Constrained consensus and optimization in multi-agent networks

Angelia Nedich, Asuman Ozdaglar, Pablo A. Parrilo

Research output: Contribution to journalArticle

793 Citations (Scopus)

Abstract

We present distributed algorithms that can be used by multiple agents to align their estimates with a particular value over a network with time-varying connectivity. Our framework is general in that this value can represent a consensus value among multiple agents or an optimal solution of an optimization problem, where the global objective function is a combination of local agent objective functions. Our main focus is on constrained problems where the estimates of each agent are restricted to lie in different convex sets. To highlight the effects of constraints, we first consider a constrained consensus problem and present a distributed "projected consensus algorithm" in which agents combine their local averaging operation with projection on their individual constraint sets. This algorithm can be viewed as a version of an alternating projection method with weights that are varying over time and across agents. We establish convergence and convergence rate results for the projected consensus algorithm.We next study a constrained optimization problem for optimizing the sum of local objective functions of the agents subject to the intersection of their local constraint sets. We present a distributed "projected subgradient algorithm" which involves each agent performing a local averaging operation, taking a subgradient step to minimize its own objective function, and projecting on its constraint set. We show that, with an appropriately selected stepsize rule, the agent estimates generated by this algorithm converge to the same optimal solution for the cases when the weights are constant and equal, and when the weights are time-varying but all agents have the same constraint set.

Original languageEnglish (US)
Article number5404774
Pages (from-to)922-938
Number of pages17
JournalIEEE Transactions on Automatic Control
Volume55
Issue number4
DOIs
StatePublished - Apr 2010
Externally publishedYes

Fingerprint

Constrained optimization
Parallel algorithms

Keywords

  • Consensus
  • Constraints
  • Distributed optimization
  • Subgradient algorithms

ASJC Scopus subject areas

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

Cite this

Constrained consensus and optimization in multi-agent networks. / Nedich, Angelia; Ozdaglar, Asuman; Parrilo, Pablo A.

In: IEEE Transactions on Automatic Control, Vol. 55, No. 4, 5404774, 04.2010, p. 922-938.

Research output: Contribution to journalArticle

Nedich, Angelia ; Ozdaglar, Asuman ; Parrilo, Pablo A. / Constrained consensus and optimization in multi-agent networks. In: IEEE Transactions on Automatic Control. 2010 ; Vol. 55, No. 4. pp. 922-938.
@article{7f1e9e01bb8f4b67902382462b00afc2,
title = "Constrained consensus and optimization in multi-agent networks",
abstract = "We present distributed algorithms that can be used by multiple agents to align their estimates with a particular value over a network with time-varying connectivity. Our framework is general in that this value can represent a consensus value among multiple agents or an optimal solution of an optimization problem, where the global objective function is a combination of local agent objective functions. Our main focus is on constrained problems where the estimates of each agent are restricted to lie in different convex sets. To highlight the effects of constraints, we first consider a constrained consensus problem and present a distributed {"}projected consensus algorithm{"} in which agents combine their local averaging operation with projection on their individual constraint sets. This algorithm can be viewed as a version of an alternating projection method with weights that are varying over time and across agents. We establish convergence and convergence rate results for the projected consensus algorithm.We next study a constrained optimization problem for optimizing the sum of local objective functions of the agents subject to the intersection of their local constraint sets. We present a distributed {"}projected subgradient algorithm{"} which involves each agent performing a local averaging operation, taking a subgradient step to minimize its own objective function, and projecting on its constraint set. We show that, with an appropriately selected stepsize rule, the agent estimates generated by this algorithm converge to the same optimal solution for the cases when the weights are constant and equal, and when the weights are time-varying but all agents have the same constraint set.",
keywords = "Consensus, Constraints, Distributed optimization, Subgradient algorithms",
author = "Angelia Nedich and Asuman Ozdaglar and Parrilo, {Pablo A.}",
year = "2010",
month = "4",
doi = "10.1109/TAC.2010.2041686",
language = "English (US)",
volume = "55",
pages = "922--938",
journal = "IEEE Transactions on Automatic Control",
issn = "0018-9286",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Constrained consensus and optimization in multi-agent networks

AU - Nedich, Angelia

AU - Ozdaglar, Asuman

AU - Parrilo, Pablo A.

PY - 2010/4

Y1 - 2010/4

N2 - We present distributed algorithms that can be used by multiple agents to align their estimates with a particular value over a network with time-varying connectivity. Our framework is general in that this value can represent a consensus value among multiple agents or an optimal solution of an optimization problem, where the global objective function is a combination of local agent objective functions. Our main focus is on constrained problems where the estimates of each agent are restricted to lie in different convex sets. To highlight the effects of constraints, we first consider a constrained consensus problem and present a distributed "projected consensus algorithm" in which agents combine their local averaging operation with projection on their individual constraint sets. This algorithm can be viewed as a version of an alternating projection method with weights that are varying over time and across agents. We establish convergence and convergence rate results for the projected consensus algorithm.We next study a constrained optimization problem for optimizing the sum of local objective functions of the agents subject to the intersection of their local constraint sets. We present a distributed "projected subgradient algorithm" which involves each agent performing a local averaging operation, taking a subgradient step to minimize its own objective function, and projecting on its constraint set. We show that, with an appropriately selected stepsize rule, the agent estimates generated by this algorithm converge to the same optimal solution for the cases when the weights are constant and equal, and when the weights are time-varying but all agents have the same constraint set.

AB - We present distributed algorithms that can be used by multiple agents to align their estimates with a particular value over a network with time-varying connectivity. Our framework is general in that this value can represent a consensus value among multiple agents or an optimal solution of an optimization problem, where the global objective function is a combination of local agent objective functions. Our main focus is on constrained problems where the estimates of each agent are restricted to lie in different convex sets. To highlight the effects of constraints, we first consider a constrained consensus problem and present a distributed "projected consensus algorithm" in which agents combine their local averaging operation with projection on their individual constraint sets. This algorithm can be viewed as a version of an alternating projection method with weights that are varying over time and across agents. We establish convergence and convergence rate results for the projected consensus algorithm.We next study a constrained optimization problem for optimizing the sum of local objective functions of the agents subject to the intersection of their local constraint sets. We present a distributed "projected subgradient algorithm" which involves each agent performing a local averaging operation, taking a subgradient step to minimize its own objective function, and projecting on its constraint set. We show that, with an appropriately selected stepsize rule, the agent estimates generated by this algorithm converge to the same optimal solution for the cases when the weights are constant and equal, and when the weights are time-varying but all agents have the same constraint set.

KW - Consensus

KW - Constraints

KW - Distributed optimization

KW - Subgradient algorithms

UR - http://www.scopus.com/inward/record.url?scp=77950834661&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77950834661&partnerID=8YFLogxK

U2 - 10.1109/TAC.2010.2041686

DO - 10.1109/TAC.2010.2041686

M3 - Article

VL - 55

SP - 922

EP - 938

JO - IEEE Transactions on Automatic Control

JF - IEEE Transactions on Automatic Control

SN - 0018-9286

IS - 4

M1 - 5404774

ER -