A Privacy-Preserving Framework for IoT Using Behavior Obfuscation and Differential Privacy in Blockchain–Cloud Architectures
Year: 2025
Type: article
Abstract: Rapidly growing Internet of Things (IoT) networks pose many fundamentally important privacy challenges, particularly when combined with transparent technologies such as the blockchain. Although blockchain provides distributed trust and auditability, its immutable logs can inadvertently leak user behavior. In this paper, we take on this problem by presenting a privacy-preserving approach for IoT systems, which is informed by the confluence of behavior obfuscation and differential privacy technology, run atop a hybrid blockchain-cloud infrastructure. The research result is a five-phase model which makes devices’ behaviors look like exactly-anonymous devices at the edge and is based on formal privacy guarantees by employing differential privacy in the cloud. To achieve this, we propose a framework that consists of light-weight obfuscation operators—timestamp perturbation, dummy event injection, and activity reordering—that, when synergistically used with the state-of-the-art, media-centric, differentially private solutions in the cloud, offer justifiable privacy against analytical and knowledge-based attacks. Cleaned–up summaries are tentatively submitted to blockchain in a safe way through smart contracts, facilitating traceable, tamper–evident logging without revealing raw behavioral information obtained from the data owner. Experimental results based on a Raspberry Pi-based IoT testbed and Ethereum-compatible blockchain show that the proposed system keeps low latency ( 2.8 s), over 400 events/min throughput, blockchain storage overhead and privacy leakage. Modular, auditable, and real-time-capable approach for secure data sharing in adversarial IoT environments (Applied to healthcare, smart homes, and industrial monitoring).
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7160896841