Network Security — Scientific Principles

Scientific Principles

Start with assets and trust boundaries. An organisation needs to know its devices, services, accounts and connections before it can decide which access is necessary. A trust boundary separates systems or users with different permissions; traffic crossing it should be evaluated rather than assumed safe because it originates inside a building.

Authentication checks an asserted identity. Authorisation determines what that identity may do. A valid login does not justify access to unrelated records or administrative controls. Least privilege limits permissions to the task and reduces the consequences of a compromised account. Segmentation separates parts of a network so that a problem in one part is less able to spread.

Encryption protects data against specified forms of exposure in transit or storage, but an authorised endpoint may still reveal or alter it. Logging and monitoring provide evidence of events; they are useful only when someone can interpret and act on them. Patching addresses known weaknesses, while secure configuration reduces unnecessary services and exposure.

NIST's Cybersecurity Framework 2.0, published in 2024, uses Govern, Identify, Protect, Detect, Respond and Recover. These are related functions rather than a once-only sequence. Governance assigns responsibility; recovery planning and tested backups support availability after disruption. In India, CERT-In's national incident-response role under section 70B of the IT Act is distinct from NCIIPC's critical-information-infrastructure role under section 70A.