Trust-Assisted Blockchain and Machine Learning-Enhanced Energy-Aware Secure Routing for Hierarchical WSNs

Abstract

As Wireless Sensor Networks (WSNs) advance towards mission-critical applications, routing frameworks must be safe, flexible, and energy-efficient (EE).Conventional clustering-based protocols (CBP) such as LEACH lack resilience against dynamic security threats and offer limited scalability in hostile environments. This paper introduces a novel framework titled Trust-Assisted Blockchain and Fuzzy SVM-Enhanced Energy-Aware Secure Routing (T-BMESR) for hierarchical WSNs. The proposed method combines adaptive trust modelling (ATM), lightweight blockchain (BC) consensus using Proof-of-Authority (PoA) and Directed Acyclic Graph (DAG), and an intelligent anomaly detection mechanism based on Fuzzy based Support Vector Machine (FBSVM). FSVM enhances classification by assigning fuzzy membership values to sensor nodes (SN), improving decision confidence in detecting malicious behavior. Trust values (TV) are dynamically updated using a fusion of direct interactions, indirect recommendations, and FBSVM-based anomaly scores. BC ensures low-overhead validation of data and node identities, securing the routing path. Simulation-based evaluations demonstrate that T-BMESR significantly outperforms baseline protocols in terms of EE, detection accuracy, throughput (T), and trust precision. The integration of FSVM and BC enables robust, scalable, and trustworthy communication across energy-constrained WSN environments.