The increased connectivity and potential insider threats make traditional network defense vulnerable. Instead of assuming that everything behind the security perimeter is safe, the zero-trust security model verifies every incoming request before granting access. This chapter draws attention to the cyber resilience within the zero-trust model. We introduce the evolution from traditional perimeter-based security to zero trust and discuss their difference. Two key elements of the zero-trust engine are trust evaluation (TE) and policy engine (PE). We introduce the design of the two components and discuss how their interplay would contribute to cyber resilience. Dynamic game theory and learning are applied as quantitative approaches to achieve automated zero-trust cyber resilience. Several case studies and implementations are introduced to illustrate the benefits of such a security model.
View on arXiv