08-EFT.WP.Core.Sea v1.0 | Appendix A — Device Field Dictionary and Examples

Home ／ Docs-Technical WhitePaper ／ 08-EFT.WP.Core.Sea v1.0

Appendix A — Device Field Dictionary and Examples

I. Field Hierarchy and Naming Conventions

• Namespaces
  sensor.* (identity & specs), cal.* (calibration & coefficients), acq.* (acquisition configuration), time.* (timebase & sync), filt.* (conditioning & anti-alias), spec.* (spectrum & features), env.* (environment & corrections), path.* (path & arrival time), quality.* (quality & drift), io.* (serialization & manifests).
• Foundation rules
  All symbols and field names are English plain text; inline symbols are wrapped in backticks (e.g., fs, ENOB, H(f), T_arr, gamma(ell)).
  Record both ts (UTC, ISO-8601) and tau_mono (monotonic clock); any expression with division, integration, or composite operators must use parentheses and declare the path or measure, e.g., L_gamma = ( ∫ 1 d ell ).
  Arrival time dual forms must co-exist:
  T_arr = ( 1 / c_ref ) * ( ∫ n_eff d ell ) and T_arr = ( ∫ ( n_eff / c_ref ) d ell ); the difference is reported as delta_form.

II. sensor.* — Device Identity and Specs

• sensor.sid : str — globally unique sensor identifier. Example: SNSR-IMU-0001.
• sensor.model : str — device model, e.g., ADX-1234.
• sensor.serial : str — factory serial number.
• sensor.class : str — device class, class ∈ {"accel","mic","rf","imu","temp","pressure"}.
• sensor.cal_id : str — active calibration version.
• sensor.A_gain : float, sensor.B_bias : float, sensor.C_offset : float — first-order linear calibration coefficients; applied as x_corr = A_gain * ( x_raw - B_bias ) + C_offset.
• sensor.ENOB : float, sensor.ADC_bits : int — effective number of bits and ADC bit width.
• sensor.DR : float — dynamic range (ratio or dB).

III. acq.* — Sampling and Acquisition Configuration

• acq.fs_nom : float, acq.fs_max : float, acq.fs : float — nominal, max, and actual sampling rate (Hz). Constraint: fs <= fs_max; Nyquist hard constraint fs >= 2 * f_max.
• acq.Delta_t : float — analysis window length (s).
• acq.window : str — window function { "rect","hann","hamming","blackman","flattop" }.
• acq.t0 : ts, acq.t1 : ts — UTC timestamps delimiting the acquisition interval.
• acq.pretrigger : float, acq.posttrigger : float — preserved durations (s) around trigger.
• acq.fs_hat : float — empirical rate estimate, fs_hat = ( N - 1 ) / ( tau_{N-1} - tau_0 ).

IV. time.* — Timebase and Synchronization

• time.clock_id : str, time.sync_ref : str — clock ID and reference, e.g., "ptp", "gps", "ntp".
• time.alpha : float, time.beta : float — skew/offset in ts_i(t) = alpha_i * tau_mono + beta_i.
• time.J : float — timing jitter (s or ms).
• time.u_alpha : float, time.u_beta : float — uncertainty of alpha/beta.
• time.ts_start : ts, time.ts_end : ts — UTC boundaries of the data block.
• time.tau_start : float, time.tau_end : float — monotonic clock boundaries (s, from system epoch or common origin).

V. filt.* — Conditioning and Anti-Aliasing

• filt.H_rev : str — filter-chain revision (must not be confused with T_fil/T_trans).
• filt.chain : list[dict] — ordered filter stages, e.g., {type:"lp", order:6, f_c:4800.0, ripple_db:0.1, stop_db:80.0}.
• filt.f_c : float, filt.BW : float — primary cutoff and passband width (Hz), recommended f_c in [0.4, 0.45] * fs.
• filt.tau_g : float — group delay estimate (s), report as tau_g(H).
• filt.AGC : dict|None — AGC settings {target_dBFS, attack_ms, release_ms, limit_db}.

VI. spec.* — Spectrum and Features

• spec.method : str — PSD estimator, default "welch".
• spec.seg : int, spec.overlap : float, spec.window : str — Welch segment count, overlap, and window.
• spec.fft_len : int, spec.df : float — FFT length and frequency resolution.
• spec.S_xx_units : str — PSD units (e.g., "power/Hz").
• spec.features : dict — suggested keys: {"band_power_[f1_f2]","crest","THD","kurtosis","SNR_dB"}.
• spec.E_psd : float — spectral error metric
  E_psd = ( ∫_{0}^{BW} | S_xx_meas(f) - S_xx_ref(f) | d f ) / ( ∫_{0}^{BW} S_xx_ref(f) d f ).

