In this paper, we present a new solution for the implementation of flexible protection mechanisms in the context of mobile agent systems, where security problems are currently a major issue. In our scheme, agents protect their code and data by carrying their own protection mechanisms. This approach improves traditional solutions, where protection was managed by the platform. The implementation is far from trivial. We have implemented this scheme in the JADE framework, using Java. Any application using mobile agents can incorporate these mechanisms to implement agent protection with a minimum impact in the its existing code base.
Joan Ametller, Sergi Robles, Jose A. Ortega-Ruiz