Search

Abstract Our environment has been significantly impacted by climate change. According to previous research, insect catastrophes induced by global climate change killed many trees, inevitably contributing to forest fires. The condition of the forest is an essential indicator of forest fires. Analysis of aerial images of a forest can detect deceased and living trees at an early stage. Automated forest health diagnostics are crucial for monitoring and preserving forest ecosystem health. Combining Modified Generative Adversarial Networks (MGANs) and YOLOv5 (You Only Look Once version 5) is presented in this paper as a novel method for assessing forest health using aerial images. We also employ the Tabu Search Algorithm (TSA) to enhance the process of identifying and categorizing unhealthy forest areas. The proposed model provides synthetic data to supplement the limited labeled dataset, thereby resolving the frequent issue of data scarcity in forest health diagnosis tasks. This improvement enhances the model's ability to generalize to previously unobserved data, thereby increasing the overall precision and robustness of the forest health evaluation. In addition, YOLOv5 integration enables real-time object identification, enabling the model to recognize and pinpoint numerous tree species and potential health issues with exceptional speed and accuracy. The efficient architecture of YOLOv5 enables it to be deployed on devices with limited resources, enabling forest-monitoring applications on-site. We use the TSA to enhance the identification of unhealthy forest areas. The TSA method effectively investigates the search space, ensuring the model converges to a near-optimal solution, improving disease detection precision and decreasing false positives. We evaluated our MGAN-YOLOv5 method using a large dataset of aerial images of diverse forest habitats. The experimental results demonstrated impressive performance in diagnosing forest health automatically, achieving a detection precision of 98.66%, recall of 99.99%, F1 score of 97.77%, accuracy of 99.99%, response time of 3.543 ms and computational time of 5.987 ms. Significantly, our method outperforms all the compared target detection methods showcasing a minimum improvement of 2% in mAP.

Power-efficient wireless sensor network routing techniques (WSN). Clustering is used to extend WSNs' lifetimes. One node act as the cluster head (CH) to represent the others in communications. The member nodes are less important than the cluster hub (CH) in the clustering procedure. Fuzzy techniques based on clustering theory may provide evenly distributed loads. In this study, we provide a fuzzy-logic-based solution that factors in distance to base station (BS), number of nodes, remaining energy, compactness, distance to communicate within a cluster, number of CH, and remaining energy. Fuzzy clustering has a preliminary and final step. First, we select CH based on distance to the base station (BS), remaining node vigor, and node compactness. In the second phase, clusters are created by combining nodes that aren't already in a CH, using density, outstanding vigor, and detachment as limitations. The proposed solution increases load balancing and node longevity. This work provides a unique hybrid routing technique for forming clusters and managing data transfer to the base station. Simulation findings confirm the protocol's functionality and competence. Reduced energy use keeps network sensor nodes online longer. The framework outperforms Stable Election Protocol (SEP), hybrid energy-efficient distributed clustering (HEED), and Low Energy Adaptive Clustering Hierarchy (LEACH). Using the nodes' energy levels to create a grid pattern for the clusters gave four clusters. In addition, the proposed method has a 4347%, 41.46%, 39.26%, 37.57% and 35.67% reduction in average energy consumption when compared with the conventional algorithms. The proposed technologies could increase the network's lifetime, stability interval, packet transfer rate (throughput), and average energy. The suggested protocol is at least 50% better in every statistic that was looked at, such as network lifetime, stability interval, packet transmission rate (throughput), and average energy use.

