MagicNET: Security architecture for discovery and adoption of mobile agents

Current research in the area of security for mobile agents deals mainly with the runtime issues of agents' protection. Mobile agent systems do not address precisely the process of acquiring mobile agents by their owners, but they assume that the agents are somehow already available for use. This assumption is acceptable for experimental or prototyping environments, but it is inadequate for the real world scenarios where agents should be trusted and reliable, but agent creators and agent owners are separated and manage agents from mutually remote locations. Thus, the issue of agents' adoption for use in serious, sensitive and business networks is very important if agents are used in real-life applications. In this paper we describe the architecture and procedures for secure, verifiable and authenticated discovery and adoption of mobile agents. The main contribution is that in this process agent's code and its functionality can be verified, so that such agents can be deployed in serious applications and scenarios.