrum eliminates EMI making it ideal for environments where Radio Frequency (RF) communication is either restricted or unreliable. To adapt to dynamic indoor conditions such as human movement or changing light intensities the system employs Q-learning to continuously optimize transmission parameters like modulation index and power levels. This reinforcement learning approach ensures enhanced link reliability, reduced Bit Error Rate (BER) and stable communication under fluctuating indoor conditions. Additionally, the incorporation of Enhanced Asymmetrically Clipped Duty-Cycled Optical OFDM (EACDO-OFDM) improves spectral efficiency and power utilization thus making the solution energy-efficient and scalable for large-scale indoor IoT deployments. By integrating adaptive machine learning techniques with VLC and IoT the system delivers robust, intelligent and energy-conscious communication, well-suited for smart building automation, indoor sensing and real-time monitoring applications.

This work introduces a Field Programmable Gate Array (FPGA) based image compression method utilizing Huffman coding (FICH) to enhance the efficiency of wireless networks, particularly within the Cloud-based Radio-Access-Network (C-RAN) architecture. The FICH method addresses image compression challenges in C-RAN, offering faster compression and decompression times compared to existing FPGA approaches. The findings include significant improvements in Bit-Error-Rate (BER), Symbol-Error-Rate (SER), and Error-Vector Magnitude (EVM), with average BER, SER, and EVM improvements of 37.85%, 24.64%, and 24.56% for fewer RRHs, and 96.10%, 91.13%, and 48.72% for more RRHs, respectively. Additionally, the FICH method demonstrated reduced encoding and decoding times, averaging 0.0545 seconds versus 0.0853 seconds when compared with existing approach. The approach also ensures robust and scalable compression, optimizing resource utilization with FPGA-based hardware acceleration. These advancements support the growing data demands of modern wireless networks.

In modern computing environments such as cloud, edge, and fog computing, as well as IoT networks and real-time systems, meeting workload deadlines is critical to ensure reliability, quality of service, and user satisfaction. The traditional scheduling algorithms often fail to adequately address the constraints associated with the workloads, particularly deadlines, the dynamic nature of workloads and resources, and the inherent resource limitations. Deadlines are the most important constraints, especially for time-sensitive applications. Achieving deadline compliance requires optimal resource provisioning, scheduling, resource allocation, resource scaling, workload migration, etc. This paper proposes a novel deadline-aware cloudlet scheduling algorithm, the Two-Stage Deadline-Aware Cloudlet Scheduling Algorithm (TSDACS), designed to minimize deadline misses through efficient resource provisioning and scaling strategies. In stage one, the algorithm provisions virtual machines with suitable configurations and quantities based on the requirements of the cloudlets to ensure they can be processed within their deadlines. Cloudlets are then scheduled onto the virtual machines in a way that minimizes deadline violations. In stage two, if the initially provisioned virtual machines fail to meet the deadlines, horizontal scaling is applied, up to a limited threshold, to enhance the performance and the deadline compliance. Experimental results demonstrate that the proposed TSDACS algorithm outperforms existing approaches, such as CPDALB, DBS, and RDLBS2, in terms of deadline miss ratio, makespan, response time, and cost efficiency, while maintaining competitive VM utilization and effective load balancing.

The increase in attacks on Internet of Things (IoT) systems in recent years is a concern, especially considering their growing application in industrial environments (IIoT) and the projected value they are expected to reach shortly. In light of this, the article describes the development of a proof of concept for an IIoT architecture that employs artificial intelligence techniques to detect real-time network attacks in IIoT private network environments. The approach included training and testing various supervised and unsupervised machine learning, as well as deep learning algorithms. The training process showed that real-time analysis is feasible, using SVM-based models for smaller networks and autoencoders for larger, more complex networks. In terms of performance, models trained with SVM clearly outperform others because they achieve perfect classification results with the training data, and SVMs have a maximum response time of about 5 seconds for 70,000 requests. In contrast, autoencoders respond to the same type of attack and requests always within 1 second. Additionally, the proof of concept shows that the proposed modular architecture can effectively visualize detected network attacks while satisfying requirements for real-time packet collection and analysis.

