This paper examines the security aspects of different pervasive scenarios involving agent communities evolved using multi agent systems (MAS). It describes the motivation and the objectives behind the development of these agent communities and analyses the security vulnerabilities, which arise within them. To counter these vulnerabilities, the paper proposes a Buddy model of security for the agent community. In this model, every agent protects its neighbour within the community, thereby sharing the responsibilities of the security function. This feature makes it a better option as compared to other hierarchical models of security, which can be brought down by a concerted attack at the controller agent. This paper also demonstrates the applicability and the effectiveness of the Buddy model in different pervasive scenarios and makes a strong case for its adoption.
|Cite as: Page, J., Zaslavsky, A. and Indrawan, M. (2004). A Buddy Model of Security for Mobile Agent Communities Operating in Pervasive Scenarios. In Proc. Second Australasian Information Security Workshop (AISW2004), Dunedin, New Zealand. CRPIT, 32. Montague, P. and Steketee, C., Eds. ACS. 17-25. |
(local if available)