An Untraceability Protocol for Mobile Agents and Its Enhanced Security Study

Abstract

Among various computational models to accommodate the mobile wireless computing paradigm, the mobile agent model has major technological advantages. Moreover, our research shows that the model has some unique features which, if used appropriately, may support user privacy. We therefore propose a security protocol for mobile agents aiming at user untraceability. We have performed a basic security analysis of the protocol and evaluated its performance. We have also implemented it, made it available to users, and applied it to an e-health service. In this paper, we present an enhanced security study of the protocol. As far as we know, such systematised study has not yet been performed. Additionally, known analyses limit their interest to particular attackers (internal attackers). We believe that the attack checklist which we developed to evaluate our protocol, may serve as a helpful reference for authors of other protocols. The paper is also addressed to designers and administrators of mobile agent systems who consider provision of user untraceability within their systems.