Twitter, Instagram and Facebook are expanding rapidly, reporting on daily news, social activities and regional or international actual occurrences. Twitter and other platforms have gained popularity because they allow users to submit information, links, photos and videos with few restrictions on content. As a result of technology advances (“big” data) and an increasing trend toward institutionalizing ethics regulation, social network analysis (SNA) research is currently confronted with serious ethical challenges. A significant percentage of human interactions occur on social networks online. In this instance, content freshness is essential, as content popularity declines with time. Therefore, we investigate how influencer content (i.e., posts) generates interactions, as measured by the number of likes and reactions. The Gradient Boosted Decision Tree (GBDT) and the Chaotic Gradient-Based Optimizer are required for estimation (CGBO). Using earlier group interactions, we develop the Influencers Prediction issue in this study’s setting of SN-created groups. We also provide a GBDT-CGBO framework and an efficient method for identifying users with the ability to influence the future behaviour of others. Our contribution is based on logic, experimentation and analytic techniques. The goal of this paper is to find domain-based social influencers using a framework that uses semantic analysis and machine learning modules to measure and predict users’ credibility in different domains and at different times. To solve these problems, future research will have to focus on co-authorship networks and economic networks instead of online social networks. The results show that our GBDT-CGBO method is both useful and effective. Based on the test results, the GBDT-CGBO model can correctly classify unclear data, which speeds up processing and makes it more efficient.

Wireless Sensor Networks (WSN) have distributed a collection of tiny sensor nodes deployed randomly in the given symmetry environment to sense natural phenomena. The sensed data are disseminated symmetrically to the control station using multi-hop communication. In WSN, the energy conservation during node coverage plays a major role in detecting node failure and providing efficient and symmetrical data transmission to the nodes of WSN. Using the cluster method and efficient localization techniques, the nodes are grouped and the precise location of the nodes is identified to establish the connection with the nearby nodes in the case of node failure. The location accuracy is achieved using the localization estimation of the anchor nodes and the nearest hop node distance estimation using the received signal strength measurement. The node optimization can be performed efficiently by the accurate estimation of the localization of the node. To optimize the node coverage and provide energy efficient and symmetrical localization among the nodes, in this paper, a cluster-based routing protocol and a novel bio-inspired algorithm, namely, Modified Bat for Node Optimization (MB−NO), to localize and optimize the unknown nodes along with the reinforcement-based Q learning algorithm is proposed with the motive of increasing the accuracy estimation between anchor nodes and the other neighbor nodes, with the objective function to optimize and improve the nodes’ coverage among the network’s nodes in order to increase the nodes’ localization accuracy. The distance metrics between the anchor nodes and other neighbor nodes have an estimated symmetry with three node positions, namely C-shape, S-shape and H-shape, using the Q learning algorithm. The proposed algorithm is implemented using the NS3 simulator. The simulation results show that the accuracy and precision of the proposed algorithm are achieved at 98% in the node coverage optimization with reduced Mean Localization Error (MLE) and computational process time compared with other bio-inspired algorithms, such as Artificial Bee Colony optimization and Genetic Algorithms.

Underwater wireless sensor networks (UWSNs) comprise numerous underwater wireless sensor nodes dispersed in the marine environment, which find applicability in several areas like data collection, navigation, resource investigation, surveillance, and disaster prediction. Because of the usage of restricted battery capacity and the difficulty in replacing or charging the inbuilt batteries, energy efficiency becomes a challenging issue in the design of UWSN. Earlier studies reported that clustering and routing are considered effective ways of attaining energy efficacy in the UWSN. Clustering and routing processes can be treated as nondeterministic polynomial-time (NP) hard optimization problems, and they can be addressed by the use of metaheuristics. This study introduces an improved metaheuristics-based clustering with multihop routing protocol for underwater wireless sensor networks, named the IMCMR-UWSN technique. The major aim of the IMCMR-UWSN technique is to choose cluster heads (CHs) and optimal routes to a destination. The IMCMR-UWSN technique incorporates two major processes, namely the chaotic krill head algorithm (CKHA)-based clustering and self-adaptive glow worm swarm optimization algorithm (SA-GSO)-based multihop routing. The CKHA technique selects CHs and organizes clusters based on different parameters such as residual energy, intra-cluster distance, and inter-cluster distance. Similarly, the SA-GSO algorithm derives a fitness function involving four parameters, namely residual energy, delay, distance, and trust. Utilization of the IMCMR-UWSN technique helps to significantly boost the energy efficiency and lifetime of the UWSN. To ensure the improved performance of the IMCMR-UWSN technique, a series of simulations were carried out, and the comparative results reported the supremacy of the IMCMR-UWSN technique in terms of different measures.

