{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T14:41:42Z","timestamp":1787323302145,"version":"build-2736575974"},"reference-count":49,"publisher":"Society for Industrial & Applied Mathematics (SIAM)","issue":"6","funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62203179"],"award-info":[{"award-number":["62203179"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62222205"],"award-info":[{"award-number":["62222205"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62173049"],"award-info":[{"award-number":["62173049"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000181","name":"AFOSR","doi-asserted-by":"publisher","award":["FA9550-21-1-0231"],"award-info":[{"award-number":["FA9550-21-1-0231"]}],"id":[{"id":"10.13039\/100000181","id-type":"DOI","asserted-by":"publisher"}]},{"name":"ARO","award":["W911NF2310266"],"award-info":[{"award-number":["W911NF2310266"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["SIAM J. Control Optim."],"published-print":{"date-parts":[[2024,12,31]]},"abstract":"<jats:p>Abstract.<\/jats:p>\n                  <jats:p>We propose an online learning algorithm that adaptively designs a decentralized linear quadratic regulator when the system model is unknown a priori and new data samples from a single system trajectory become progressively available. The algorithm uses a disturbance-feedback representation of state-feedback controllers coupled with online convex optimization with memory and delayed feedback. Under the assumption that the system is stable or given a known stabilizing controller, we show that our controller enjoys an expected regret that scales as [Formula: see text] with the time horizon [Formula: see text] for the case of partially nested information pattern. For more general information patterns, the optimal controller is unknown even if the system model is known. In this case, the regret of our controller is shown with respect to a linear suboptimal controller.<\/jats:p>","DOI":"10.1137\/23m161714x","type":"journal-article","created":{"date-parts":[[2024,12,16]],"date-time":"2024-12-16T04:17:16Z","timestamp":1734322636000},"page":"3341-3368","source":"Crossref","is-referenced-by-count":2,"title":["Learning Decentralized Linear Quadratic Regulators with \\({\\sqrt{T}}\\) Regret"],"prefix":"10.1137","volume":"62","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8608-5815","authenticated-orcid":true,"given":"Lintao","family":"Ye","sequence":"first","affiliation":[{"name":"Key Laboratory of Image Processing and Intelligent Control, Ministry of Education and School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan 439974 China."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ming","family":"Chi","sequence":"additional","affiliation":[{"name":"Corresponding author. Key Laboratory of Image Processing and Intelligent Control, Ministry of Education and School of Artificial Intelligence and Automation, Huazhong University of Science and Technology, Wuhan 439974 China."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ruiquan","family":"Liao","sequence":"additional","affiliation":[{"name":"College of Petroleum Engineering, Yangtze University, Wuhan 434023 China."}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Vijay","family":"Gupta","sequence":"additional","affiliation":[{"name":"Elmore School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN 47907 USA."}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"351","published-online":{"date-parts":[[2024,12,16]]},"reference":[{"key":"ref1","unstructured":"Y. Abbasi-Yadkori and C. Szepesv\u00e1ri, Regret bounds for the adaptive control of linear quadratic systems, in Proceedings of the 24th Conference on Learning Theory 19, 2011, pp. 1\u201326."},{"key":"ref2","unstructured":"N. Agarwal, B. Bullins, E. Hazan, S. Kakade, and K. Singh, Online control with adversarial disturbances, in Proceedings of the 36th International Conference on Machine Learning 97, 2019, pp. 111\u2013119."},{"key":"ref3","first-page":"784","volume":"28","author":"Anava O.","year":"2015","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1016\/S0005-1098(00)00050-9"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1017\/CBO9780511804441"},{"key":"ref6","unstructured":"J. Bu, A. Mesbahi, M. Fazel, and M. Mesbahi, LQR Through the Lens of First Order Methods: Discrete-Time Case, preprint, arXiv:1907.08921, 2019."