Bb84Wcp
Tier A: GLLP + decoy BB84 with weak coherent pulses vs the GYS parameter set
9/9 passed in 88ms
| Test | What it does | Result |
|---|---|---|
test_bb84_guards | bb84_rate raises a ValueError naming the slot for an error rate outside [0, 1/2], a gain outside [0, 1], q_sift outside (0, 1], NaN anywhere, and q1 > q_mu. | ✅ pass |
test_gys_linear_scaling | The GYS decoy BB84 rate's log-log slope against eta over 25-75 km is 1.0 within 0.05. | ✅ pass |
test_gys_nondecoy_deficit | GLLP with no decoys reaches 40 km optimised and no key at GYS's mu = 0.1, a different theorem from the 49.0 km Brassard-Lutkenhaus-Mor-Sanders condition of test_gys_pns_reading. | ✅ pass |
test_gys_optimal_mu | The 25 km optimum intensity is within 0.02 of Ma et al.'s mu_optimal = 0.48 at f(e) = 1.22. | ✅ pass |
test_gys_rate_scale | At mu = 0.48 the GYS rate is 2.55e-3 bit/pulse at 0 km and 1.72e-5 at 100 km. | ✅ pass |
test_gys_secure_distance | Ma, Qi, Zhao & Lo 2005's "over 140 km of Telecom fibers" on the GYS parameters: the rate is positive at 140 km and zero at 145 km. | ✅ pass |
test_rate_falls_distance | The rate falls monotonically with fibre length and is exactly 0.0 from 150 to 400 km. | ✅ pass |
test_recon_floor | bb84_rate and cow_rate refuse an inefficiency f_ec < 1, which beats Shannon, and the rate falls from f_ec = 1 to 1.22. | ✅ pass |
test_zero_error_closed_form | At zero error bb84_rate is q*Q1 exactly. | ✅ pass |