Exploiting Locality and Structure for Distributed Optimization in Multi-Agent Systems

被引:0
作者
Brown, Robin [1 ]
Rossi, Federico [3 ]
Solovey, Kiril [2 ]
Wolf, Michael T. [3 ]
Pavone, Marco [2 ]
机构
[1] Stanford Univ, Inst Computat & Math Engn, Stanford, CA 94305 USA
[2] Stanford Univ, Dept Aeronaut & Astronaut, Stanford, CA 94305 USA
[3] CALTECH, Jet Prop Lab, 4800 Oak Grove Dr, Pasadena, CA 91109 USA
来源
2020 EUROPEAN CONTROL CONFERENCE (ECC 2020) | 2020年
关键词
CONSENSUS; ALGORITHM;
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
A number of prototypical optimization problems in multi-agent systems (e.g. task allocation and network load-sharing) exhibit a highly local structure: that is, each agent's decision variables are only directly coupled to few other agent's variables through the objective function or the constraints. Nevertheless, existing algorithms for distributed optimization generally do not exploit the locality structure of the problem, requiring all agents to compute or exchange the full set of decision variables. In this paper, we develop a rigorous notion of "locality" that relates the structural properties of a linearly-constrained convex optimization problem (in particular, the sparsity structure of the constraint matrix and the objective function) to the amount of information that agents should exchange to compute an arbitrarily high-quality approximation to the problem from a cold-start. We leverage the notion of locality to develop a locality-aware distributed optimization algorithm, and we show that, for problems where individual agents only require to know a small portion of the optimal solution, the algorithm requires very limited inter-agent communication. Numerical results show that the convergence rate of our algorithm is directly explained by the locality metric proposed, and that the proposed theoretical bounds are remarkably tight; comparison to the projected sub-gradient algorithm shows that our locality-aware algorithm requires orders of magnitude fewer communication rounds to achieve similar solution quality.
引用
收藏
页码:440 / 447
页数:8
相关论文
共 20 条
[1]  
Bererton C. A., 2004, THESIS
[2]  
Brown R. A., 2020, EXPLOITING LOCALITY
[3]  
Chen A., 2012, CORR
[4]   Dual Averaging for Distributed Optimization: Convergence Analysis and Network Scaling [J].
Duchi, John C. ;
Agarwal, Alekh ;
Wainwright, Martin J. .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2012, 57 (03) :592-606
[5]   Fast Distributed Gradient Methods [J].
Jakovetic, Dusan ;
Xavier, Joao ;
Moura, Jose M. F. .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2014, 59 (05) :1131-1146
[6]   LOW DIAMETER GRAPH DECOMPOSITIONS [J].
LINIAL, N ;
SAKS, M .
COMBINATORICA, 1993, 13 (04) :441-454
[7]   Optimization flow control - I: Basic algorithm and convergence [J].
Low, SH ;
Lapsley, DE .
IEEE-ACM TRANSACTIONS ON NETWORKING, 1999, 7 (06) :861-874
[8]   Convergence of Min-Sum Message-Passing for Convex Optimization [J].
Moallemi, Ciamac C. ;
Van Roy, Benjamin .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2010, 56 (04) :2041-2050
[9]   Network Newton Distributed Optimization Methods [J].
Mokhtari, Aryan ;
Ling, Qing ;
Ribeiro, Alejandro .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2017, 65 (01) :146-161
[10]  
Nedic A., 2017, NETWORK TOPOLOGY COM