Trust Enhanced Security - A New Philosophy for Secure Collaboration of Mobile Agents

Abstract

The mobile agent computing model violates some of the fundamental assumptions of conventional security techniques. Consequently, this has rendered many of the existing conventional security countermeasures less effective for mobile agents. In this paper, we propose a new philosophy of trust enhanced security, which advocates a paradigm shift for mobile agent security solutions: from security-centric to trust-centric with the aim of providing improved security and performance of mobile agents. We first examine the problem of uncertainty in behavior induced by the security assumption violations by mobile agents; we then propose a trust enhanced security approach and argue for the need for a paradigm shift to trust-centric solutions. Next we identify a list of general design requirements for the trust-centric solutions and outline the new architectural design which supports the new trust enhanced security philosophy in practice. Finally we discuss the emergent properties of the new architecture and introduce the experimental results for validating the properties