Enhancing the software maintenance greatly depends on the precise and prompt handing out of bug reports according to their bug-category and importance. To resolve the aforementioned problems, an automated method of classifying and ranking bug reports is required. Numerous scholars have recently looked into the automated classification and prioritization of bug reports. But not much has been accomplished in this area. During software development, the most crucial stages are testing and maintenance. In these phases of development activity, bug reports are essential. When software modules are being tested, the software quality assurance team creates a bug report. But the main issue that comes up while analysing bug data that is written in normal text. As a result, processing and extracting information from it is extremely challenging. The aforementioned requirements are the driving force for this research. The Proposed research suggested creating a hybrid model that takes advantage of machine learning models' contextual awareness as well as more conventional feature extraction methods (such as TF-IDF). A downstream classifier (such as an SVM, logistic regression) can receive these two feature sets (one from TF-IDF and the other from BERT) after they have been concatenated. This enables the model to take advantage of the extensive contextual relationships that BERT captures as well as the statistical importance of phrases (TF-IDF) These two approaches were used separately in the earlier research, which resulted in less performance. The research made use of a confidential dataset that was acquired from a private company upon request for performing testing, the data included from eight hundred employees. To aid in model training, bug keywords were first taken out of the bug description field. The results shows that proposed model achieves 89% accuracy.

In the era of digital transformation, the Internet of Things (IoT) has revolutionized everyday objects, and IoT gateways play a pivotal role in managing data flow within these networks. However, the dynamic and expansive nature of IoT networks poses significant cybersecurity challenges, demanding the development of adaptive security systems to protect against evolving threats. The research paper presents the development of the CoralMatrix Security framework, a novel approach to IoT cybersecurity using advanced machine learning algorithms. The framework incorporates the AdaptiNet Intelligence Model, integrating deep learning and reinforcement learning for effective real-time threat detection and response. To comprehensively evaluate the framework's performance, the study utilized the N-BaIoT dataset, facilitating a quantitative analysis that provided valuable insights into the model's capabilities. The results of the analysis showcased the CoralMatrix Security framework's robustness across various dimensions of IoT cybersecurity. Notably, the framework achieved a high detection accuracy rate of approximately 83.33%, underscoring its efficacy in identifying and responding to Cybersecurity threats in real-time. Furthermore, the research examined the framework's scalability, adaptability, resource efficiency, and robustness against diverse cyber-attack types, all quantitatively assessed to provide a comprehensive understanding of its capabilities. The paper suggests future work to optimize the framework for larger IoT networks and adapt continuously to emerging threats, aiming to expand its application across diverse IoT scenarios. The CoralMatrix Security framework, with its proposed algorithms, emerges as a promising, efficient, effective, and scalable solution for the dynamic challenges of IoT cybersecurity.

The intersection of agricultural activities and natural habitats presents several implications, one of the most pressing being the intrusion of various species into crop fields. To address the continued and growing challenge of discovering methods for successful mitigation and protection of crops, a new method must be introduced. Current practices, such as using pheromones and monitoring by humans, yield suboptimal results due to inefficiency, labour-intensive processes, and environmental damage. This paper proposes a novel method based on deep learning technologies, namely UNet and EfficientNet-B7 architectures, for semantic segmentation of crop-damaging species in agricultural habitats. The method provides a reliable means to detect species intrusion by integrating it into automated monitoring operations. Whether identifying the segmentation point of target species or utilizing feature-based detection with the proposed model, the application of advanced convolutional neural networks ensures the accuracy of these systems. This method demonstrates higher performance in species identification and proposes newer and ecologically friendly methods of implementing non-invasive monitoring for conservation alongside agriculture. The experimental results confirm that the proposed UNet with EfficientNet-B7 achieves a high accuracy ratio of 95.3%, an F1-score ratio of 92.4%, and an efficiency ratio of 97.6% compared to other existing models.

