37-EFT.WP.EDX.HighSpeed v1.0 | Chapter 14 — Simulation Stack & Benchmark Cases (Methods.SimStack)

Home ／ Docs-Technical WhitePaper ／ 37-EFT.WP.EDX.HighSpeed v1.0

Chapter 14 — Simulation Stack & Benchmark Cases (Methods.SimStack)

I. Chapter Objectives & Structure

1. Objective: Build a multi-layer Simulation Stack (SimStack) for high-speed scenarios that, under a unified dialect, executes the full forward → inversion → validation → release workflow from netlist/layout/binding and 3D-EM artifacts to Z_eft(omega), T_arr, T_group, w_{p,m}(omega), and ΔZ_rad(omega), and provide reproducible sim–meas alignment and a benchmark case library.
2. Structure: Stack layers → Shared dialect & records → Numerical methods & stability → Benchmark cases → Sim–meas/EM alignment → Falsifiability → Data structures & export → Compliance templates → Cross-chapter closure.
3. Shared time-of-arrival dialect (equivalent; explicit gamma(ell) & d ell; record delta_form):
   • Constant-factored: T_arr = ( 1 / c_ref ) * ( ∫ n_eff d ell )
   • General: T_arr = ( ∫ ( n_eff / c_ref ) d ell )

II. Simulation Stack Layers (HF, bottom-up)

1. Physics Core (S20-HF / S30-HF / S50-HF):
   • S20-HF: group delay and linear-phase approximation; S30-HF: alpha/beta/k and amplitude/phase dispersion;
   • S50-HF: Z_eft = Z_ref + ΔZ_T [+ ΔZ_rad], with Re{ΔZ_rad} ≥ 0 and K–K consistency.
2. Path/Mode Engine (I30-HF / Chapter 6):
   layout ↔ gamma(ell) binding; segment {Δell_i, n_eff(i)}; w_{p,m}(omega) and modal coupling C(omega).
3. Radiation Gate (Chapter 5):
   ΔZ_rad(omega) estimation and gating, coordinated with near/far fields and I_CM.
4. S↔Z & Normalization (Chapter 8):
   Multiport S→Z mapping and mixed-mode normalization; consistency of Znorm(omega)/Z_c(omega).
5. Solvers:
   Frequency-domain sparse/ABCD composition and FFT convolution; stable derivatives ∂Z/∂θ and regularization for sensitive parameters.
6. UQ & Inversion (Mx-*):
   Priors/likelihood/evidence (Laplace/nested sampling), PPC, and uncertainty propagation to KPIs.
7. I/O & Cards:
   Data/model/pipeline cards; mandatory arrival and qa_gates; version locks and reproducible seeds.

III. Shared Dialect & Records (mandatory)

• Arrival: choose one dialect; explicitly record gamma(ell), d ell, c_ref, and delta_form.
• Units/dimensions: SI; key equalities must pass check_dim = pass.
• Passivity/causality: Re{Z_eft} ≥ 0, K–K consistency; if ΔZ_rad is enabled, Re{ΔZ_rad} ≥ 0.
• Normalization/mixed-mode: unify Znorm(omega) and T_mm/Z0_mm before S→Z; after mapping, re-check passivity/K–K.
• Coherence-window KPIs: E_phase/GDR compliant; use energy composition out of window.

IV. Numerical Methods & Stability (HF focus)

• Group delay/phase: after sync correction, T_group = d/domega(arg Z_eft); linear-phase fitting limited to the window Ω.
• Propagation-constant extraction: beta ≈ unwrap(arg Z_eft)/L (approx.) or supplied by transmission/3D-EM; enforce T_group ≈ L·dβ/dω as a consistency gate.
• Amplitude dispersion: under matching, alpha ≈ −ln|S21|/L; apply band-limited regularization near singularities.
• Modal coupling: C(omega) at bends/splits constrained by power conservation; |C_{mn}|∈[0,1].
• Stability tactics: residual whitening and spectral flatness checks; smooth/band-limit priors for σ_eff(ω)/n_eff(ω); forbid right-half-plane poles.

V. Benchmark Case Library (HF, minimally reproducible)

• SIM-HF-01 | Path switch A/B (P1): equal-length dual traces, guards/returns differ; metrics: k_φ ≈ ΔT_arr, E_phase/GDR.
• SIM-HF-02 | Dielectric/conductor small perturbations (P2): slow sweeps of temperature/dispersion; metrics: smooth renormalization + passivity/K–K.
• SIM-HF-03 | Radiation gap/stub (P3): sealable; metrics: Re{ΔZ_rad} ≥ 0 and decreased after sealing, field consistency.
• SIM-HF-04 | Via/plane transition: tunable stub length/returns; metrics: ΔW drift and E_phase/GDR change.
• SIM-HF-05 | Connector/launch transition: taper + ground-finger optimization; metrics: ΔZ_c(Ω) and no KPI degradation.
• SIM-HF-06 | Board-to-cable coupling (CM): return strategies; metrics: I_CM consistent with ΔZ_rad/ΔW trends.

