Equipment failure data trends focused privacy preserving scheme for Machine-as-a-Service

Abstract

In the Machine-as-a-Service (MaaS) model, enterprises lease equipment from original equipment manufacturer (OEM) to reduce production costs, and share equipment failure data to assist OEM improve equipment quality. However, The failure data trends formed by the frequency of multi-type failure may reveal private information about the enterprises. Most previous studies did not consider the issue of privacy leakage through failure data trends. Therefore, we propose an equipment failure data trends focused privacy preserving scheme for MaaS to prevent the leakage of enterprise privacy data trends. Firstly, our scheme safeguards the multi-dimensional data privacy of equipment through local differential privacy. Secondly, the differential privacy mechanism is integrated into the blockchain to build a fine-grained privacy-preserving categorical query algorithm for enterprise privacy, which decouples the correlation between failure data trends and enterprise privacy. Finally, theoretical analysis proves the privacy preservation capabilities of our scheme. The experimental analysis confirms that our scheme effectively protects data trends privacy, and the results indicate that our scheme has lower computational and time expenditures compared to similar schemes.