Lightweight Game-Theoretic Trust Aggregation and Bio-Inspired Rerouting for Secure Industrial Wireless Sensor Networks

Abstract

Industrial Wireless Sensor Networks (IWSNs) deployed in smart manufacturing and critical process automation face extreme resource constraints alongside heightened exposure to targeted physical and routing attacks. Conventional cryptographic mechanisms impose prohibitive memory and latency penalties, whereas existing soft-security schemes frequently succumb to collusion and bad-mouthing attacks. This paper presents a novel Game-Theoretic Trust Aggregation and Bio-Inspired Rerouting (GTA-BR) framework for IWSNs. The proposed architecture models node interaction dynamics using a non-cooperative game formulation to derive Bayesian-updated Direct Trust Values (BDTV) while filtering dishonest third-party recommendations. A dynamic fuzzy-clustering stage selects robust Cluster Heads (CHs) based on energy density and trust variance. Inter-cluster routing is resolved via a Multi-Objective Grey Wolf Optimizer (MOGWO) that balances path trust ratings, residual energy levels, and hop latency. Extensive NS-3 simulations under 0% to 30% malicious node density show that GTA-BR maintains a Packet Delivery Ratio (PDR) exceeding 95.1%, reduces end-to-end latency to under 98 ms, and extends total operational network lifetime by 31.2% relative to contemporary baseline protocols.