Distributed subgradient methods for multi-agent optimization

Angelia Nedich, Asuman Ozdaglar

Research output: Contribution to journalArticle

1330 Citations (Scopus)

Abstract

We study a distributed computation model for optimizing a sum of convex objective functions corresponding to multiple agents. For solving this (not necessarily smooth) optimization problem, we consider a subgradient method that is distributed among the agents. The method involves every agent minimizing his/her own objective function while exchanging information locally with other agents in the network over a time-varying topology. We provide convergence results and convergence rate estimates for the subgradient method. Our convergence rate results explicitly characterize the tradeoff between a desired accuracy of the generated approximate optimal solutions and the number of iterations needed to achieve the accuracy.

Original languageEnglish (US)
Pages (from-to)48-61
Number of pages14
JournalIEEE Transactions on Automatic Control
Volume54
Issue number1
DOIs
StatePublished - 2009
Externally publishedYes

Fingerprint

Topology

Keywords

  • Convex optimization
  • Cooperative control
  • Distributed optimization
  • Multi-agent network
  • Subgradient method

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Control and Systems Engineering
  • Computer Science Applications

Cite this

Distributed subgradient methods for multi-agent optimization. / Nedich, Angelia; Ozdaglar, Asuman.

In: IEEE Transactions on Automatic Control, Vol. 54, No. 1, 2009, p. 48-61.

Research output: Contribution to journalArticle

@article{13bda96971ad45f4970739942fd0b86c,
title = "Distributed subgradient methods for multi-agent optimization",
abstract = "We study a distributed computation model for optimizing a sum of convex objective functions corresponding to multiple agents. For solving this (not necessarily smooth) optimization problem, we consider a subgradient method that is distributed among the agents. The method involves every agent minimizing his/her own objective function while exchanging information locally with other agents in the network over a time-varying topology. We provide convergence results and convergence rate estimates for the subgradient method. Our convergence rate results explicitly characterize the tradeoff between a desired accuracy of the generated approximate optimal solutions and the number of iterations needed to achieve the accuracy.",
keywords = "Convex optimization, Cooperative control, Distributed optimization, Multi-agent network, Subgradient method",
author = "Angelia Nedich and Asuman Ozdaglar",
year = "2009",
doi = "10.1109/TAC.2008.2009515",
language = "English (US)",
volume = "54",
pages = "48--61",
journal = "IEEE Transactions on Automatic Control",
issn = "0018-9286",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "1",

}

TY - JOUR

T1 - Distributed subgradient methods for multi-agent optimization

AU - Nedich, Angelia

AU - Ozdaglar, Asuman

PY - 2009

Y1 - 2009

N2 - We study a distributed computation model for optimizing a sum of convex objective functions corresponding to multiple agents. For solving this (not necessarily smooth) optimization problem, we consider a subgradient method that is distributed among the agents. The method involves every agent minimizing his/her own objective function while exchanging information locally with other agents in the network over a time-varying topology. We provide convergence results and convergence rate estimates for the subgradient method. Our convergence rate results explicitly characterize the tradeoff between a desired accuracy of the generated approximate optimal solutions and the number of iterations needed to achieve the accuracy.

AB - We study a distributed computation model for optimizing a sum of convex objective functions corresponding to multiple agents. For solving this (not necessarily smooth) optimization problem, we consider a subgradient method that is distributed among the agents. The method involves every agent minimizing his/her own objective function while exchanging information locally with other agents in the network over a time-varying topology. We provide convergence results and convergence rate estimates for the subgradient method. Our convergence rate results explicitly characterize the tradeoff between a desired accuracy of the generated approximate optimal solutions and the number of iterations needed to achieve the accuracy.

KW - Convex optimization

KW - Cooperative control

KW - Distributed optimization

KW - Multi-agent network

KW - Subgradient method

UR - http://www.scopus.com/inward/record.url?scp=59649103668&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=59649103668&partnerID=8YFLogxK

U2 - 10.1109/TAC.2008.2009515

DO - 10.1109/TAC.2008.2009515

M3 - Article

AN - SCOPUS:59649103668

VL - 54

SP - 48

EP - 61

JO - IEEE Transactions on Automatic Control

JF - IEEE Transactions on Automatic Control

SN - 0018-9286

IS - 1

ER -