A Blockchain-Enabled Privacy-Preserving IoT Framework With Domain-Adaptive Anomaly Detection in Zero-Trust Environments

Although the fast growth of Internet of Things (IoT) across multiple industries has significantly enhanced operational efficiency, it also brings major security and privacy management obstacles—particularly in zero-trust environments. Centralized security systems demonstrate built-in scalability constraints and susceptibility to attacks, whereas blockchain solutions provide decentralized audit capabilities but face challenges with latency and storage capacity limitations. Our work advances IoT security and privacy protection through a blockchain-based framework that combines a permissioned blockchain system with sophisticated cryptographic protocols (including post-quantum cryptography) and implements off-chain storage with IPFS alongside an adversarial domain adaptation method for real-time anomaly detection. Our approach addresses common criticisms found in the literature by publishing detailed simulation parameters and comparing our system with leading frameworks—such as vanilla PoW, Hyperledger Fabric, and DAG-based chains (e.g., IOTA)—while assessing performance through NS-3 network simulations. We include extensive threat model analyses that examine collusion and Sybil attacks on off-chain storage nodes as well as quantum computing risks, and we confirm anomaly detection performance with standardized datasets (e.g., N-BaIoT, UNSW-Sensor) using robust statistical measures. Our experiments demonstrate up to a 35 % reduction in end-to-end latency (compared to baseline PoW systems), a 40% increase in throughput (compared to baseline PoW systems), and a 70% reduction in on-chain storage overhead (compared to direct on-chain storage without IPFS and coding) while maintaining anomaly detection accuracy above 90 % under domain shifts. These performance improvements highlight the potential of our approach to enable secure, real-time monitoring in resource-constrained deployments such as remote patient health tracking, industrial automation lines, and large-scale smart city sensor networks. Comprehensive testing on platforms like Raspberry Pi 4 and ESP32 demonstrates that the framework delivers strong performance and reliability even under strict zero-trust security requirements.