Minal Patil

Background: In rural and forest-bordering areas, human-wildlife conflict has escalated, resulting in substantial crop damage, property loss, and human mortality. Conventional mitigating solutions such as fencing, manual surveillance, and local alarms, have limited coverage, time taking and high operating costs. Automated real-time animal detection systems have become a practical for early warning and conflict avoidance due to recent upgrades in edge computing and deep learning particularly in distant locations without dependable internet access. Methods: In this study, embedded hardware and deep learning-based object detection are used to create a real-time wild animal detection and warning system. Roboflow was used to develop and enhance a dataset of 1,554 real-world photographs from seven classes (tiger, bear, elephant, lion, cheetah, leopard, and human) to 5,917 images. Using GPU acceleration, three object detection models YOLOv8(n)(n), SSD, and RetinaNet were trained on Google Colab and assessed using precision, recall, F1-score, and mean Average Precision (mAP). A GSM 900 A module is used to send SMS notification with no requirement of internet connectivity, and the optimized model was installed on a Raspberry Pi 5 with a Pi Camera for real-time inference. Results: According to experimental data, YOLOv8(n) performs better than SSD and RetinaNet, with an mAP@50 84.4% and recall of 81% as opposed to SSD’s 62.3% precision and 58% recall and RetinaNet’s around 77% precision and 80% recall. Additionally, YOLOv8(n) showed improved robustness and quicker inference time across a range of background and lighting circumstances. While SSD and RetinaNet demonstrated weaker or scattered attention in complex situations, explainable AI analysis utilizing Grad-CAM and LIME verified greater and more concentrated activation on pertinent animal traits for YOLOv8(n). Conclusion: For real-time wildlife monitoring and early warning, the suggested system offers a precise, affordable, and scalable option. Through the integration of YOLOv8(n), Raspberry Pi 5 edge computing, and GSM-based connectivity, the system efficiently functions in distant settings and dramatically shortens reaction times during wildlife infiltration incidents. The findings demonstrate that YOLOv8(n) is the best model for integrated animal detection, enhancing wildlife conservation, lowering agricultural loss, and improving human safety.

In the ad-hoc network, where we have mobile and wireless connections, traditional grouping, generally faced with many challenges, including the weakness of memory in Local and ineffective energy use. These problems lead to the reduction of network life and service quality (QOS). Moreover, in a large network that has a high movement of node changes, frequent weaving, and energy levels that require volunteers to increase complexity to maintain the stability of the network and effective communication. To meet these challenges, the algorithm of the gray wolf enhancement (GWOCA) focuses on the use of energy and choosing the best cluster (CHS). GWOCA procures the ideal cluster center. Guaranteed a balanced load distribution and the ability to adjust the size quickly, even for large and dynamic networks. Moreover, the cluster definition protocol (CBRP) has been developed to improve the efficiency of determination by organizing the structure itself as a group and selecting CHs for internal communication and between effective groups. The preliminary simulation operated using the NS2 model, showing that the GWOCA cluster is significantly higher than the traditional grouping methods, including cluster, cluster, Parent Cluster, and CBRP, in the main network performance. Our research states that GWOCA can reduce the delay by 30.66%, reduce energy consumption by 4.55% and the PDR (PDR) ratio is reduced by 10% compared to general methods. When compared to the method, the method that is presented is also evaluated for the ability to expand, which effectively helps to ensure the use of reliable energy and sending packets. And show stable performance under the effective node density. In addition, in-depth energy analysis for different network phases—Creating a cluster. The route and maintenance have been implemented. By emphasizing the importance of energy saving that can be done by our modified group. These results confirm that the proposed GWOCA-based clustering method presents a robust, scalable solution for energy-efficient, QoS-aware MANETs, capable of addressing the critical challenges of energy depletion, topology changes, and scalability, making it suitable for real-world applications in dynamic, mobile, and resource-constrained environments.

Mobile Ad-Hoc Networks are incapable of energizing themselves due to their limited energy. The effort is to develop an energy-adequate power management plan for the MANET. Cluster heads may malfunction or operate incorrectly as a result of power problems while based on different cluster routing methods. Consequently, during information collecting and interaction, the cluster heads encounter instability. Finding the unstable cluster heads and swapping out for another node to use the re configurable clustering technique is the primary goal of this study. In order to correctly define the cluster heads, the proposed a Scikit-Learn's K-Means clustering method. The anomalous or superfluous modifications in cluster heads and the shift in the cluster nodes in the count are detected by the suggested in Scikit-Learn's K-Means technique. The proposed work represents Scikit learning k-means clustering in MANET and calculates QoS parameters such as Energy, Throughput, Delay and Packet delivery ratio. By leveraging Scikit-Learn's K-Means Clustering (SLKMC), a novel approach in MANET can achieve an optimized trade-off between energy, delay, PDR, and Throughput, making it a practical and efficient choice for QoS enhancement as Energy efficiency is expected to be up to 10-30% energy savings by reducing redundant communication. The delay is reduced with anticipation of a 15-25% decrease in average delay with efficient cluster-based routing. Packet Delivery Ration might improve by 5-20%, ensuring more reliable data delivery. Throughput is improved coordination and reduced collisions can enhance Throughput by 10-25%. Thus, the expected benefits quantify the impact of these enhancements in terms of QoS metrics and improve Network Performance.

Clustering is a highly structured analysis for coordinating routing and improving network performance in MANET using different attributes and metrics. The collaboration and disjoint of mobile nodes from various clusters in the network (MANETs) with their movable nodes' behavior disrupts the network structure unstable, and this issue becomes more severe if the nodes are Cluster Heads (CH). In this article comprehensively investigates of clustering schemes and classifies them into various categories based on the Cluster Head (CH) selection criteria. It helps other researchers to understand how individual forms of clustering algorithms differ from one another. In this review article, the authors emphasize energy and mobility as pivotal criteria for CH selection and cluster maintenance. The article has used quality of service (QoS) indicators to evaluate the efficacy of different existing systems. In research carried out by the different systems, the authors have figured out some key QoS metric commutation and detected several critical factors influencing the effectiveness of clustering methods. Generally, each cluster in the network routing protocol of a MANET includes a CH and multiple neighbor nodes. The benefits of this approach are evident, including higher scalability and stability, reduced management overhead, and a decrease in the number of nodes participating in the same channel.

Manet have several nodes that are linked together wirelessly and the nodes are mobile in nature. Nodes position is flexible. The primary issue with the traditional clustering method is that it is prone to become trapped in the regional optimum path. While the nodes are being transmitted or received, energy is used. Through the use of the GWOGA clustering technique, CHs collect data from cluster participants and communicate other nodes with the cluster. While choosing the best CH to lengthen the network's lifespan is the Manet is most vital responsibility. Clustering based on the Grey Wolf Optimization and Global Algorithm (GWOGA), an attempt has been made to address this issue. GWOGA search function is employed for the best cluster centres in the supplied feature span. The cluster centres are encoded using the agent representation. When choosing the best CHs, the GWOGA dynamically balances the process of increasing and diversifying search. In addition, choosing the best CHs for the network is aided by factors like node degree, energy, distance, node centrality, Throughput, Delay, Path Loss Ratio, etc. The computational findings show that GWOGA offers improved values of excellent throughput, packet delivery ratio, end delay, energy a better lifespan segregated with the performance of normal clustering method. NS2 simulator is used for the simulation for finding QoS parameters.