Probabilistic Trust Inference Theory to Optimizing Multi-Level Trust in Software Defined Networks

In the evolving landscape of Software Defined Networks (SDNs) and Wireless Sensor Networks (WSNs), ensuring secure and reliable communication is paramount, particularly in the presence of potential malicious nodes and denial-of-service (DoS) attacks. This paper proposes a novel model named Probabilistic Trust Inference Theory (PTIT) for optimizing multi-level trust management with Enhanced Lion Optimization (ELO) for routing. Our model incorporates a dynamic trust evaluation framework that leverages both direct and indirect interactions to compute trust scores through a Bayesian inference mechanism. By establishing multiple trust levels, we enhance the granularity of trust assessments, facilitating the formation of robust trust-based clusters that mitigate the risk of attacks. Each cluster is led by a high-trust node, ensuring reliable data aggregation and communication pathways. The proposed routing protocol employs ELO to integrate trust considerations into path selection, significantly reducing the likelihood of routing through untrustworthy nodes while defending against malicious activities and DoS attacks. We evaluate the performance of our approach through extensive simulations, demonstrating substantial improvements in network throughput, latency, and energy efficiency compared to existing trust management frameworks. Our findings highlight the effectiveness of PTIT-ELO in creating a resilient and adaptive trust management system that enhances the security and reliability of WSNs against malicious threats. This work contributes to the growing body of research aimed at fortifying network infrastructures while optimizing overall performance.