DeQED: an Efficient DivideandCoordinate Algorithm for DCOP (Extended Abstract)

12th International Conference on Autonomous Agents and Multiagent Systems

Daisuke Hatano ; Katsutoshi Hirayama

PID  :  123936

来源: CEUR

【 摘 要 】

This paper presents a new DCOP algorithm called DeQED (De composition with Quadratic Encoding to Decentralize). DeQED is based on the DivideandCoordinate (DaC) framework, where the agents repeat solving their updated local subproblems (the di vide stage) and exchanging coordination information that causes to update their local subproblems (the coordinate stage). Unlike other DaCbased DCOP algorithms, DeQED does not essentially increase the complexity of local subproblems and allows agents to avoid exchanging variable values in the coordinate stage. Our experimental results show that DeQED significantly outperformed other incomplete DCOP algorithms for both random and structured