VII. env.* — Environment and Corrections

• env.temp : float, env.hum : float, env.press : float — environmental readings (units per sensor domain, e.g., degC, %RH, Pa).
• env.RefCond : dict — reference conditions, e.g., {temp:23.0, hum:45.0, press:101325.0}.
• env.correction_applied : bool — whether corr_env(x; RefCond) has been applied.
• env.U : float, env.u_c : float, env.k : float — expanded uncertainty U = k * u_c.

VIII. path.* — Path and Arrival Time

• path.gamma_desc : str — geometric description or key of the gamma(ell) path.
• path.L_gamma : float — path length L_gamma = ( ∫ 1 d ell ).
• path.c_ref : float, path.n_eff_model : str — reference speed and effective index model revision.
• path.T_arr_meas : float, path.T_arr_model : float — measured/modelled arrival time (s).
• path.delta_form : float — dual-form difference
  delta_form = | ( 1 / c_ref ) * ( ∫ n_eff d ell ) - ( ∫ ( n_eff / c_ref ) d ell ) |.
• path.toa_model_rev : str — arrival-time model revision.

IX. quality.* — Quality and Drift

• quality.q_score : float — quality score q_score ∈ [0,1].
• quality.clip_rate : float, quality.missing_rate : float — clipping ratio and missingness (from m ∈ {0,1} source mask).
• quality.D : float — drift magnitude (consistent with Chapter 9), e.g., D = || mu_x - mu_ref || / sigma_ref.
• quality.alert : dict|None — last alert summary {kind, ts, level}.

X. io.* — Serialization and Manifests

• io.fmt : str — file/stream format, fmt ∈ {"jsonl","csv","parquet","nc","tfrecord"}.
• io.compress : str|None — compression algorithm (e.g., "zstd").
• io.partition_by : list[str] — lake partition keys, e.g., ["ts_date","sid"].
• io.bytes_raw : int, io.bytes_stored : int, io.bytes_compressed : int — raw/stored/compressed sizes; storage efficiency SE = bytes_raw / bytes_stored, compression ratio CR = bytes_raw / bytes_compressed.
• io.manifest_rev : str — manifest revision; must anchor window, window function, filter and model versions.

XI. Integrity and Consistency Checks (Extract)

• Sampling stability: fs >= 2 * f_max, record fs_hat and |fs_hat - fs_nom| / fs_nom.
• Timebase consistency: all delays and jitter on tau_mono; ts for publication/audit.
• Anti-aliasing: provide H(f) and f_c, report tau_g(H).
• Arrival time: provide gamma(ell), d ell, c_ref, n_eff, and delta_form.
• Quality loop: missing marked with m=0; q_score and D must be traceable to raw chunks and parameter versions.

XII. Example 1 — LAB Accelerometer

{

"sensor": {

"sid": "SNSR-ACC-0007",

"model": "ADX-1234",

"serial": "SN-A1B2C3",

"class": "accel",

"cal_id": "CAL-2025-01",

"A_gain": 0.9987,

"B_bias": 0.0012,

"C_offset": 0.0000,

"ENOB": 17.6,

"ADC_bits": 24,

"DR": 98.0

},

"acq": {

"fs_nom": 20000.0,

"fs_max": 25600.0,

"fs": 20000.0,

"Delta_t": 10.0,

"window": "hann",

"t0": "2025-08-30T09:00:00Z",

"t1": "2025-08-30T09:00:10Z",

"fs_hat": 19999.8

},

"time": {

"clock_id": "clk_lab_01",

"sync_ref": "ptp",

"alpha": 0.9999997,

"beta": 2.3e-06,

"J": 8.0e-06,

"u_alpha": 1.0e-07,

"u_beta": 5.0e-07,

"ts_start": "2025-08-30T09:00:00Z",

"ts_end": "2025-08-30T09:00:10Z",

"tau_start": 12345.0,

"tau_end": 12355.0

},

"filt": {

"H_rev": "H-acc-v2",

"chain": [

{"type": "lp", "order": 6, "f_c": 9000.0, "ripple_db": 0.1, "stop_db": 80.0}

],

"f_c": 9000.0,

"BW": 9000.0,

"tau_g": 2.1e-04,

"AGC": null

},

"spec": {

"method": "welch",

"seg": 16,

"overlap": 0.5,

"window": "hann",

"fft_len": 262144,

"df": 0.076,

"S_xx_units": "power/Hz",

"features": {"band_power_[0_2000]": 1.23, "crest": 3.7, "THD": 0.012, "kurtosis": 3.05, "SNR_dB": 42.1},

