PNT-resilience framework cross-walk
June 20, 2026 · View on GitHub
Aligned to, not certified. This maps Kshana's simulated capabilities to the
published PNT-resilience frameworks. It is a simulation-derived self-assessment
aligned to the DHS/CISA Resilient PNT Conformance Framework (RPCF) v2.0, the
RethinkPNT/Firesmith Resist-Detect-Respond-Recover (RDRR) model, and Yang
Yuanxi's resilient-PNT criteria. It is not a certification, accreditation, or
compliance statement, and it carries no endorsement from DHS, IEEE, or any
authority. Every row is labelled with its honest verification status
(VALIDATED against an external oracle, or MODELLED from first principles with
tests); the resilience-scoring layer itself is MODELLED.
The scoring engine (src/resilience/) consumes these capabilities and emits
per-dimension sub-scores, never a single phantom number. See
src/verification.rs for the machine-checked status invariants.
DHS RPCF v2.0 technique categories (the seven; "OLVIDMR" is our mnemonic)
| Category | Kshana capability used as the sub-score driver | Module | Status |
|---|---|---|---|
| Obfuscate | Declared technique × source quality (no direct behavioural driver) | resilience::score | MODELLED |
| Limit | Declared technique × source quality | resilience::score | MODELLED |
| Verify | Impairment-detector AUC under the scenario (spoof/jam/meacon monitors) | impairment_eval, resilience::score | MODELLED |
| Isolate | Declared technique × source quality | resilience::score | MODELLED |
| Diversify | Independent-group count → inverse-Simpson effective diversity | resilience::diversity | MODELLED |
| Mitigate | Availability under denial | fom, resilience::score | MODELLED |
| Recover | Holdover coast × bounded-degradation gate | fom, holdover, resilience::timeline | MODELLED |
RethinkPNT / Firesmith RDRR functions
| Function | Driver | Module | Status |
|---|---|---|---|
| Resist | Mean of Obfuscate/Limit/Isolate/Diversify sub-scores | resilience::score | MODELLED |
| Detect | Verify sub-score (detector AUC) | resilience::score | MODELLED |
| Respond | Mitigate sub-score (availability) | resilience::score | MODELLED |
| Recover | Holdover × bounded gate; resilience-timeline KPIs (detect/react/recover/loss) | resilience::timeline | MODELLED |
Yang Yuanxi criteria (the subset the timing/detection FoMs speak to)
| Criterion | Driver | Module | Status |
|---|---|---|---|
| Availability | Availability under denial | resilience::score | MODELLED |
| Reliability | Filter integrity fraction | fom, resilience::score | MODELLED |
| Continuity | Holdover × bounded gate | resilience::timeline | MODELLED |
| Accuracy | Timing/quality proxy — position-domain accuracy is NOT modelled (fom::positioning_performance returns an explicit gap) | resilience::score | MODELLED (timing-domain only) |
What this cross-walk does not claim
- No certified RPCF Level. The assigned Level is a tentative, simulation-derived reading with an explicit bounded-degradation gate, not a conferred maturity.
- No position-domain accuracy. Kshana's figures of merit are timing-domain and detection metrics; position CEP/SEP/HPL require a multi-axis nav solution Kshana does not yet provide.
- No field validation. The reference panel and threat ensemble are synthetic,
parameter-grounded reductions (
resilience::panel), stated as a first-class limitation.
The companion study (resilience::study) exists precisely to show that collapsing
these dimensions into one composite score or one Level produces a rating whose
architecture ranking is unstable under defensible weighting and threat choices —
so this cross-walk is a measurement layer to be read per-dimension, not a single
grade to certify against.