Journal of Shandong University(Engineering Science) ›› 2026, Vol. 56 ›› Issue (1): 14-25.doi: 10.6040/j.issn.1672-3961.0.2024.231

• Machine Learning & Data Mining • Previous Articles    

Distributed online optimization algorithm based on differential privacy mechanism and one-point feedback

ZHANG Bo1, XU Yue1, KANG Le1, ZHANG Guijun2   

  1. ZHANG Bo1, XU Yue1, KANG Le1, ZHANG Guijun2( 1. State Grid Ningxia Electric Power Co., Ltd., Information Communication Company, Yinchuan 750002, Ningxia, China;
    2. College of Finance and Economics, Taiyuan University of Technology, Taiyuan 030024, Shanxi, China
  • Published:2026-02-03

PDF (PC)

2

Abstract: For distributed online optimization problem with privacy protection on directed networks, this research proposed a distributed online optimization algorithm based on differential privacy mechanism. The state of the node was disturbed by random noise which conformed to the Laplacian distribution, and the privacy information of the node was effectively protected. To solve the problem of unknown gradient information explicitly, this research introduced an one-point feedback to estimate the real gradient, and used the estimated gradient information to guide the update of decision variables, so that the algorithm could adapt to the scenario where the gradient information was unavailable. The theoretical results showed that the proposed algorithm could not only protect the privacy information of nodes but also realize the sublinear regret, and the distributed online optimization problem could be effectively solved. The simulation results verified the effectiveness of the algorithm.

Key words: distributed optimization, online optimization, multi-agent system, differential privacy mechanism, one-point feedback

CLC Number: 

  • TP181
[1] ZHAO Z Y, XIA L C, JIANG L Y, et al. Zeroth-order gradient tracking for decentralized learning with privacy guarantees[J]. ISA Transactions, 2024, 152: 1-14.
[2] 王程, 刘念. 基于交替方向乘子法的互联微电网系统分布式优化调度[J]. 电网技术, 2016, 40(9): 2675-2681. WANG Cheng, LIU Nian. Distributed optimal dispatching of interconnected microgrid system based on alternating direction method of multipliers[J]. Power System Technology, 2016, 40(9): 2675-2681.
[3] WANG S N, LI C G. Distributed robust optimization in networked system[J]. IEEE Transactions on Cybernetics, 2016, 47(8): 2321-2333.
[4] ZHAO Z Y. Sample-based dynamic event triggered algorithm for optimization problem of multi-agent systems[J]. International Journal of Control, Automation and Systems, 2022, 20(8): 2492-2502.
[5] 杨涛, 徐磊, 易新蕾, 等. 基于事件触发的分布式优化算法[J]. 自动化学报, 2022, 48(1): 133-143. YANG Tao, XU Lei, YI Xinlei, et al. Event-triggered distributed optimization algorithms[J]. Acta Automatica Sinica, 2022, 48(1): 133-143.
[6] XI C G, KHAN U A. Distributed subgradient projection algorithm over directed graphs[J]. IEEE Transactions on Automatic Control, 2017, 62(8): 3986-3992.
[7] LI H Q, WANG Z, CHEN G, et al. Distributed robust algorithm for economic dispatch in smart grids over general unbalanced directed networks[J]. IEEE Transactions on Industrial Informatics, 2020, 16(7): 4322-4332.
[8] 陈刚, 李志勇. 集合约束下多智能体系统分布式固定时间优化控制[J]. 自动化学报, 2022, 48(9): 2254-2264. CHEN Gang, LI Zhiyong. Distributed fixed-time optimization control for multi-agent systems with set constraints[J]. Acta Automatica Sinica, 2022, 48(9): 2254-2264.
[9] YAN F, SUNDARAM S, VISHWANATHAN S V N, et al. Distributed autonomous online learning: regrets and intrinsic privacy-preserving properties[J]. IEEE Transactions on Knowledge and Data Engineering, 2012, 25(11): 2483-2493.
[10] 刘洋. 时变非平衡图下多智能体分布式优化算法设计[D]. 大连: 大连理工大学, 2021: 35-43. LIU Yang. Design of distributed optimization algorithm for multi-agent systems over time-varying unbalanced graphs[D]. Dalian:Dalian University of Technology, 2021: 35-43.
[11] MATEOSNU(~overN)EZ D, CORTÉS J. Distributed online convex optimization over jointly connected digraphs[J]. IEEE Transactions on Network Science and Engineering, 2014, 1(1): 23-37.
[12] 吴庆涛, 朱军龙, 葛泉波, 等. 一种基于条件梯度的加速分布式在线学习算法[J]. 自动化学报, 2024, 50(2): 386-402. WU Qingtao, ZHU Junlong, GE Quanbo, et al. An accelerated distributed online learning algorithm based on conditional gradient[J]. Acta Automatica Sinica, 2024, 50(2): 386-402.
[13] PANG Y P, HU G Q. Randomized gradient-free distributed online optimization with time-varying cost functions[C] //2019 IEEE 58th Conference on Decision and Control(CDC). Nice, France:IEEE, 2019: 4910-4915.
[14] PANG Y P, HU G Q. Randomized gradient-free distributed optimization methods for a multiagent system with unknown cost function[J]. IEEE Transactions on Automatic Control, 2020, 65(1): 333-340.
[15] WANG C, XU S Y, YUAN D M, et al. Push-sum distributed online optimization with bandit feedback[J]. IEEE Transactions on Cybernetics, 2022, 52(4): 2263-2273.
[16] DWORK C. Differential privacy[C] // International Colloquium on Automata, Languages, and Programming. Berlin, Germany: Springer, 2006: 1-12.
[17] DING T, ZHU S Y, HE J P, et al. Differentially private distributed optimization via state and direction perturbation in multiagent systems[J]. IEEE Transactions on Automatic Control, 2022, 67(2): 722-737.
[18] XIONG Y Y, XU J M, YOU K Y, et al. Privacy-preserving distributed online optimization over unbalanced digraphs via subgradient rescaling[J]. IEEE Transactions on Control of Network Systems, 2020, 7(3): 1366-1378.
[19] WEI M L, YANG Z Q, JI Q T, et al. Privacy-preserving distributed projected one-point bandit online optimization over directed graphs[J]. Asian Journal of Control, 2023, 25(6): 4705-4720.
[20] ZHAO Z Y, YANG J, GAO W, et al. Differentially private distributed online optimization via push-sum one-point bandit dual averaging[J]. Neurocomputing, 2024, 572: 127184.
[1] CUI Yang, ZHANG Ke, JIANG Bin. Fault estimation design for linear multi-agent systems with switching topologies [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2017, 47(5): 263-270.
[2] WU Yanming, WANG Ruiyun, WANG Zhanshan. Fault detection for multi-agent systems based on intermediate observer [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2017, 47(5): 96-102.
[3] YE Dan, ZHANG Tianyu, LI Kui. Adaptive fault-tolerant containment control for multi-agent systems with unknown global information [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2017, 47(5): 1-6.
[4] LI Ming, LIU Wei, ZHANG Yanduo. Mulit-Agent dynamic task allocation based on improved contract net protocol [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2016, 46(2): 51-56.
[5] YU Zhen-mei1, YANG Rong-rong2, QI Feng-mei3, ZHANG Hui3, LIU Feng-ming4. A load balance model for Web services based on  immune stability maintenance [J]. JOURNAL OF SHANDONG UNIVERSITY (ENGINEERING SCIENCE), 2012, 42(3): 39-44.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
Copyright 2018 © Editorial office of Journal of Shandong University (Engineering Science)
Supported by Beijing Magtech Co.Ltd