},{"key":"ref7","series-title":"Lecture Notes","volume-title":"Introduction to Online Optimization","author":"Bubeck S.","year":"2011"},{"key":"ref8","unstructured":"A. Cassel, A. Cohen, and T. Koren, Logarithmic regret for learning linear quadratic regulators efficiently, in Proceedings of the 37th International Conference on Machine Learning, PMLR 119, 2020, pp. 1328\u20131337, https:\/\/proceedings.mlr.press\/v119\/cassel20a.html."},{"key":"ref9","unstructured":"A. B. Cassel and T. Koren, Online policy gradient for model free learning of linear quadratic regulators with \\(\\sqrt{T}\\) regret, in Proceedings of the 38th International Conference on Machine Learning, PMLR 139, 2021, pp. 1304\u20131313, https:\/\/proceedings.mlr.press\/v139\/cassel21a.html."},{"key":"ref10","unstructured":"X. Chen and E. Hazan, Black-box control for linear dynamical systems, in Proceedings of the 34th Conference on Learning Theory, 2021, pp. 1114\u20131143, https:\/\/proceedings.mlr.press\/v134\/chen21c.html."},{"key":"ref11","unstructured":"A. Cohen, T. Koren, and Y. Mansour, Learning linear-quadratic regulators efficiently with only \\(\\sqrt{T}\\) regret, in Proceedings of the 36th International Conference on Machine Learning 97, 2019, pp. 1300\u20131309, https:\/\/proceedings.mlr.press\/v97\/cohen19b.html."},{"key":"ref12","unstructured":"S. Dean, H. Mania, N. Matni, B. Recht, and S. Tu, Regret bounds for robust adaptive control of the linear quadratic regulator, in Proceedings of the 32nd International Conference on Neural Information Processing Systems (NIPS 2018), 2018, pp. 4192\u20134201, https:\/\/dl.acm.org\/doi\/10.5555\/3327144.3327332."},{"key":"ref13","doi-asserted-by":"crossref","unstructured":"S. Fattahi, N. Matni, and S. Sojoudi, Learning sparse dynamical systems from a single sample trajectory, in 2019 IEEE 58th Conference on Decision and Control, IEEE, 2019, pp. 2682\u20132689.","DOI":"10.1109\/CDC40024.2019.9029815"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1137\/19M1291108"},{"key":"ref15","unstructured":"M. Fazel, R. Ge, S. Kakade, and M. Mesbahi, Global convergence of policy gradient methods for the linear quadratic regulator, in Proceedings of the 35th International Conference on Machine Learning, PMLR 80, 2018, pp. 1467\u20131476."},{"key":"ref16","doi-asserted-by":"crossref","unstructured":"H. Feng and J. Lavaei, On the exponential number of connected components for the feasible set of optimal decentralized control problems, in 2019 American Control Conference, IEEE, 2019, pp. 1430\u20131437.","DOI":"10.23919\/ACC.2019.8814952"},{"key":"ref17","unstructured":"L. Furieri, Y. Zheng, and M. Kamgarpour, Learning the globally optimal distributed LQ regulator, in Proceedings of the 2nd Conference on Learning for Dynamics and Control, PMLR 120, 2020, pp. 287\u2013297, https:\/\/proceedings.mlr.press\/v120\/furieri20a.html."},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1137\/120880811"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2020.3037046"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.1972.1099850"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1017\/CBO9781139020411"},{"key":"ref22","unstructured":"S. Lale, K. Azizzadenesheli, B. Hassibi, and A. Anandkumar, Logarithmic Regret Bound in Partially Observable Linear Dynamical Systems, preprint, arXiv:2003.11227, 2020."},{"key":"ref23","doi-asserted-by":"crossref","unstructured":"A. Lamperski and J. C. Doyle, Dynamic programming solutions for decentralized state-feedback LQG problems with communication delays, in 2012 American Control Conference (ACC), IEEE, 2012, pp. 6322\u20136327, https:\/\/doi.org\/10.1109\/ACC.2012.6315282.","DOI":"10.1109\/ACC.2012.6315282"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1016\/j.automatica.2015.05.010"},{"key":"ref25","unstructured":"J. Langford, A. Smola, and M. Zinkevich, Slow Learners Are Fast, preprint, arXiv:0911.0491, 2009."},{"key":"ref26","unstructured":"Y. Li, S. Das, J. Shamma, and N. Li, Safe Adaptive Learning-Based Control for Constrained Linear Quadratic Regulators With Regret Guarantees, preprint, arXiv:2111.00411, 2021."},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2021.3128592"},{"key":"ref28","first-page":"1","volume":"21","author":"Malik D.","year":"2020","journal-title":"J. Mach. Learn. Res."},{"key":"ref29","unstructured":"H. Mania, S. Tu, and B. Recht, Certainty equivalence is efficient for linear quadratic control, in Proceedings of the 33rd International Conference on Neural Information Processing Systems, ACM, 2019, pp. 10154\u201310164."