文献类型：期刊文献

英文题名：HMM Noise Injection for Privacy Protection in a Cooperative User-Server LQG Control System

作者：Lu, Mengqi[1]; Ni, Yuqing[1]; Yang, Wen[1]; Yang, Chao[1]

机构：[1] East China University of Science and Technology, Key Laboratory of Smart Manufacturing in Energy Chemical Process, Ministry of Education, Dept. of Automation, School of Information Science and Engineering, Shanghai, 200237, China

年份：2026

外文期刊名：IEEE Transactions on Automatic Control

收录：EI(收录号：20262721048254);Scopus(收录号：2-s2.0-105043746143)

语种：英文

外文关键词：Background noise - Optimization - Privacy-preserving techniques

摘要：In this paper, we address the challenge of privacy preservation in a cooperative user-server LQG control system comprising a user and a server. To protect the user's privacy, we propose a scheme of adding noise to the state measurements sent to the server, with the noise generated by a hidden Markov model (HMM). This noise introduces a deviation in the server's state estimation, thereby enhancing the privacy performance. We also propose a privacy metric to quantify this deviation. Furthermore, we analyze the impact of our privacy scheme on both privacy performance and potential service loss in LQG control. Finally, we formulate an optimization problem to balance the tradeoff between privacy protection and LQG control performance, utilizing a novel frequency domain method to solve the hidden Markov model for generating appropriate noise. ? 1963-2012 IEEE.