AttacksThreshold
Blinding, efficiency mismatch and dead-time blanking against click receivers
12/12 passed in 1ms
| Test | What it does | Result |
|---|---|---|
test_blank_deadtime | The dead time impairments.saturate charges as lost clicks the blanking attack charges as Eve's information, leaving the QBER unmoved at 16.52 and 0.37 photons. | ✅ pass |
test_blank_hidden | blank_hidden passes Weier's 200 ns lead against a 5 ns window and a 2 microsecond or 400 ns dead time, and refuses a pulse inside the window or older than the dead time. | ✅ pass |
test_blank_leak | Published anchor, Weier et al., New J. Phys. 13, 073024 (2011) Table I: the model sits above all seven measured I_EB and reproduces the top point 0.908 to 0.023, I_EB = 1 - h(QBER_BE) holds to 6e-4, and the doubled diagonal reading lands on Fig. 2b. | ✅ pass |
test_blank_model | The blanking model has Weier's structure: the parallel detector's blind probability composes from two diagonal halves, zero intensity blinds and leaks nothing, and the error rate is independent of signal intensity to 2e-3. | ✅ pass |
test_blinding_anchor | Published anchor, Lydersen et al., Nat. Photonics 4, 686 (2010), arXiv:1008.4593: the measured Clavis2 thresholds give P_always/P_never = 1.4405 < 2, their Eq. (1), and both detectors blind at 397 and 765 microwatt, below the 1.08 mW control power. | ✅ pass |
test_blinding_break | Under detector control the observed QBER is Eve's choice while she holds one bit per sifted bit, and a 1% gain costs a 2% trigger share, 1% passive. | ✅ pass |
test_blinding_control | blind_break refuses a detector set violating Eq. (1) and a gain Eve cannot cover, and the trigger share halves when Bob's basis choice is passive. | ✅ pass |
test_mismatch_rate | Fung, Tamaki, Qi, Lo and Ma, QIC 9, 131 (2009), arXiv:0802.3788 Eqs. (32) to (34): the two forms join at matched detectors and otherwise differ by (1 - share) h(e_bit), and at Zhao's 7.13 mismatch and 5.68% QBER 1 - 2h(e) overstates the rate 4.07 times. | ✅ pass |
test_shift_anchor | Published anchor, Qi et al. Eq. (4): their measured mismatch of 2 bounds the rate at h(2/3) = 0.9183 against a naive 1, symmetric under r to 1/r, with no induced error. | ✅ pass |
test_shift_balance | Published anchor, Zhao, Fung, Qi, Chen and Lo, PRA 78, 042333 (2008) Table I: shift_balance derives the equalising mixture 23.031% and 3479.07 detections per detector against their 23.0% and 3479, and that mixture gives the overall QBER 5.681% against 5.68%. | ✅ pass |
test_shift_measured | Zhao's two measured shifts give mismatches 7.133 and 3.297, rate ceilings 0.5378 and 0.7827 under Qi Eq. (4), and their per-shift QBERs 6.135% and 5.365%. | ✅ pass |
test_shift_qber | Qi, Fung, Lo and Ma, QIC 7, 73 (2007), arXiv:quant-ph/0512080 Eq. (2): the faked-state QBER rises from 0 at r = 0 to 1/2 at r = 1 and crosses intercept-resend's 1/4 at their stated r = 0.2. | ✅ pass |