29-EFT.WP.TBN.Measurement v1.0 | Appendix B Contract Library & Strategy Cards (C50 Full Set)

Home ／ Docs-Technical WhitePaper ／ 29-EFT.WP.TBN.Measurement v1.0

Appendix B Contract Library & Strategy Cards (C50 Full Set)

• One-line Objective: Provide executable contracts C50-* and “strategy card” handling flows that span the TBN metrology chain from mathematical baseline → acquisition → estimation → delay & corrections → instrument metrology → spectrum/time-domain consistency → analytics/measurement/anchor → uncertainty → runtime. The goal is to rigidly guarantee: two-form parity, same window/bandwidth/units, dimensional compliance, and RefCond traceability—while supporting manifest-based publication and replay auditing.
• I. Scope & Subjects
• Subjects: y(t), S_phi/S_y, sigma_y(τ), T_arr^{form1/form2}, T_arr*, lat_*, corr_env, ΔT_map, ΔT_obs, ΔT_inst/ΔT_proc/ΔT_asym, and runtime metrics delta_form_rt, hit, stale_ratio, latency.
• Inputs: analytical/configuration form (models/paths/corrections/instrument/network) and observational/implementation form (PLL/IQ/FFT/timestamps); RefCond (hash/age/coverage); u/U with thresholds; panel window W and RBW/ENBW.
• Outputs: contracts.rules, contracts.report, strategy.cards, and the handling records in manifest.tbn.* (including evidence URIs and signature chain).
• II. Naming, Severity & Common Fields
• Naming: C50-<domain><index>; domain→chapter mapping:
  2x Mathematical Baseline, 3x Acquisition/Timebase, 4x Estimators, 5x Allan Family, 6x Link Delay, 7x Path/Environment Corrections, 8x Instrument/Processing Chain, 9x Spectrum⇄Time Consistency, 10x Analytics/Measurement/Anchor Replay, 11x Uncertainty/Guardband, 12x Contracts & Strategy, 13x Runtime Streaming, 14x Security & Integrity, 15x End-to-End.
• Severity: sev ∈ {blocker, high, medium, low}; Action: action ∈ {fail, warn, degrade, bypass}.
• Common record: {rule_id, sev, metric, unit, dim, threshold|range, window, samples, RefCond.hash, evidence_uri, pass, msg, ts}.
• Hard prerequisites: two-form parity / same window & bandwidth / dimensional compliance / traceability — if any fails, handle it before comparing other metrics.
• III. Domain-Level Contracts Master Table (C50 Full Set)
  If scenario-specific thresholds conflict with defaults, the use-case/environment thresholds prevail (recorded in contracts.thresholds). Default unit/dimension: unit(z)=1, unit(T)='[T]'.
• 2x Mathematical Baseline (phase/frequency/arrival-time)
• C50-201 Two-form delta: delta_form_p95 ≤ tol_Tarr.
• C50-202 Window/bandwidth consistency: RBW/ENBW/window_fn must match the manifest.
• C50-203 SNR/CRLB: estimators satisfy SNR ≥ SNR_min and T_obs ≥ T_min.
• C50-205 Dimensional compliance: check_dim(S_phi)="rad^2/Hz", check_dim(S_y)="1/Hz", check_dim(sigma_y)=1, check_dim(T_arr)='[T]'.
• 3x Acquisition Front-Ends & Timebase
• C50-301 Timebase anchoring: |offset| ≤ offset_max, |skew| ≤ skew_max, |J|_p95 ≤ J_max.
• C50-302 Sampling/trigger jitter: t_j,rms ≤ t_j,max, t_trig,rms ≤ t_trig,max.
• C50-303 ENOB/SNR: ENOB ≥ ENOB_min or SNR_meas ≥ SNR_min.
• C50-305 Multichannel consistency: inter-channel delay/phase biases ≤ τ_ch,max/φ_ch,max.
• C50-306 Two-form budget: front-end contribution satisfies u(delta_form | front-end) ≤ budget.
• 4x Phase & Frequency-Offset Estimation (PLL/IQ/FFT/Hilbert)
• C50-401 CRLB/window length/SNR: var_emp ≤ α·CRLB (α ≤ 3).
• C50-402 Window/leakage: persisted RBW/ENBW/window_fn must match; side-lobe limit L_sidelobe within threshold.
• C50-404 Two-form alignment: estimator sequences align in the same window with T_arr^{form1/2}, recording method components.
• 5x Allan Family
• C50-501 Units/dimensions: check_dim(sigma_y)=1, check_dim(TDEV)='[T]'.
• C50-502 Measure conventions: overlap/no-dead-time, τ grid, and ENBW corrections must be persisted.
