Resilient PNT, Timing & Navigation Integrity

Five "independent" sources can secretly be two

A PNT receiver can show GNSS, network time, inertial and visual sources all agreeing — and still be wrong, because several of those sources quietly share the same upstream clock, calibration, or map. Five agreeing readings built on two real dependencies isn't five times more confident; it's the same blind spot repeated five times.

A clock that keeps ticking after losing its reference doesn't announce that it's lost authority

An oscillator can continue producing timestamps after its external time reference disappears — and unless something is tracking how long it's been in that holdover state and how much its error has grown, it looks exactly as authoritative as it did before the reference was lost.

Reviewing that disagreement doesn't require raw receiver control

This runs on redacted source-agreement, timing and holdover-drift telemetry your PNT system already logs — never raw GPS/GNSS receiver control, classified timing-source data, or live navigation-system access.

Scenario: Four "independent" sources, one hidden shared dependency.

A PNT receiver ingests GNSS, cellular network time, an inertial unit, and a visual odometry feed — four sources that look independent. Unknown to the integrator, the cellular time service and the GNSS receiver both derive their epoch from the same upstream NTP server after a recent network reconfiguration.

WITHOUT this check: All four sources agree within tolerance. The fusion filter reports HIGH_CONFIDENCE based on four independent confirmations — when two of those four are really the same dependency counted twice.

WITH this check: The dependency graph flags that GNSS and cellular-time share a common upstream clock. Confidence is reported as MODERATE — 2 INDEPENDENT ROOTS, NOT 4, and the shared dependency is disclosed in the review record.

Scenario: A receiver tracking multiple satellite constellations sees GPS suddenly disagree with its own recent trajectory.

A coordinated spoofing attack can manipulate several GPS satellites together, so the GPS solution still looks internally consistent even though it's wrong.

WITHOUT this check: The GPS solution is trusted because nothing inside the GPS constellation itself contradicts it — there's no obvious internal red flag to catch.

WITH this check: The same position and time solution is cross-checked against Galileo and BeiDou, which are still reporting consistent, expected values. GPS disagreeing while two independent constellations agree with each other is flagged as a likely single-constellation spoofing candidate — a real environmental effect would typically show some correlated impact across constellations, not isolate to just one.

Do not buy the claim. Test the gate.

Feed it a synthetic multi-source PNT log where two "independent" sources actually share an upstream dependency you've deliberately hidden, alongside a scenario where only one constellation shows a coordinated anomaly. It succeeds only if it flags the false independence and correctly isolates the single-constellation spoofing candidate — without you having to tell it where to look.