"E_psd": 0.018

},

"env": {

"temp": 23.2,

"hum": 45.1,

"press": 101322.0,

"RefCond": {"temp": 23.0, "hum": 45.0, "press": 101325.0},

"correction_applied": true,

"u_c": 0.003,

"k": 2.0,

"U": 0.006

},

"path": {

"gamma_desc": "bench-straight-1m",

"L_gamma": 1.000,

"c_ref": 343.0,

"n_eff_model": "n_air_std_v1",

"T_arr_meas": 0.00295,

"T_arr_model": 0.00292,

"delta_form": 3.0e-05,

"toa_model_rev": "toa-air-1.0"

},

"quality": {

"q_score": 0.992,

"clip_rate": 0.0,

"missing_rate": 0.0,

"D": 0.12,

"alert": null

},

"io": {

"fmt": "parquet",

"compress": "zstd",

"partition_by": ["ts_date","sid"],

"bytes_raw": 160000000,

"bytes_stored": 52000000,

"bytes_compressed": 48000000,

"manifest_rev": "mf-sea-1.0"

}

}

XIII. Example 2 — FIELD Microphone (Long-duration)

{

"sensor": {

"sid": "SNSR-MIC-1145",

"model": "UMK-500",

"serial": "SN-9F2E7Q",

"class": "mic",

"cal_id": "CAL-2025-02",

"A_gain": 1.002,

"B_bias": 0.0,

"C_offset": 0.0,

"ENOB": 19.2,

"ADC_bits": 24,

"DR": 102.5

},

"acq": {

"fs_nom": 48000.0,

"fs_max": 96000.0,

"fs": 43200.0,

"Delta_t": 60.0,

"window": "hann",

"t0": "2025-08-30T00:00:00Z",

"t1": "2025-08-30T01:00:00Z",

"fs_hat": 43199.7

},

"time": {

"clock_id": "clk_field_02",

"sync_ref": "ntp",

"alpha": 1.0000011,

"beta": -1.1e-05,

"J": 2.2e-04,

"u_alpha": 2.0e-07,

"u_beta": 2.0e-06,

"ts_start": "2025-08-30T00:00:00Z",

"ts_end": "2025-08-30T01:00:00Z",

"tau_start": 501234.0,

"tau_end": 505,

"note": "tau_end trimmed per rotation"

},

"filt": {

"H_rev": "H-mic-v3",

"chain": [

{"type": "hp", "order": 4, "f_c": 20.0, "ripple_db": 0.1, "stop_db": 60.0},

{"type": "lp", "order": 6, "f_c": 19000.0, "ripple_db": 0.2, "stop_db": 80.0},

{"type": "notch", "f0": 50.0, "Q": 30.0}

],

"f_c": 19000.0,

"BW": 18980.0,

"tau_g": 3.6e-04,

"AGC": {"target_dBFS": -12.0, "attack_ms": 10.0, "release_ms": 200.0, "limit_db": -1.0}

},

"spec": {

"method": "welch",

"seg": 8,

"overlap": 0.5,

"window": "hann",

"fft_len": 131072,

"df": 0.33,

"S_xx_units": "power/Hz",

"features": {"band_power_[20_20000]": 0.87, "crest": 4.1, "THD": 0.018, "kurtosis": 3.20, "SNR_dB": 36.7},

"E_psd": 0.041

},

"env": {

"temp": 31.5,

"hum": 65.0,

"press": 100850.0,

"RefCond": {"temp": 23.0, "hum": 45.0, "press": 101325.0},

"correction_applied": true,

"u_c": 0.006,

"k": 2.0,

"U": 0.012

},

"path": {

"gamma_desc": "urban-canyon-seg42",

"L_gamma": 142.3,

"c_ref": 343.0,

"n_eff_model": "n_air_urban_v2",

"T_arr_meas": 0.420,

"T_arr_model": 0.417,

"delta_form": 3.4e-03,

"toa_model_rev": "toa-air-1.1"

},

"quality": {

"q_score": 0.956,

"clip_rate": 0.002,

"missing_rate": 0.004,

"D": 0.38,

"alert": {"kind": "drift", "ts": "2025-08-30T00:47:00Z", "level": "warn"}

},

"io": {

"fmt": "parquet",

"compress": "zstd",

"partition_by": ["ts_date","sid"],

"bytes_raw": 2.4e9,

"bytes_stored": 8.3e8,

"bytes_compressed": 7.1e8,

"manifest_rev": "mf-sea-1.0"