• C50-503 ν_eff & coverage: coverage(U) ≥ 95%; short sequences must degrade or be extended.
• 6x Link Delay & Jitter
• C50-601 Two-form delta: delta_form_p95 ≤ tol_Tarr.
• C50-602 Symmetry/asymmetry: without one-way anchors, |ΔT_asym|_p95 ≤ τ_asym_max; with anchors, match-gate ≤ τ_match.
• C50-603 Jitter & spectrum: σ_t_p95 ≤ σ_t,max; in-band S_t(f) respects spectral gates.
• C50-604 CRLB/SNR: var_emp(T̂) ≤ α·CRLB, SNR ≥ SNR_min.
• 7x Path & Environmental Corrections
• C50-701 Two-form delta (post-correction): delta_form_p95 ≤ tol_Tarr.
• C50-702 Freshness/coverage: age(met/TEC/n_field) ≤ Δt_max, coverage ≥ cov_min.
• C50-703 Mapping consistency: ΔT_map_p95 ≤ tol_map, with u(ΔT_map) persisted.
• C50-704 Multipath control: MER_p95 ≤ MER_max or MP_res ≤ MP_res_max.
• 8x Instrument & Processing Chain
• C50-801 Two-form delta: delta_form_p95 ≤ tol_Tarr.
• C50-802 Asymmetry ceiling: |ΔT_asym|_p95 ≤ τ_asym_max.
• C50-803 Evidence completeness: loopback/peer/triangle timestamp sets or correlation evidence URIs must be persisted.
• C50-804 Spectrum/jitter: σ_t_p95 ≤ σ_t,max, S_t(f) compliant.
• 9x Spectrum ⇄ Time-Domain Consistency
• C50-901 Same window & bandwidth: mappings use the same W/RBW/ENBW/window_fn.
• C50-902 Unit compliance: check_dim(σ_t)='[T]', check_dim(σ_φ)='rad', check_dim(EVM)=1.
• C50-903 Two-form companions: delta_form_p95 ≤ tol_Tarr, ΔT_obs_p95 ≤ tol_align.
• C50-904 EVM consistency: |EVM_pred − EVM_meas|_p95 ≤ tol_evm.
• 10x Analytics vs Measurement vs Anchor/Replay
• C50-1001 Hard two-form gate: if delta_form_p95 ≤ tol_Tarr fails → reject publish.
• C50-1002 Same window/bandwidth/units: if inconsistent → recompute first.
• C50-1003 Residual upper bounds: |resid_*| + k·u(resid) ≤ threshold_*.
• C50-1004 Anchor/replay completeness: anchor and replay.core evidence URIs/hashes must be persisted.
• 11x Uncertainty & Guardband
• C50-1101 GUM/MC consistency: 0.8 ≤ ρ ≤ 1.25.
• C50-1102 Coverage for publication: report U for key quantities with confidence ≥ 95%.
• C50-1103 Two-form gate with uncertainty: delta_form + k·u(delta_form) ≤ tol_Tarr.
• C50-1104 Mapping/alignment gates: ΔT_map + k·u(ΔT_map) ≤ tol_map, ΔT_obs + k·u(ΔT_obs) ≤ tol_align.
• 12x Contracts & Strategy
• C50-1201…1222: see Ch.12 (two-form/alignment/spectral gates/freshness/coverage/signatures/bypass, etc.).
• 13x Runtime Streaming
• C50-1301 Two-form & alignment at runtime: p95(delta_form_rt_T) ≤ tol_Tarr, p95(ΔT_obs) ≤ tol_align.
• C50-1302 Frequency/statistics gates: p95(|resid_Sy|_B) ≤ τ_Sy; sigma_y slope consistency.
• C50-1303 Latency/jitter: latency_p95 ≤ SLO.latency_p95, jitter_p95 ≤ τ_jit_max.
• C50-1304 Cache/freshness: hit ≥ hit_min, stale_ratio ≤ tol_stale, panel updates ≤ Δt_panel_max.
• 14x Security & Integrity
• C50-1401 Two-form hard gate first: unmet → reject/degrade.
• C50-1402 Spectral integrity: D_spec ≤ τ_spec, ACLR ≥ ACLR_min, OBB ≤ OBB_max.
• C50-1403 Timing integrity: Timing_anom_p95 ≤ τ_T.
• C50-1404 Energy balance: E_balance_p95 ≤ τ_E.
• C50-1405 Signatures/replay: Sig_ok=1 and Replay_ok=1.
• 15x End-to-End
• C50-1501 End-to-end gates: delta_form_p95 ≤ tol_Tarr, ΔT_obs_p95 ≤ tol_align, resid_Sy_p95 ≤ τ_Sy, coverage ≥ 95%, signatures verifiable.
