A Trust Aware Authentication Scheme for Wireless Sensor Networks Optimized by Salp Swarm Optimization and Deep Belief Networks

Abstract

Presently, the integration of Internet of Things (IoT) and wireless sensor networks (WSN) offers a broad research field for enabling advanced networked services. It remains popular due to its applicability in various real time areas such as healthcare, environmental monitoring, factory configuration, and many more. While the benefits of WSNs are many, security is still a major concern due to the intrinsic prevalence of wireless links in the network. In order to achieve security and reliable communication, an optimized authentication scheme becomes necessary. Therefore, this research work introduces a novel salp swarm optimization with deep belief network based trust aware authentication (SSDBN-TAA) scheme for WSN. Primarily, the SSDBN-TAA technique undergoes a weighted clustering scheme to partition the network into a collection of clusters. Additionally, a trust factor is collectively derived between the nodes that exist in the network, and the nodes exceeding the threshold trust value are considered as valid. An SSDBN model is utilized for dynamically selecting the threshold trust value, and the hyperparameters of the DBN model are optimally adjusted using the salp swarm algorithm (SSA). The design of SSA is efficient and thereby enhances the authentication performance. To explore the enhanced outcomes of the SSDBN-TAA technique, we conduct extensive comparative experiments to ensure the enhanced outcomes of the SSDBN-TAA system dominate the present state of art approaches.