Blockchain-Federated TrustGuardNet Model for Robust and Privacy-Preserving Consumer IIoT Intelligence

The exponential expansion of Industrial Internet of Things (IIoT) networks and consumer technology ecosystems has introduced complex challenges in securing distributed data sharing while preserving user privacy and system trust. Traditional federated learning (FL) frameworks often falter under adversarial model poisoning, gradient inference attacks, and inefficient consensus mechanisms, especially in heterogeneous, resource-constrained IIoT and consumer technology environments. To address these limitations, this study proposes TrustGuardNet, a novel blockchain-enhanced FL model that enables intelligent, privacy-aware, and secure collaboration across IIoT and consumer technology platforms. TrustGuardNet integrates a lightweight and trust-centric Adaptive Reputation-Aware Proof-of-Contribution (AR-PoC) mechanism to promote fairness and reliability during model aggregation. To mitigate gradient-level privacy threats, it deploys the AGOSA pipeline, blending differential privacy, secure multiparty computation, and adaptive gradient obfuscation (AGO). Further defense against malicious contributions is achieved via the TWRAD module, which incorporates trust-weighted aggregation, anomaly detection, and clustering-based gradient filtering. Experimental evaluation across real-world IIoT and consumer technology scenarios demonstrates superior performance, achieving 97% accuracy, 98% detection rate, and a 0.99 privacy preservation score, while maintaining low computational overhead and strong energy efficiency. By harmonizing blockchain trust mechanisms with decentralized learning, TrustGuardNet delivers a scalable, resilient, and ethically grounded approach for the future of secure industrial and consumer intelligence.