A Novel Routing Protocol for Data Transmission in WSNs with Security enhancement and Extended Lifetime using A Hybrid Deep Learning and Optimization Approach

Abstract

In many applications like smart infrastructure, healthcare facilities, army surveillance, and environmental monitoring, the Wireless Sensor Network (WSN) is crucial. The effectiveness of the WSN is limited by the following factors: low power sources, poor routing, security susceptibilities, prolonged delay, and the crucial demands in preserving Quality of Service (QoS). Optimizing QoS and robustness are the main emphasis of conventional methods. These conventional methods fail to offer this solution. This comprehensive solution is an integrated strategy that tackles a challenging issue by simultaneously considering several significant variables. For resolving these issues, a novel routing protocol (RP) is suggested in this study. Implementation of this suggested method may support in improving data transmission (DT), security, and the lifetime of WSN. The process begins with data pre-processing method. For effectively training deep learning (DL) module, this suggested method utilizes Robust Scaling approach. The cluster head (CH) is selected by utilizing hybrid Particle Swarm Optimization-Fuzzy C-Means (PSO-FCM) clustering technique, and this may result in optimizing the cluster formation process of the network. After the clustering method, route discovery method is implemented. The Graph neural network (GNN) and symmetric butterfly optimization algorithm (SBOA) are integrated in this route discovery method. This GNN facilitate precise data classification. The dynamic routing optimization is done via SBOA. For securing networks from these cyber threats, blockchain (BC)-based authentication is used. Thus, overall security of the system is improved. This hybrid method promotes in data management, energy efficiency (EE), and network security. The suggested strategy considerably increases the operational lifespan of WSNs when compared to current methods, and it was demonstrated by simulation outcomes.