• DocumentCode
    3139642
  • Title

    Multi-agent Coordination Based on Contract Net Protocol

  • Author

    Sun, Defu ; Wu, Juhua

  • Author_Institution
    Sch. of Inf. Sci. & Technol., Beijing Univ., Beijing, China
  • fYear
    2009
  • fDate
    15-16 May 2009
  • Firstpage
    353
  • Lastpage
    357
  • Abstract
    In multi-agent system, agents need autonomy and collaboration to accomplish a complex task. Contract net protocol (CNP) is coordination mechanism often used in a multi-agent system. Due to the limitations inherent in the conventional CNP, this paper proposes a Norm-based CNP to improve the efficiency and effectiveness of the coordination processes in a multi-agent system. First, a three-dimensional taxonomy of norm is put forward in terms of the hierarchy, type and flexibility of norms. Then a coordination process guided by Norm-based CNP is developed under the taxonomy framework. The new coordination process consists of two principal stages and five sub-divided phases, providing a feasible solution for the optimization of the candidate selection. The merits of dividing the process into two stages are converging coordination process as soon as possible and controlling whole system effectively. In collaborative process, norms ensure to achieve global performance and agents pursuing maximal local performance to achieve satisfied result. Finally a case study is performed to illustrate the real application of the proposed approach.
  • Keywords
    groupware; multi-agent systems; collaborative process; contract net protocol; coordination mechanism; multiagent coordination; multiagent system; norm three-dimensional taxonomy; Contracts; Control systems; Intelligent agent; International collaboration; Law; Multiagent systems; Protocols; Sun; Taxonomy; Ubiquitous computing; contract net protocol; coordination; multi-agent; norm;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Intelligent Ubiquitous Computing and Education, 2009 International Symposium on
  • Conference_Location
    Chengdu
  • Print_ISBN
    978-0-7695-3619-4
  • Type

    conf

  • DOI
    10.1109/IUCE.2009.111
  • Filename
    5222886