Security

Subjects, Objects, Policies, and Adversaries

Security is the discipline of protecting systems against threats in the presence of adversaries who are assumed to be clever, persistent, and well-resourced.

The fundamental elements are subjects (principals), objects (resources), policies (what is allowed), mechanisms (how policy is enforced), and threats (what the adversary can do).

Cryptographic primitives, access control models, and security protocols are the higher-order structures that make protection practical at scale.

This note connects deeply to operating systems (the reference monitor and TCB), networking (secure channels and protocols), algorithms (cryptographic primitives and protocols), and the general theory of systems (protection as a control problem).

Core Security Principles

The CIA triad, Kerckhoffs’s principle (security through obscurity is insufficient), the computational hardness assumptions underlying cryptography, and the principle of least privilege and complete mediation form the deductive foundation.

From these follow the design of modern cryptographic protocols, access control systems, and secure system architectures.

Measuring Security in Practice

Attack success rates, time-to-compromise, false positive/negative rates for detection systems, and measurable side-channel leakage are the observables. Implementation details, key sizes, and the accuracy of the threat model have direct causal effects.

The Core Security Procedures

Modern authenticated encryption, key exchange and secure channel establishment (TLS), access control enforcement, and intrusion/anomaly detection are the production-grade algorithms that real systems depend on.

Each has a clear specification, security proof under stated assumptions, and well-understood engineering trade-offs.

Protection Domains and Controlled Information Flow

A secure system can be modeled as a set of protection domains with carefully controlled information flows between them. Threats act as disturbances. Security mechanisms (prevention, detection, response, recovery) form the feedback loops that keep the system in an acceptable security state despite ongoing attacks.

The Eternal Arms Race under Human and Economic Constraints

Security engineering is uniquely difficult because the adversary is intelligent and adaptive, the requirements are often in direct tension with usability and performance, and the economic incentives frequently favor features and speed over hardening. The field is defined by the constant need to raise the cost of attack while keeping the cost of defense manageable.

The substrate here captures the essential objects, flows, and trade-offs that security engineers and researchers work with every day.