Advanced Mathematical Modelling for Energy-Efficient Data Transmission and Fusion in Wireless Sensor Networks

Abstract: Wireless Sensor Networks (WSNs) are indispensable for data-intensive applications, necessitating efficient energy management and robust data fusion techniques. This paper proposes an integrated framework leveraging fuzzy logic and backpropagation neural networks (BPNN) to enhance energy efficiency and data accuracy in WSNs. The model focuses on optimizing Cluster Head (CH) selection using fuzzy logic, considering parameters such as energy levels, proximity to the base station, and local density centrality. A Minimum Spanning Tree (MST) algorithm is employed for energy-efficient data transmission from sensor nodes to CHs, minimizing energy consumption during data routing. BPNN-based data fusion at CHs reduces redundant data transmissions to the base station, thereby optimizing energy utilization and enhancing overall network performance. Simulation results demonstrate substantial improvements over conventional methods, including a 30% increase in network longevity, 25% improvement in data accuracy, and a 40% reduction in energy consumption. These gains are attributed to the intelligent CH selection strategy enabled by fuzzy logic, which ensures efficient resource allocation and minimizes energy wastage. The effectiveness of the proposed approach is validated through comprehensive simulations, showcasing its ability to prolong network lifetime, maintain data integrity, and improve energy efficiency. The integration of fuzzy logic and BPNN not only addresses the challenges of energy management and data fusion in WSNs but also provides a scalable and adaptable framework for future applications requiring reliable and sustainable sensor network operations. The source code is available at https://github.com/hikomal/BPNN_WSN.git