Hybrid optimization-based clustering and cross-layer channel access routing with wavelet-based Elman network for effective transmission in wireless sensor network

Hybrid optimization-based clustering and cross-layer channel access routing with wavelet-based Elman network for effective transmission in wireless sensor network

연구 분야: Infrastructure

학회: International Journal of Data Science and Analytics

With the fast growth of the Internet of Things, Wireless Sensor Networks are being adopted across a wide range of fields, such as defence uses, environment-related control, medical care and manufacturing fabrication, transportation administration, and more. Because of the traits of limited resources and dynamic topology, WSNs are encountering significant challenges like security and power utilization. Various routing protocols have been developed, but these approaches are complicated, leading to configuration errors and causing incorrect routing. Thus, a deep learning-based attack prediction and effective routing protocol is developed to enable data transmission that is secure and energy-efficient in this research. The sensor nodes are first positioned randomly within the experimental area. Once the nodes are deployed, identify the attack nodes from normal nodes using a Wavelet-Based Elman Neural Network. Identified attack nodes are subsequently isolated from the network to prevent unauthorized access. The remaining legitimate nodes are then organized into clusters using an Optimized Threshold Sensitive Adaptive Efficient Clustering approach to reduce the energy consumption. In this clustering process, the cluster head is selected using a hybrid of the Coyote Optimization Algorithm and Magnificent Frigate Bird Optimization, depending on criteria such as power and node importance. After clustering, the system employs the Efficient Cross-Layer Channel Access and Routing Protocol to identify the optimal path for efficient data transmission from the source to the destination node. The proposed method achieves an attack detection accuracy of 95.20%, and a positive predictive value of 94% while improving overall security and preserving energy in the network.