Abstract

This paper investigates the role of trust metrics in attack-resistant public key certification. We present an analytical framework for understanding the effectiveness of trust metrics in resisting attacks, including a characterization of the space of possible attacks. Within this framework, we establish the theoretical best case for a trust metric. Finally, we present a practical trust metric based on network flow that meets this theoretical bound. independent sources of certification, and rejects (by assigning low trust values) assertions with insufficient certification. The previous work raises many questions, including: To which kinds of attack is a trust metric resistant? Which trust metric is best? How well do these trust metrics work? 1

Citations