• C50-1502 Uncertainty & runtime: coverage(U) ≥ 95%, 0.8 ≤ ρ(GUM/MC) ≤ 1.25, runtime hit ≥ hit_min.
• IV. Strategy Cards (YAML Template & Canonical Cards)
• Template
• card_id: <string>
• triggers: [C50-xxxx, ...]
• sev: blocker|high|medium|low
• decide:
• if: "<metric> comparator <threshold>" # or a boolean expression
• act:
• - api: I50-xx
• params: {...}
• - api: I50-yy
• params: {...}
• verify:
• metric: <name>
• window: "W=5min,p95"
• rollback:
• plan: "<bypass|lower_order|rollback_version>"
• notes: "<why/impact/evidence_uri>"
• Canonical strategy cards
• card.twoform_guard (two-form delta out of bounds)
• triggers: C50-201/601/801/1301
• act: recompute via I50-66 twoform_consistency; if needed, unify to form2 + I50-116 design_guardband.
• verify: delta_form_p95 ≤ tol_Tarr.
• card.align_fix (alignment error out of bounds)
• triggers: C50-1202/1301/1403
• act: I50-112 propagate_gum to update u(ΔT_obs); re-solve sync via I50-112/34; shrink window and recompute.
• verify: ΔT_obs_p95 ≤ tol_align.
• card.spectrum_recompute (spectral gate violation)
• triggers: C50-1204/1302/1402
• act: I50-33 correct_enbw / adjust RBW/ENBW/window_fn; write-back model.
• verify: p95(|resid_Sy|_B) ≤ τ_Sy.
• card.mapping_dualpub (phase/group mapping out of bounds)
• triggers: C50-703/1104
• act: publish phase/group dual forms side-by-side; increase tol_map guardband.
• verify: ΔT_map_p95 meets gate.
• card.source_refresh (freshness/coverage insufficient)
• triggers: C50-1211/1304
• act: switch sources / supplement acquisition; I50-103 interp_sources; degrade or reject.
• verify: age ≤ Δt_max, coverage ≥ cov_min.
• card.security_failclosed (signature/bypass/anomaly)
• triggers: C50-1405/1402/1403
• act: bypass/rollback, rate-limit; persist evidence and replay package.
• verify: Sig_ok=1, Replay_ok=1, anomaly rate reduced.
• V. Threshold Guidance & Calibration
• Two-form / mapping / alignment: tol_Tarr ≈ 1e-3 · T_arr; tol_map ≈ 0.1 · tol_Tarr; tol_align ≈ 0.02 · T_sym or 1e-3 · T_f.
• Spectral/statistical residuals: τ_Sy and tol_sigma depend on SNR / window / bandwidth and scenario calibration.
• Jitter/asymmetry: σ_t,max and τ_asym_max come from the standard and device specifications.
• Freshness/coverage: Δt_max tied to source refresh rate; cov_min ≥ 0.90.
• GUM/MC consistency: ρ ∈ [0.8, 1.25].
• All threshold changes must be written to contracts.thresholds and change_log with source hash and replay evidence.
• VI. Manifest Fields (manifest.tbn.contracts Minimum Keys)
• rules[]: status, window, threshold, RefCond.hash, and evidence_uri for each C50-*.
• metrics: { Q:{delta_form_p95, ΔT_obs_p95, resid_Sy_p95, resid_sigma_y_p95}, Int:{age, coverage, ΔT_map_p95}, Sec:{D_spec, Timing_anom, Sig_ok} }.
• decisions: executed strategy cards, actions, and verifications; rollback_to pointer.
• thresholds_uri: threshold table and source hash; signature: manifest signature.
• VII. Interface Mapping
• Contract computation: chapter-level M50-*.
• Remediation execution: this volume’s I50-* (called directly by strategy-card act.api).
• Manifest writing: emit_*_manifest for baseline/frontend/estimator/delay/corr/anchor/u/contracts/rt.
• VIII. Cross-References
• Chapters 2–15 (P/S/M/I/C); Appendix A (Interfaces) / C (Manifests) / D (Metrics) / E (Uncertainty); companion volumes TimeBase/Sync, PathCorrection, Packets.Light, PathRedshift, Instrument, Methods.Cleaning.
• Summary
• This appendix compiles the full C50-* set and standard strategy cards, grounded on four rigid constraints: two-form parity, same window/bandwidth, dimensional compliance, and traceability.
• Combined with I50-* and manifest.tbn.*, engineering teams can enforce a closed-loop governance for TBN metrology that is detectable, actionable, verifiable, and rollback-ready across the entire chain.