Results 1  10
of
12
Quality Guarantees on kOptimal Solutions for Distributed Constraint Optimization Problems
, 2007
"... A distributed constraint optimization problem (DCOP) is a formalism that captures the rewards and costs of local interactions within a team of agents. Because complete algorithms to solve DCOPs are unsuitable for some dynamic or anytime domains, researchers have explored incomplete DCOP algorithms t ..."
Abstract

Cited by 44 (10 self)
 Add to MetaCart
A distributed constraint optimization problem (DCOP) is a formalism that captures the rewards and costs of local interactions within a team of agents. Because complete algorithms to solve DCOPs are unsuitable for some dynamic or anytime domains, researchers have explored incomplete DCOP algorithms that result in locally optimal solutions. One type of categorization of such algorithms, and the solutions they produce, is koptimality; a koptimal solution is one that cannot be improved by any deviation by k or fewer agents. This paper presents the first known guarantees on solution quality for koptimal solutions. The guarantees are independent of the costs and rewards in the DCOP, and once computed can be used for any DCOP of a given constraint graph structure.
On Koptimal distributed constraint optimization algorithms: new bounds and algorithms
 In AAMAS ’08
, 2008
"... Distributed constraint optimization (DCOP) is a promising approach to coordination, scheduling and task allocation in multi agent networks. In largescale or lowbandwidth networks, finding the global optimum is often impractical. Koptimality is a promising new approach: for the first time it provi ..."
Abstract

Cited by 16 (7 self)
 Add to MetaCart
Distributed constraint optimization (DCOP) is a promising approach to coordination, scheduling and task allocation in multi agent networks. In largescale or lowbandwidth networks, finding the global optimum is often impractical. Koptimality is a promising new approach: for the first time it provides us a set of locally optimal algorithms with quality guarantees as a fraction of global optimum. Unfortunately, previous work in koptimality did not address domains where we may have prior knowledge of reward structure; and it failed to provide quality guarantees or algorithms for domains with hard constraints (such as agents ’ local resource constraints). This paper addresses these shortcomings with three key contributions. It provides: (i) improved lowerbounds on koptima quality incorporating available prior knowledge of reward structure; (ii) lower bounds on koptima quality for problems with hard constraints; and (iii) koptimal algorithms for solving DCOPs with hard constraints and detailed experimental results on largescale networks.
Solving multiagent networks using distributed constraint optimization
, 2007
"... In many cooperative multiagent domains, the effect of local interactions between agents can be compactly represented as a network structure. Given that agents are spread across such a network, agents directly interact only with a small group of neighbors. A distributed constraint optimization proble ..."
Abstract

Cited by 7 (5 self)
 Add to MetaCart
In many cooperative multiagent domains, the effect of local interactions between agents can be compactly represented as a network structure. Given that agents are spread across such a network, agents directly interact only with a small group of neighbors. A distributed constraint optimization problem (DCOP) is a useful framework to reason about such networks of agents. Given agents’ inability to communicate and collaborate in large groups in such networks, we focus on an approach called koptimality for solving DCOPs. In this approach, agents form groups of one or more agents until no group of k or fewer agents can possibly improve the DCOP solution; we define this type of local optimum, and any algorithm guaranteed to reach such a local optimum, as koptimal. The article provides an overview of three key results related to koptimality. The first set of results are worstcase guarantees on the solution quality of koptima in a DCOP. These guarantees can help determine an appropriate koptimal algorithm, or possibly an appropriate constraint graph structure, for agents to use in situations where the cost of coordination between agents must be weighed against the quality of the solution reached. The second set of results are upper bounds on the number of koptima that can exist in a DCOP. These results are useful in domains where a DCOP must generate a set of solutions rather than single solution. Finally, we sketch algorithms for koptimality and provide some experimental results for 1, 2 and 3optimal algorithms for several types of DCOPs.
Centralized, Distributed or Something Else? Making Timely Decisions in MultiAgent Systems
"... In multiagent systems, agents need to share information in order to make good decisions. Who does what in order to achieve this matters a lot. The assignment of responsibility influences delay and consequently affects agents ’ abilities to make timely decisions. It is often unclear which approaches ..."
Abstract

