Towards a Framework for Autonomic Security Protocols.

We've been looking at what we call the security protocol negotiation problem. People who use security protocols to authenticate, do key exchange, or whatever, are typically stuck with using whatever protocol is available. If the participants can't agree on a suitable protocol, then they don't get to use each other's services. We are interested in self-configuring security protocols where the participants have to say what their requirements are and synthesise or con.gure their own suitable protocol that would meet their respective goals. If we have this idea of self-configuring security protocols then, as a consequence, we also have selfhealing security protocols. If the requirements change, or if the assumptions about our principals and keys change, then we have this ability to recon.gure and produce a security protocol that will meet those new goals, based on those assumptions. [ABSTRACT FROM AUTHOR]