ExpGordonB92
Tier B: Gordon et al. 2004, the B92 QBER curve to 11.85 km at 100 MHz
5/5 passed in 1ms
| Test | What it does | Result |
|---|---|---|
test_gordon_curve | One visibility per window fitted at zero length, with their 180 counts/s and 2.2 dB/km, reproduces the QBER from 0.4% to 15.6% out to 9.96 km, worst miss 0.73 points at 5 ns and 0.54 at 9 ns. | ✅ pass |
test_gordon_leak | The fitted leakage is 0.256% per beam splitter, a 25.9 dB extinction, and their background carries 36.1% of the 0.4% zero-length QBER. | ✅ pass |
test_gordon_tail | Declared miss: 11.07 and 11.85 km read 22.24% and 27.14% against their 24.3% and 31.8%, their counts falling 10.7% from 9.96 to 11.07 km where 2.2 dB/km asks 43.0%. | ✅ pass |
test_gordon_takeover | Background-free the model holds at 0.256% at every length; halving their 180 counts/s puts 9.96 km at 9.41% and doubling it at 23.85%, both excluded by their 15.6%. | ✅ pass |
test_gordon_window | The two 100 MHz windows agree at zero length while the 1 GHz rows split 3.7 points, more than the whole 100 MHz error, a temporal term b92_detect does not model. | ✅ pass |