Cited by 2 (1 self)
 Add to MetaCart
In multiagent systems, agents need to share information in order to make good decisions. Who does what in order to achieve this matters a lot. The assignment of responsibility influences delay and consequently affects agents ’ abilities to make timely decisions. It is often unclear which approaches are best. We develop a model where one can easily test the impact of different assignments and information sharing protocols by focusing only on the delays caused by computation and communication. Using the model, we obtain interesting results that provide insight about the types of assignments that perform well in various domains and how slight variations in protocols can make great differences in feasibility.
Locally Optimal Algorithms and Solutions for Distributed Constraint Optimization
"... This paper summarizes the author’s recent work in distributed constraint optimization (DCOP). New local algorithms, as well as theoretical results about the types of solutions that these algorithms can reach, are discussed. ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
This paper summarizes the author’s recent work in distributed constraint optimization (DCOP). New local algorithms, as well as theoretical results about the types of solutions that these algorithms can reach, are discussed.
Multiagent Teamwork: Hybrid Approaches
"... Today within the multiagent community, we see at least four competing methods to building multiagent systems: beliefdesireintention (BDI), distributed constraint optimization (DCOP), distributed POMDPs, and auctions or gametheoretic methods. While there is exciting progress within each approach, t ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
Today within the multiagent community, we see at least four competing methods to building multiagent systems: beliefdesireintention (BDI), distributed constraint optimization (DCOP), distributed POMDPs, and auctions or gametheoretic methods. While there is exciting progress within each approach, there is a lack of crosscutting research. This article highlights the various hybrid techniques for multiagent teamwork developed by the teamcore group. In particular, for the past decade, the TEAMCORE research group has focused on building agent teams in complex, dynamic domains. While our early work was inspired by BDI, we will present an overview of recent research that uses DCOPs and distributed POMDPs in building agent teams. While DCOP and distributed POMDP algorithms provide promising results, hybrid approaches allow us to use the complementary strengths of different techniques to create algorithms that perform better than either of their component algorithms alone. For example, in the BDIPOMDP hybrid approach, BDI team plans are exploited to improve POMDP tractability, and POMDPs improve BDI team plan performance. 1.
unknown title
"... In many multiagent domains, including sensor networks, teams ofunmanned air vehicles, or teams of personal assistant agents, a set of agents chooses a joint action as a combination of individual actions. Often, the locality of agents ’ interactions means that the utility generated by each agent’s a ..."
Abstract
 Add to MetaCart
In many multiagent domains, including sensor networks, teams ofunmanned air vehicles, or teams of personal assistant agents, a set of agents chooses a joint action as a combination of individual actions. Often, the locality of agents ’ interactions means that the utility generated by each agent’s action depends only on the actions of a subset of the other agents. In this case, the outcomes of possible joint actions can be compactly represented by graphical models, such as a distributed constraint optimization problem (DCOP) (Modi et al. 2005, Mailler and Lesser 2004) for cooperative domains or by a graphical game (Kearns, Littman, and Singh 2001; Vickrey and Koller 2002) for noncooperative domains. Each of these models can take the form of a graph in which each node is an agent and each (hyper)edge denotes a subset of locally interacting agents. In particular, associated with each such hyperedge is a reward matrix that indicates the costs or rewards incurred due to the joint action of the subset of agents involved, either to the agent team (in DCOPs) or to individual agents (in graphical games). Here, costs refer to negative real numbers and rewards refer to positive real numbers to reflect intuition; we could use either costs or rewards exclusively if we assume they can span all real numbers. Local interaction is a key property captured in such graphs; not all agents interact with all other agents. This article focuses on the team setting, using DCOP, whose applications include multiagent plan coordination (Cox, Durfee, and Bartold 2005), sensor networks (Zhang et al. 2003), meeting scheduling (Petcu and Faltings 2005), and RoboCup soccer (Vlassis, Elhorst, and Kok 2004). Traditionally, researchers have focused on obtaining a single, globally optimal solution to DCOPs, introducing complete algorithms such as Adopt (Modi et al.
Acknowledgements
, 2010
"... First and foremost, I would like to thank my advisor, Sven Koenig, for his guidance and support throughout this journey, as well as the other members of my committee, ..."
Abstract
 Add to MetaCart
First and foremost, I would like to thank my advisor, Sven Koenig, for his guidance and support throughout this journey, as well as the other members of my committee,
Balancing Local Resources and Global Goals in MultiplyConstrained DCOP
, 2010
"... Distributed constraint optimization (DCOP) is a useful framework for cooperative multiagent coordination. DCOP focuses on optimizing a single team objective. However, in many domains, agents must satisfy constraints on resources consumed locally while optimizing the team goal. Yet, these resource co ..."
Abstract
 Add to MetaCart
(Show Context)
Distributed constraint optimization (DCOP) is a useful framework for cooperative multiagent coordination. DCOP focuses on optimizing a single team objective. However, in many domains, agents must satisfy constraints on resources consumed locally while optimizing the team goal. Yet, these resource constraints may need to be kept private. Designing DCOP algorithms for these domains requires managing complex tradeoffs in completeness, scalability, privacy and efficiency. This article defines the multiplyconstrained DCOP (MCDCOP) framework and provides complete (globally optimal) and incomplete (locally optimal) algorithms for solving MCDCOP problems. Complete algorithms find the best allocation of scarce resources while optimizing the team objective, while incomplete algorithms are more scalable. The algorithms use four main techniques: (i) transforming constraints to maintain privacy; (ii) dynamically setting upper bounds on resource consumption; (iii) identifying the extent to which the local graph structure allows agents to compute exact bounds; and (iv) using a virtual assignment to flag problems rendered unsatisfiable by resource constraints. Proofs of correctness are presented for all algorithms. Experimental results illustrate the strengths and weaknesses of both the complete and incomplete algorithms.