Application Security Models for Mobile Agent Systems

Mobile agents are a distributed computing paradigm based on mobile autonomous programs. Mobile applications must balance security requirements with available security mechanisms in order to meet application level security goals. We introduce a trust framework to reason about application security requirements, trust expression, and agent protection mechanisms. We develop application security models that capture initial trust relationships and consider their use for mobile agent security.