},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2021.3087455"},{"key":"ref31","doi-asserted-by":"crossref","unstructured":"Y. Ouyang, M. Gagrani, and R. Jain, Learning-Based Control of Unknown Linear Systems With Thompson Sampling, preprint, arXiv:1709.04047, 2017.","DOI":"10.1109\/ALLERTON.2017.8262873"},{"key":"ref32","doi-asserted-by":"crossref","unstructured":"A. Quattoni, X. Carreras, M. Collins, and T. Darrell, An efficient projection for \\(\\ell_{\\infty }\\) regularization, in ICML \u201909: Proceedings of the 26th Annual International Conference on Machine Learning, ACM, 2009, pp. 857\u2013864, https:\/\/doi.org\/10.1145\/1553374.1553484.","DOI":"10.1145\/1553374.1553484"},{"key":"ref33","first-page":"1984","volume":"50","author":"Rotkowitz M.","year":"2005","journal-title":"IEEE Trans. Automat. Control"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2011.2173411"},{"key":"ref35","first-page":"1","volume":"22","author":"Sarkar T.","year":"2021","journal-title":"J. Mach. Learn. Res."},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2013.2281881"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1137\/22M1489836"},{"key":"ref38","unstructured":"M. Simchowitz and D. Foster, Naive exploration is optimal for online LQR, in Proceedings of the 37th International Conference on Machine Learning, PMLR 119, 2020, pp. 8937\u20138948, https:\/\/proceedings.mlr.press\/v119\/simchowitz20a.html."},{"key":"ref39","unstructured":"M. Simchowitz, K. Singh, and E. Hazan, Improper learning for non-stochastic control, in COLT 2020 Conference on Learning Theory, 2020, pp. 3320\u20133436."},{"key":"ref40","unstructured":"S. Tu and B. Recht, The gap between model-based and model-free methods on the linear quadratic regulator: An asymptotic viewpoint, in Proceedings of the Thirty-Second Conference on Learning Theory, PMLR 99, 2019, pp. 3036\u20133083."},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1137\/0306011"},{"key":"ref42","doi-asserted-by":"crossref","unstructured":"L. Xin, L. Ye, G. Chiu, and S. Sundaram, Identifying the dynamics of a system by leveraging data from similar systems, in 2022 American Control Conference (ACC), IEEE, 2022, pp. 818\u2013824, https:\/\/doi.org\/10.23919\/ACC53348.2022.9867413.","DOI":"10.23919\/ACC53348.2022.9867413"},{"key":"ref43","unstructured":"L. Ye, M. Chi, R. Liao, and V. Gupta, Learning Decentralized Linear Quadratic Regulator with \\(\\sqrt{T}\\) Regret, preprint, arXiv:2210.08886v4, 2022."},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2022.3215940"},{"key":"ref45","doi-asserted-by":"crossref","unstructured":"J. Yu, D. Ho, and A. Wierman, Online adversarial stabilization of unknown networked systems, in Proceedings of the ACM on Measurement and Analysis of Computing Systems, 7 (2023), 26, https:\/\/doi.org\/10.1145\/3579452.","DOI":"10.1145\/3579452"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.1981.1102603"},{"key":"ref47","unstructured":"K. Zhang, B. Hu, and T. Basar, Policy optimization for \\(\\mathcal{H}_2\\) linear control with \\(\\mathcal{H}_{\\infty }\\) robustness guarantee: Implicit regularization and global convergence, in Proceedings of the 2nd Conference on Learning for Dynamics and Control, PMLR 120, 2020, pp. 179\u2013190, https:\/\/proceedings.mlr.press\/v120\/zhang20a.html."},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1109\/TAC.2020.2979785"},{"key":"ref49","unstructured":"I. Ziemann and H. Sandberg, Regret Lower Bounds for Learning Linear Quadratic Gaussian Systems, preprint, arXiv:2201.01680, 2022."}],"container-title":["SIAM Journal on Control and Optimization"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/epubs.siam.org\/doi\/pdf\/10.1137\/23M161714X","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T13:54:09Z","timestamp":1787320449000},"score":1,"resource":{"primary":{"URL":"https:\/\/epubs.siam.org\/doi\/10.1137\/23M161714X"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,12,16]]},"references-count":49,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2024,12,31]]}},"alternative-id":["10.1137\/23M161714X"],"URL":"https:\/\/doi.org\/10.1137\/23m161714x","relation":{},"ISSN":["0363-0129","1095-7138"],"issn-type":[{"value":"0363-0129","type":"print"},{"value":"1095-7138","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,12,16]]}}}