Protection of multiagent systems

Abstract

The use of intelligent agents to improve business to business (B2B) communications is starting to be adopted by more and more companies. The main reason is that intelligent agents are able to multiply the number of agreements between suppliers and customers. However, due to the type of information managed by agents those negotiations must be protected in order to avoid fraud and misrepresentation. Communications between agents must be authenticated and, once an agreement is signed, none of the parties should be able to repudiate its signature, and hence their agreement. Furthermore, some confidentiality services must be supplied, allowing to hide some message contents or identities.Security has important implications to the participants of the market, because they will trust the system according to the level of security supplied. This paper shows how RETSINA agents may obtain those security services using trust management and security mechanisms. Our approach uses the trust infrastructure defined by [3] for RETSINA and the security services supplied by the new IPSEC protocols [1]. © Springer-Verlag 2003.