This note focuses on the convergence problem of a distributed recursive least squares (RLS) estimator, which converges fast in many circumstances. We find that the convergence behavior of this distributed RLS is quite different for the scalar-parameter and high-dimensional-parameter cases. 

　　Publication: 

　　IEEE Transactions on Automatic Control (Volume: 67, Issue: 12, December 2022) 

　　DOI: 10.1109/TAC.2022.3194882 

　　Author: 

　　Zhaobo Liu 

　　Key Laboratory of Systems and Control, Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing, China 

　　School of Mathematical Sciences, University of Chinese Academy of Sciences, Beijing, China 

　　Chanying Li 

　　Key Laboratory of Systems and Control, Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing, China 

　　School of Mathematical Sciences, University of Chinese Academy of Sciences, Beijing, China 

　　Email: cyli@amss.ac.cn