Principles: Threat Models, Trusted Computing Bases, and the “Gold” Standard
Topics and Goals
- Threat modeling and case study
- Goal: Develop (formal and informal) threat models for modern systems.
- Trusted Computing Bases (TCBs)
- Goal: Understand the definition of a TCB
- Goal: Identify the TCB for a novel system
- Lampson’s “gold” standard techniques
- Authentication, Authorization, Audit
- Goal: Identify mechanisms for achieving each
- Goal: Distinguish between authentication & authorization
Threat Models
- What is the defining characteristic of security?
- What is a security mindset?
- What does a threat model include?
- Assets
- System’s goals
- What are typical system goals?
- Adversary definition
- How do we define an adversary model?
- Why don’t we include the adversary’s strategy?
- How can we compare adversaries?
- How do we define an adversary model?
-
How can we categorize defenses?
- E-Voting case study
Trusted Computing Bases (TCB)
- What is the TCB?
- What’s the difference between something that is Trusted vs Trustworthy?
- Why do we need a TCB?
- What are the qualities of an ideal TCB?
“Gold” Standard
- Three core princples for reasoning about secure systems:
- Authentication: Who is it?
- Authorization: What can they do?
- Audit: What happened?