Asking the fleet what it is doing…
monad-knowledge Wi-Fi sensing lab · FIIT STU
Campaign session

01KVZ2FNS4DHTQ99R8MSCNFES5

finished 2026-06-25 09:41:24.388582+00:00 → 2026-06-25 09:57:37.124930+00:00 · 30 runs · supervisor: none

“Room scale ELIMINATED as the band-gap confound. On the small 135 m² floor (resplan-12419, 6 authored seats), walking-CV is high (0.07-0.12) but SATURATED — occupants near the 3 short links nearly max-out perturbation even at N=1 → weak grading (ρ 2.4=+0.29 [-0.27,0.73], 5.0=+0.29 [-0.28,0.73]), and STILL NO band gap (+0.00). Three-way: sim-small no gap, sim-large no gap (+0.01), real +0.43. Scenario ✗, body_loss ✗, room-scale ✗ all eliminated → the sim→real 5 GHz advantage is the ray-tracer's FREQUENCY PHYSICS (identical per-band num_subcarriers/bandwidth/body-scatterer-response/max_depth). The deeper exp-csi-calibration target is now pinned.”

Archive snapshot, as of 22 h ago — the run corpus is rebuilt once a day, so this page is not a live reading. The fleet panel is the live one; it refreshes every 30 s.

Success criteria

CriterionResolved
30/30 coupled runs gate-passed on the small floor (resplan-12419, 135 m², 6 seats authored in bedroom-0 near the Rx); all per-run scalars within PHYSICAL_BOUNDS; runs explicitly attached (MCP). yes
Walking CV clears the floor decisively (0.07-0.12) — small room + short links strongly perturb the channel — but is SATURATED (weakly graded), the opposite failure mode to the large floor's starvation. yes
Per-band ρ with bootstrap CI + multi_seed_ci (0% inert, genuine 3 seeds): 2.4=+0.29 [-0.27,0.73], 5.0=+0.29 [-0.28,0.73] — both bands identical. yes
Three-way comparison (sim-small / sim-large / WiMANS real) shows NEITHER sim scale reproduces the real 5 GHz>2.4 GHz gap → room scale eliminated; defect localised to the RT frequency model. yes

Synthesis

c-csi-roomscale-calibration — session 01KVZ2FNS4DHTQ99R8MSCNFES5 (2026-06-25)

Verdict: room scale is ELIMINATED as the sim→real band-gap confound. By exclusion, the 5 GHz advantage the simulator lacks is the ray-tracer's frequency physics.

This closes the diagnostic chain opened by c-csi-band-calibration: three candidate causes for "the sim doesn't reproduce the WiMANS 5 GHz > 2.4 GHz occupancy-discriminability gap" — scenario, body_loss, room scale — are now all tested and rejected.

Setup

The smallest ResPlan floor carrying a csi-link-* experiment, resplan-12419 (135 m²) vs the 350 m² resplan-12439, to test whether occupants nearer the links recover the band gap. The floor had no seats, so 6 seats were authored in bedroom-0 (near csi-rp12419-rx1; provenance computed) and a smoke confirmed the walker reaches them (walking-CV 0.96). 30 coupled runs N∈{1..5}×{2.4,5.0 GHz}×seeds{0,1,2}, refit scenario duration_units=1200, all gate-passed, explicitly attached.

What we found

band sim small (12419) sim large (12439) WiMANS real
2.4 GHz +0.29 [−0.27, +0.73] +0.74 +0.28
5.0 GHz +0.29 [−0.28, +0.73] +0.75 +0.71
band gap +0.00 +0.01 +0.43
  • Opposite failure mode, same conclusion. The small floor does not starve the signal — walking-CV is high (0.07–0.12) because occupants near the 3 short links strongly perturb the channel. But it saturates: the perturbation nearly maxes out by N=1, so the CV is only weakly graded (ρ=+0.29, CIs spanning 0). The large floor under-couples (better dynamic range, ρ=+0.74); the small floor over-couples (saturated, ρ=+0.29). Different pathologies — but neither shows any band gap (both +0.00/+0.01).
  • Only real data has the 5 GHz advantage (+0.43). 3 genuine seeds (0% inert), honest CIs.

Conclusion — the defect is pinned

Scenario (fixed at dur=1200), body_loss (inert 5→20 dB), and room scale (this run, both floors equal-band) are all eliminated. The simulator treats 2.4 and 5 GHz identically for occupancy variance regardless of floor or scenario, because the ray-tracer uses the same per-band parametersnum_subcarriers/bandwidth_hz, the homogeneous body-scatterer response, and max_depth=3 for both carriers. It physically cannot manufacture the shorter-wavelength sensitivity real 5 GHz hardware shows. The exp-csi-calibration refit target is therefore the frequency-dependent RT model, not scenario/coupling/geometry: per-band subcarrier count + bandwidth, a frequency-dependent body permittivity/scattering response, and possibly higher max_depth at 5 GHz.

In-silico calibration probe; rank + band-gap only. 6 seats added to resplan-12419 (persistent DB change). Figure + metrics under the session prefix and _attachments/wimans-zeroshot/fig_roomscale_3way.png. 30 runs attached.

Attached runs

Run Gate Purpose Replay
3ZJKX3KK n1_24g_s1 replay
K9AC4SW5 n1_24g_s0 replay
MD6TBZ4A n1_24g_s2 replay
AJ0BRNKN n1_50g_s0 replay
J849FH6P n1_50g_s2 replay
9BVQTP1F n1_50g_s1 replay
32BTR3RX n2_24g_s0 replay
BV1M9MWS n2_24g_s1 replay
QN0VBAYA n2_24g_s2 replay
1JQ5ZA04 n2_50g_s0 replay
SA5A6CGH n2_50g_s1 replay
BWC7PGK1 n2_50g_s2 replay
GN9P6T6P n3_24g_s0 replay
R8Z185F7 n3_24g_s1 replay
KEB94EN0 n3_50g_s0 replay
ZGWZHZ7W n3_24g_s2 replay
KWJFRHG1 n3_50g_s1 replay
Q8J667MT n3_50g_s2 replay
837SHAZG n4_24g_s0 replay
CWR0C015 n4_24g_s1 replay
T1VACD0W n4_50g_s0 replay
PC1KS135 n4_24g_s2 replay
FDMFNG64 n4_50g_s1 replay
Y5JWF2CX n4_50g_s2 replay
EY3BKWPX n5_24g_s0 replay
RGB5PGDK n5_24g_s1 replay
JRE22HYZ n5_24g_s2 replay
46EAPM5X n5_50g_s0 replay
C8YD8C6Y n5_50g_s1 replay
KEQR8DJE n5_50g_s2 replay