`araim-reference-check` (scenario kind)

Scenario kind of the engine v0.38.0, family GNSS measurements, positioning and integrity. Advisory: analysis output from a research and engineering tool, not type-approved navigation equipment and not a safety-of-life service. Every label here is the engine's or the verification matrix's; a kind is MODELLED unless the matrix says otherwise (how a label is earned: the README's Evidence section).

What it does#

Published ARAIM certification reference vectors: the engine's ARAIM horizontal and vertical protection levels run against the WG-C ARAIM Technical Subgroup's OWN worked numerical examples, at the tolerance (TOL_PL = 5e-2 m) those documents themselves state. Two vectors are committed as fixtures, each carrying its retrieval URL, retrieval date, source-file SHA-256 and page: the Reference Airborne Algorithm Description Document v3.1 (2019) Appendix D — self-consistent, and the acceptance vector — and the Milestone 3 Report (2016) Annex A section A.IX, the statement the research bibliographies cite. Every published input (the 10-satellite 2-constellation geometry matrix, the C_int and C_acc variance diagonals, b_nom, P_sat, P_const and the LPV-200 constant set) and every published output (VPL, HPL, EMT, sigma_v_acc, K_fa_3 and the two constellation-fault modes' sigma_3, sigma_ss_3 and b_3) is reported beside the engine's own number and their absolute difference. The protection-level equation is NOT re-solved here: the reference's mode list is handed to the engine's existing raim::araim_protection_level / araim_integrity_risk, so the bisection and the Gaussian-tail algebra under test are the engine's own; only the geometry, the sub-solutions, the detection thresholds and the published budget split are built for it. EXTERNALLY CHECKED against the 2019 vector: VPL 18.2926 m vs 18.3 m published, HPL 13.4063 m vs 13.45 m, EMT 7.2997 m vs 7.2998 m, sigma_v_acc 1.3694 m vs 1.3694 m, and all six published intermediates to within half a unit in their last printed decimal. The 2016 vector carries two internal defects the report states rather than hides — a sign typo in row 3 of G, and a K_fa_3 evaluated at 57 fault modes while the document states N_fault_max = 1 (12 modes) — so its geometry intermediates reproduce exactly while its protection levels are reported as measured discrepancies and excluded from the acceptance figure; the tolerance is never widened to absorb them. Every emitted numeric field carries a unit and a provenance class, and a transcribed figure is labelled published so it can never be mistaken for a computed one. SCOPE: only N_fault_max = 1 (single-satellite and single-constellation fault modes) is implemented, and a case whose priors require simultaneous multi-event fault subsets is REFUSED rather than silently truncated; fault detection, exclusion, the chi-square consistency check and the double-counting re-allocation step of the reference algorithm are out of scope. This reproduces a published reference algorithm's worked example — it is not a certification, an airworthiness artefact or an approval.

One run of scenarios/araim-reference-check.toml summarises as: araim-reference-check: 2 published WG-C reference vector(s); worst VPL/HPL agreement 0.0437 m against the reference's own TOL_PL = 0.05 m over 1 acceptance vector(s)

Inputs#

From the schema (kshana kinds --json):

Required fields: none
Optional fields: vector

The example scenarios/araim-reference-check.toml sets, among others:

Field	Value	Unit
`kind`	araim-reference-check	text

Run it#

kshana scenarios/araim-reference-check.toml

What it gives#

The scalar results of that run, with the units the engine states:

Result	Value	Unit
`acceptance_met`	true	flag
`acceptance_tolerance_m`	0.05	m
`acceptance_vectors`	1	count
`vectors_checked`	2	count
`worst_acceptance_error_m`	0.043709	m

Result label (the engine's): EXTERNALLY CHECKED — the engine's ARAIM protection levels run against the published worked numerical examples of the EU-U.S. Working Group C ARAIM Technical Subgroup reference airborne algorithm, at the tolerance (TOL_PL) those documents themselves state. Every published figure carries its source UR

The chart the engine writes for scenarios/araim-reference-check.toml, kind araim-reference-check

Evidence#

VALIDATED: the verification matrix rows that name this kind.

VALIDATED The multiple-hypothesis solution-separation protection-level equation itself: all-in-view and per-fault-mode weighted least squares on an East/North/Up geometry with one clock state per constellation, the sub-solution sigmas, solution-separ Tests: araim_reference_vectors.rs.

The rows the run itself exercised, with the oracle the matrix gives each (the public sources are named there):

MODELLED Deterministic, scenario-hashed, SBOM + cross-platform golden gates. Oracle (ExternalDataset): SBOM conformance to the official CycloneDX 1.5 JSON Schema (+ valid SPDX identifiers) — an external published standard, zero validation errors over the full dependency graph; the FoM-determinism / byte-reproducibility part remains a pinned self-consistency check, so the row stays MODELLED
VALIDATED The multiple-hypothesis solution-separation protection-level equation itself: all-in-view and per-fault-mode weighted least squares on an East/North/Up geometry. Oracle (ExternalDataset): EU-U.S. Working Group C ARAIM Technical Subgroup, Reference Airborne Algorithm Description Document v3.1 (20 June 2019), Appendix D worked numerical example — published INPUTS (10-satellite 2-constellation geometry matrix, C_int/C_acc diagonals, b_nom, P_sat, P_const, LPV-200 constants) and published OUTPUTS (VPL 18.3 m, HPL 13.45 m, EMT 7.2998 m, σ_v_acc 1.3694 m, K_fa_3 5.1083, and six constellation-fault intermedi

Limitations#

What the run's own report says is not modelled or not validated:

Reproducibility & software assurance is MODELLED, not validated: a sub-claim is externally checked, but the whole capability composes modelled pieces, so the capability stays Modelled. (verification matrix (docs/MODELLED-RATIONALE.md))

araim-reference-check (scenario kind)