Title: An adaptive casteship mechanism for developing multi-agent systems

Authors: Xinjun Mao, Lijun Shan, Hong Zhu, Ji Wang

Addresses: Department of Computer Science and Technology, National University of Defense Technology, Changsha, Hunan Province, 410073, PR China. ' Department of Computer Science and Technology, National University of Defense Technology, Changsha, Hunan Province, 410073, PR China. ' Department of Computing, Oxford Brookes University, Oxford OX33 1HX, UK. ' National Laboratory for Parallel and Distributed Processing, National University of Defense Technology, Changsha, Hunan Province, PR China.

Abstract: In this paper, we propose an adaptive casteship mechanism for modelling and designing adaptive Multi Agent Systems (MAS). In our approach, caste is the modular unit and abstraction that specify agents| behaviour. Adaptive behaviours of agents are captured as the change of castes during their lifecycles by executing |join|, |quit|, |activate| and |deactivate| operations on castes. The formal semantics of caste operations are rigorously defined. The properties of agent|s adaptive behaviours are formally specified and proved. A graphical notation of caste transition diagrams and a number of rules for check consistency are designed. An example is also presented throughout the paper.

Keywords: multi-agent systems; MAS design; MAS modelling; adaptive agents; caste; adaptive casteship; agent-based systems.

DOI: 10.1504/IJCAT.2008.017716

International Journal of Computer Applications in Technology, 2008 Vol.31 No.1/2, pp.17 - 34

Published online: 27 Mar 2008 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article