Road crack detection is a crucial safety measure for any country, especially in regions with complex road networks. In India, most roads are well-connected to cities and urban areas. However, urban roads often suffer frequent damage due to various factors. This study focuses on detecting road damage in Indian urban areas using a Deep Convolutional Neural Network (DCNN). We have developed a model specifically designed for identifying cracks in Indian urban roads. To evaluate the proposed model, we collected over 700 images of damaged roads from different urban locations across Tamil Nadu. In this work, we employed the DCNN algorithm for road crack detection, which has proven to be an efficient approach. The proposed method incorporates Histogram Equalization (HE) and Contrast Limited Adaptive Histogram Equalization (CLAHE) to enhance the contrast of road images. This is the first initiative aimed at developing a road damage detection system tailored to urban roadways in India. A comparative analysis was conducted during the preprocessing stage by applying both HE and CLAHE techniques. The model was trained using 5,000 roadside images, including both cracked and non-cracked surfaces. During training, image enhancement was performed using HE and CLAHE, and the processed images were then used to train DCNN. For testing, 700 roadside images (with and without cracks) were utilized, following the same preprocessing steps. The model's accuracy was determined based on its ability to correctly identify road cracks. Results indicate that the proposed model performs effectively for crack detection on CLAHE-enhanced images of Indian urban roads. Additionally, the proposed model's performance was compared with existing models such as ResNet, VGG16, and VGG19 using the same dataset. Evaluation metrics including accuracy, precision, recall, and F1-score were used. The proposed model achieved 98.6% accuracy, 98.5% precision, 99.6% recall, and a 99% F1-score.

Coronary artery disease (CAD) is the most common cardiovascular disease. Risk factors impact CAD progression. Diagnostic and therapy methods for this illness include several costly side effects. Consequently, researchers are seeking economical and precise techniques for diagnosing this condition. Machine learning algorithms may assist doctors in the early diagnosis of the condition. Hence this work presents an efficient approach for feature selection and classification of abnormal heart rate patterns by combining Joint Mutual Information (JMI), Quantum Annealing, and a Bayesian ensemble model using CatBoost and XGBoost classifiers. The method starts with Joint Mutual Information to rank features based on their dependency with the target variable, identifying the most informative features for classification. Quantum Annealing, specifically simulated annealing in this case, is then used to optimize the subset of features by exploring the feature space and selecting the most relevant combinations, thus improving the model's performance by avoiding suboptimal solutions. The selected characteristics are then input into a Bayesian ensemble of CatBoost and XGBoost classifiers, which are trained to forecast heart rate irregularities. This ensemble method integrates the advantages of gradient boosting models to improve forecast accuracy and mitigate overfitting. The proposed technique is termed Probability-based Bayesian Statistics with Ensemble Boost (PBSEnsBoost), achieving an accuracy of 98.2%, specificity of 93.6%, sensitivity of 92.6%, and an F1-score of 95.7%.

Constant video data streaming to central servers’ costs high bandwidth and storage. This article proposes a lightweight and cost-effective secure smart home infrastructure employing a single board computer and a software motion for camera surveillance. Video screens from several cameras are monitored by the motion program, which acts when movement is detected. The presented framework also sends email and smartphone messages to the smart homeowner efficiently if motion is detected. To increase the sustainability of the framework above and beyond, we have integrated renewable energy to power the NVIDIA Jetson Nano and the cameras as opposed to conventional sources of energy, making our framework eco-friendly. Four advanced technology and alert notification methodologies are compared. For both indoor and outdoor environments, the effectiveness and adaptability of the Edge AI (AI on edge) powered IoT framework for smart surveillance has been evaluated. The framework could achieve 94% accuracy, 92% precision, and 96% recall in indoor scenarios and showed its ability to detect motion in challenging indoor scenarios. Despite difficulties like weather, foliage, and animal disturbances, it remained accurate to 87%, precise at 85%, and recalled 92% in outdoor areas.