Blockchain and digital twin empowered trustworthy self-healing for Edge-AI enabled Industrial Internet of Things

Xinzheng Feng,Jun Wu,Yulei Wu,Jianhua Li,Wu Yang

DOI: https://doi.org/10.1016/j.ins.2023.119169

IF: 8.1

2023-05-19

Information Sciences

Abstract:The public has regarded Edge-AI enabled Industrial Internet of Things (IIoT) as the crucial foundation in the intelligent digital factories in Industry 4.0. It can fully catch the massive production data derived from the complex production process, and provide efficient, intelligent services. However, the deployment of edge AI aggravates the complexity and security risks caused by the massive heterogeneous resource-constrained and vulnerable edge IIoT devices. Effective fault prevention is crucial to ensure the security and robustness of the IIoT with numerous vulnerable edge devices. Most existing solutions are based on the history log, which can hardly defend against attacks and is easy to cause excessive maintenance. To address this issue, we propose a trustworthy self-healing scheme based on the combination of distributed digital twin (DT) and blockchain, to ensure the security and robustness of the industrial system network. We first propose an implementation architecture of the distributed DT based self-healing IIoT to apply the distributed DT simulation capability fully. In addition, we provide a DT simulation operating mechanism for the controlled industrial devices, considering the requirement of users and constrained resources of edge servers. Moreover, this work proposes a blockchain-based decentralized trust management mechanism to ensure the reliability of self-healing. The security analysis and performance evaluation show the security and efficiency of our proposal.

computer science, information systems