DCOP Games for Multi-agent Coordination
نویسندگان
چکیده
Many challenges in multi-agent coordination can be modeled as distributed constraint optimization problems (DCOPs) but complete algorithms do not scale well nor respond effectively to dynamic or anytime environments. We introduce a transformation of DCOPs into graphical games that allows us to devise and analyze algorithms based on local utility and prove the monotonicity property of a class of such algorithms. The game-theoretic framework also enables us to characterize new equilibrium sets corresponding to a given degree of agent coordination. A key result in this paper is the discovery of a novel mapping between finite games and coding theory from which we can determine a priori bounds on the number of equilibria in these sets, which is useful in choosing the appropriate level of coordination given the communication cost of an algorithm.
منابع مشابه
Distributed Multi-Criteria Coordination in Multi-Agent Systems
Distributed constraint optimization (DCOP) has emerged as a key technique for multiagent coordination. Unfortunately, while previous work in DCOP focuses on optimizing a single team objective, domains often require satisfying additional criteria. This paper provides a novel multi-criteria DCOP algorithm, based on two key ideas: (i) transforming multi-criteria problems via virtual variables to h...
متن کاملChallenges for multi-agent coordination theory based on empirical observations
Significant research progress and understanding about the nature of coordination has been made over the years. Development of the DCOP and DEC-MDP frameworks in the past decade has been especially important. Although these advances are very important for multi-agent coordination theory, they overlook a set of coordination behaviors and phenomena that have been observed empirically by many resea...
متن کاملDistributed Multi-Criteria Coordination: Privacy vs. Efficiency
Distributed constraint optimization (DCOP) has emerged as a key technique for multiagent coordination. Unfortunately, while previous work in DCOP focuses on optimizing a single team objective, domains often require satisfying additional criteria. This paper provides a novel multi-criteria DCOP algorithm, based on two key ideas: (i) transforming multi-criteria problems via virtual variables to h...
متن کاملCoordinating multi-agent reinforcement learning with limited communication
Coordinated multi-agent reinforcement learning (MARL) provides a promising approach to scaling learning in large cooperative multiagent systems. Distributed constraint optimization (DCOP) techniques have been used to coordinate action selection among agents during both the learning phase and the policy execution phase (if learning is off-line) to ensure good overall system performance. However,...
متن کاملA distributed constraint optimization approach for coordination under uncertainty
Distributed Constraint Optimization (DCOP) provides a rich framework for modeling multi-agent coordination problems. Existing problem domains for DCOP focus on small (<100 variables), deterministic domains. We present a mapping to DCOP for large-scale team coordination problems that were used in the DARPA Coordinators program. This domain requires distributed, scalable algorithms to meet diffic...
متن کامل