ExpDiamantiDps
Tier B: Diamanti et al. 2006, DPS measured over 100 km
7/7 passed in 351ms
| Test | What it does | Result |
|---|---|---|
test_dps_deadtime | Their Eq. (5) dead-time factor exp(-numuT*t_d/2) agrees with the sequential scan in run_clicks to 4% at 10 km and 80 ns. | ✅ pass |
test_dps_montecarlo | At 10 km, 4e6 pulses of run_clicks reproduce the closed form to 3% on sift and 0.4 points on QBER, about three standard errors. | ✅ pass |
test_dps_near_deficit | Declared miss at 10 km: QBER 2.16% against 2.2% but sift 2.55 Mbit/s against 2, needing ~1.1 dB of unpublished loss that moves the QBER to 2.35%. | ✅ pass |
test_dps_qber | The 100 km QBER is 2.691% against the 2.7% jitter-free residual, splitting into 1.726% dark against their 1.7% and 0.965% interferometric against their 1%. | ✅ pass |
test_dps_secure | The Waks-Takesue-Yamamoto rate at 100 km is 168.2 bit/s against their 166, 1.3% high. | ✅ pass |
test_dps_three_ways | QBER, dark share and rate invert to alpha = 0.20763, 0.20673 and 0.20799 dB/km, a 0.61% spread inside [0.20, 0.22]; only the rate leg is independent of the error decomposition. | ✅ pass |
test_dps_window_reading | Reading "eta=0.4%" as post-window is excluded: it needs alpha = 0.2404-0.2417 dB/km, and gives 1.79% QBER against 2.7% and 359 bit/s against 166. | ✅ pass |