Abstract

An approach to solve the security aspects raised by agent systems, especially with mobile code and in an open environment, in the context of telecommunication applications is presented. A scenario is given that represents a typical user approach to tomorrow’s integrated services. Typical applications and services are thoroughly characterized, electronic markets are introduced, the players in the new market are shown, and the new market models and related security questions are put into relation to applicable law. The technology involved in the framework is presented. Various security related aspects are shown, Java is discussed as a programming language facilitating security and providing a secure runtime environment. Software agents and their communities are given. An advanced security infrastructure is then devised. It comprises of mechanisms on several dierent abstraction and realization layers. On a low level basic mechanisms are given irrespective of the "agentness" of the system. Security functionality is added to agents and they are enabled to reason about security, service provisioning is secured. Communities of agents are enriched with security functionality to prevent malicious intruders to enter the protected domain, “own” agents are prevented from migrating to mischieveous other agent places. Crossing borders of domains, means for secured e-commerce even between unknown agents are provided.