Each case ships dataset_card/pipeline_card/env_lock and reference outputs for regression and release gates.

VI. Sim–Meas / 3D-EM Alignment (I30-HF / I40-HF / M10-HF)

• Port harmonization: map_ports / em_port_align; unify Znorm(ω) and, if used, T_mm/Z0_mm.
• De-embed/renorm: deembed → renorm, record baseline_id and artifact hashes.
• S→Z mapping: Z_eft = Z0^{1/2}(I+S)(I−S)^{-1}Z0^{1/2}; immediately check passivity/K–K.
• Timebase correction: arg Z_corr = arg Z_raw − ( ω·Δt_sync ).
• Path correction & arrival: write arrival{form,gamma,measure,c_ref,Tarr,u_Tarr,delta_form}; only after two-dialect T_arr agreement compute KPIs.
• KPI/consistency: E_phase/GDR pass; optionally T_group ≈ L·dβ/dω; compute ΔW.
• Radiation channel (if enabled): ΔZ_rad = Z_eft − (Z_ref+ΔZ_T); verify Re{ΔZ_rad} ≥ 0 and pre/post-sealing monotonicity.
• Dataset card output: summarize Z_eft/argZ/T_group/KPIs/ΔZ_rad and QA results.

VII. Falsifiability (SimStack-specific)

• F-HF-1 (Window consistency): failure of E_phase/GDR or T_group ≈ L·dβ/dω → reject window or β extraction.
• F-HF-2 (S→Z/normalization): Re{Z_eft} < 0 or K–K failure → reject normalization/mapping.
• F-HF-3 (Radiation positive-real): Re{ΔZ_rad} < 0 or no decrease after sealing → reject radiation equivalence.
• F-HF-4 (Path/mode): k_φ ≉ ΔT_arr or anomalous ΔW trend → reject path modeling or C(ω).
• F-HF-5 (Regression): rerun vs reference outputs exceeds thresholds → reject release.

VIII. Data Structure & Export (minimal template)

simstack_hf:

case_id: "SIM-HF-01"

model_id: "EDX-HS-eft-ms"

freq_grid_Hz: [...]

layout_ref: "LAY-2025-001"

binding_ref: "LAY2PATH-HF-0001"

deemb: {method:"TRL", version:"1.2", artifact:"/artifacts/deemb.json", baseline_id:"BLSN-EDX-001"}

Znorm_ohm: [50.0, 50.0]

sync: {dt_sync_s: 2.0e-12}

arrival:

form: "n_over_c" # or "one_over_c_times_n"

gamma: "explicit"

measure: "d_ell"

c_ref: 299792458.0

Tarr_s: 1.234e-09

u_Tarr_s: 6.0e-12

delta_form: "n_over_c"

outputs:

Z_eft: {real:[...], imag:[...]}

argZ: [...]

T_group_s: [...]

weights: {w_main:[...], w_side:[...]}

deltaZ_rad: {Re_ohm:[...], Im_ohm:[...]} # optional

kpis:

E_phase_rad: 0.043

GDR_s: 1.8e-10

ΔW: 0.17

qa_gates: {check_dim:"pass", passivity:"pass", KK:"pass"}

seed: 20250915

IX. Compliance Templates (APIs & computation snippets)

• API prototypes

api:

- id: "SimStackHF.build"

proto: "build(netlist, layout, binding_hf, options) -> sim_handle"

- id: "SimStackHF.forward"

proto: "forward(sim_handle, theta, grid) -> {Z_eft, argZ, T_group, w_{p,m}, ΔZ_rad?}"

- id: "SimStackHF.invert"

proto: "invert(sim_handle, data, priors, sampler:'NUTS') -> {posterior, logZ, summary}"

- id: "SimStackHF.ppc"

proto: "ppc(sim_handle, posterior, grid) -> {residual_spectrum, gates}"

- id: "SimStackHF.export"

proto: "export(sim_handle, format:'cards|json') -> artifacts[]"

• HF KPI computation (pseudocode)

phi = unwrap(argZ_corr[ω1:ω2]) # coherence window

T_group = grad(phi, ω)

E_phase = max_abs(phi - (ω*Tarr + phi0_opt))

GDR = max_abs(T_group - median(T_group))

assert E_phase <= E_phase_gate and GDR <= GDR_gate

X. Cross-Chapter Links & Closure

• Dependencies: Chapter 4 (S20-HF/S30-HF), Chapter 5 (S50-HF), Chapter 6 (S40-HF), Chapter 7 (KPIs), Chapter 8 (S↔Z), Chapters 9–11 (Metrology/Falsification/Co-simulation).
• Downstream: Chapter 12 (Design Rules — layout review & sign-off KPIs), Chapter 15 (Data & Repro — data/pipeline/env-lock), Chapter 16 (Design Protocol & Checklist — include regression gates in sign-off).