Accounting for Uncertainty in Deceptive Signaling for Cybersecurity

Abstract

Deceptive signaling has proven an effective method that can aid security analysists and deter attacks on unprotected targets by strategically revealing information to an attacker. However, recent research has shown that uncertainty in real-time information processing can have a negative impact on the effectiveness of the defense algorithm. The current research developed a new algorithm, dubbed Confusion Signaling, that aims to account for uncertainty in an abstracted insider attack scenario. The results of cognitive model simulations and a human behavioral experiment reveal interesting and unexpected reactions under uncertainty. We discuss the implications of these findings for signaling algorithms that aim to account for uncertainty using deceptive signaling for cybersecurity.