Secured Routing Protocol for Improving the Energy Efficiency in WSN Applications

A staggering number of applications rely on the network architecture to carry out their tasks, which has led to a fast growth in wireless sensor networks (WSN). The possibility of harmful activity and data theft is growing as a result of the growth in devices and data. Thus, the network’s regular users have an impact on legitimate data delivery, which lowers customer happiness and worsens network standards. The data have been saved using a variety of security procedures that have been developed in past research studies. However, harmful activity continues to engage in its illegal operations despite their efforts to safeguard data transmission in the network. As a result, a number of recent research projects have concentrated on predicting innovative techniques and processes to offer security in WSN. In comparison to existing methods, this work attempted to offer an effective tighter security for WSN and suggested an ML-Based Secured Routing Protocol (MLSRP) for WSN with improved energy efficiency and overall performance. Energy efficiency is the main requirement of WSNs, hence a clustered network is proposed where the data are routed through the cluster head nodes. In this paper, a multicriteria based decision-making (MCDM) model is used by the MLSRP to perform data routing, clustering, and cluster head election while also analyzing a number of network characteristics that might affect the quality of a node, route, and data. In NS2 software, the suggested framework is put into practice and simulated. The results are then validated to gauge performance. The observed quantitative results reveal that the proposed MLSRP method attains an improved network lifetime by 5% and network throughput of 6%. It reduces energy consumption by 40%, curtails overhead to 37%, and minimizes end-to-end delay by 30% than the other conventional methods. The suggested framework performs better than others when its total performance is compared to that of older methods.