}

}

XIV. Example 3 — STREAM Manifest (Micro-batch Ingestion)

{

"sensor": {"sid": "SNSR-RF-2201", "model": "SDR-X9", "serial": "RF-77C1", "class": "rf", "cal_id": "CAL-2025-03", "ENOB": 12.8, "ADC_bits": 14, "DR": 70.0},

"acq": {"fs_nom": 2000000.0, "fs_max": 5000000.0, "fs": 2000000.0, "Delta_t": 1.0, "window": "hann", "fs_hat": 1999995.0},

"time": {"clock_id": "clk_sdr_01", "sync_ref": "gps", "alpha": 1.0, "beta": 0.0, "J": 4.0e-06, "ts_start": "2025-08-30T09:10:00Z", "ts_end": "2025-08-30T09:10:01Z", "tau_start": 201.0, "tau_end": 202.0},

"filt": {"H_rev": "H-rf-v1", "chain": [{"type": "bpf", "order": 8, "f1": 915000000.0, "f2": 916000000.0, "ripple_db": 0.5, "stop_db": 60.0}], "f_c": 915500000.0, "BW": 1000000.0, "tau_g": 1.0e-06},

"spec": {"method": "welch", "seg": 4, "overlap": 0.5, "window": "hann", "fft_len": 2097152, "df": 0.96, "S_xx_units": "power/Hz", "features": {"band_power_[915_916]": 2.78, "crest": 5.2, "SNR_dB": 28.4}},

"env": {"temp": 27.0, "hum": 40.0, "press": 101000.0, "RefCond": {"temp": 23.0, "hum": 45.0, "press": 101325.0}, "correction_applied": false},

"path": {"gamma_desc": "rf-los-200m", "L_gamma": 200.0, "c_ref": 299792458.0, "n_eff_model": "n_rf_free_v1", "T_arr_meas": 6.72e-07, "T_arr_model": 6.67e-07, "delta_form": 5.0e-09, "toa_model_rev": "toa-rf-1.0"},

"quality": {"q_score": 0.941, "clip_rate": 0.0, "missing_rate": 0.0, "D": 0.21},

"io": {"fmt": "tfrecord", "compress": "zstd", "partition_by": ["ts_minute","sid"], "bytes_raw": 8.0e7, "bytes_stored": 2.4e7, "bytes_compressed": 2.0e7, "manifest_rev": "mf-sea-1.0"}

}

XV. Minimal Required Manifest Set (Pre-commit Checklist)

• Identity: sensor.sid, sensor.model, sensor.serial, sensor.cal_id.
• Acquisition: acq.fs, acq.Delta_t, acq.window, acq.t0/acq.t1 or time.ts_start/time.ts_end.
• Timebase: time.clock_id, time.sync_ref, time.alpha, time.beta, time.J.
• Conditioning: filt.H_rev, filt.f_c, filt.BW, filt.tau_g.
• Spectrum: spec.method, spec.seg, spec.overlap, spec.window, spec.fft_len or spec.df.
• Environment: env.RefCond, env.correction_applied.
• Path: path.c_ref, path.n_eff_model, path.gamma_desc, path.T_arr_meas or path.T_arr_model, path.delta_form.
• Quality: quality.q_score, quality.clip_rate, quality.missing_rate.
• Lake I/O: io.fmt, io.compress, io.partition_by, io.manifest_rev.

XVI. Constraints and Conflict Checks (Execution Points)

• Naming conflicts
  Never confuse T_fil (tension) with T_trans (transmission coefficient); path/arrival-time fields must only use the c_ref, n_eff, T_arr conventions.
• Stability
  Any streaming channel must satisfy rho = lambda / mu < 1; backpressure policy must be recorded in the runtime configuration (see Core.Threads).
• Convention locking
  Spectrum and arrival-time reports must record H_rev, spec.window, spec.method, path.toa_model_rev, and env.RefCond in io.manifest_rev.

XVII. Cross-Volume Anchor References

• Arrival time & path: c_ref, gamma(ell), d ell, L_gamma, n_eff(x,t), T_arr (see Energy Filaments and Chapter 8 herein).
• Concurrency & ingestion: chan, cap, q_len, bp, W_q (see Core.Threads).
• Metrology & uncertainty: U = k * u_c, u(alpha), u(beta) (see Core.Metrology).

XVIII. Change Tracking and Versioning (Recommended Fields)