In recent years, the underwater wireless sensor network (UWSN) has received a significant interest among research communities for several applications, such as disaster management, water quality prediction, environmental observance, underwater navigation, etc. The UWSN comprises a massive number of sensors placed in rivers and oceans for observing the underwater environment. However, the underwater sensors are restricted to energy and it is tedious to recharge/replace batteries, resulting in energy efficiency being a major challenge. Clustering and multi-hop routing protocols are considered energy-efficient solutions for UWSN. However, the cluster-based routing protocols for traditional wireless networks could not be feasible for UWSN owing to the underwater current, low bandwidth, high water pressure, propagation delay, and error probability. To resolve these issues and achieve energy efficiency in UWSN, this study focuses on designing the metaheuristics-based clustering with a routing protocol for UWSN, named MCR-UWSN. The goal of the MCR-UWSN technique is to elect an efficient set of cluster heads (CHs) and route to destination. The MCR-UWSN technique involves the designing of cultural emperor penguin optimizer-based clustering (CEPOC) techniques to construct clusters. Besides, the multi-hop routing technique, alongside the grasshopper optimization (MHR-GOA) technique, is derived using multiple input parameters. The performance of the MCR-UWSN technique was validated, and the results are inspected in terms of different measures. The experimental results highlighted an enhanced performance of the MCR-UWSN technique over the recent state-of-art techniques.

Abstract The coastal flood during the tropical cyclone Tauktae, 2021, at Chellanam coast, Kerala, India, has invited wide attention as the wave overtopping severely affected coastal properties and livelihood. We used a combination of WAVEWATCHIII and XBeach to study the coastal inundation during high waves. The effect of low-frequency waves and the rise in the coastal water level due to wave setup caused the inundation at Chellanam, even during low tide with negligible surge height. Wave setup raised the water level at the coast with steep slopes to more than 0.6 m and peaked during low tide, facilitating wave breaking at the nearshore region. The coastal regions adjacent to these steep slopes were subjected to severe inundation. The combined effect of long and short waves over wave setup formed extreme wave runups that flooded inland areas. At gently sloping beaches, the longwave component dominated and overtopped the seawalls and damaged households along the shoreline. The study emphasizes the importance of longwave and wave setup and its interaction with nearshore bathymetry during the high wave. The present study shall lead to the development of a coastal inundation prediction system for the low-lying hot spots using the combination of WAVEWATCHIII and XBeach models.

Recently, human action recognition (HAR) has become an important focus of computer science research because of its applications in surveillance, robotics, sentiment analysis, and other areas. Human activity classification is a time-consuming operation, especially when photos are cluttered, and the background is unclear. In addition, conventional machine learning models fail to achieve robust performance due to the increasing number of activities. Comparatively, deep learning (DL) approaches help to automatically learn and describe the necessary features of the input data with low manual work and robust discriminant abilities. In addition, a preprocessing stage is included in this study's HAR utilizing hyperparameter tuned DL (HAR-HPTDL) model that removes undesired background and improves the quality of the input. The model also implements a bidirectional long short-term memory model as a feature extractor, the sparrow search algorithm to tune the hyperparameters, and a SoftMax layer for the effective classification of human actions. In addition, the curse of dimensionality can be overcome via entropy-based feature reduction and Chi square-based feature selection. Based on a variety of measures, the HAR-HPTDL methodology has been put through its paces with other published techniques. The results indicate that the HAR-HPTDL technique outperforms current state-of-the-art techniques in simulations. The outcome of this work demonstrates that an HAR-HPTDL model may achieve comparable or even superior recognition accuracy of 0.949 than the prior best deep classifier(s) on all databases with proper parameter optimization. [ABSTRACT FROM AUTHOR]