The Signal
Quantum Dots Outpace Laser Pulses in an MDI-QKD Model
A theoretical comparison finds that quantum-dot sources can outperform optimized weak coherent laser pulses in modeled MDI-QKD key rates, assuming perfect photon indistinguishability for that comparison. A separate analysis finds that imper
Quantum.world Team · Written with AI · fact-checked ·
A 0.3-photon setting beats an optimized laser baseline
At a modeled mean photon number of 0.3, quantum-dot sources produce higher secret-key rates than weak coherent laser pulses across the transmission distances the researchers analyze—even though the laser intensity is optimized separately at every loss level. This comparison assumes perfect photon indistinguishability. The theoretical study is not a demonstration of a working quantum-key-distribution link.
The researchers compare two measurement-device-independent quantum key distribution (MDI-QKD) systems that differ in their photon sources. In each, Alice and Bob send encoded photons to a central, untrusted measurement unit, which performs a Bell-state measurement. The protocol’s security does not depend on trusting that unit’s detectors, but its performance depends on the photons arriving there and interfering correctly.
“Notably, QD sources using decoy states outperform the optimised WCP protocol across all distances already for μ=0.3, despite the latter being optimised at every transmission loss.”
Janka Memmen and co-authors, the paper’s authors
Sub-Poissonian emissions change the trade-off
Weak coherent pulses follow Poissonian photon-number statistics: raising a laser pulse’s mean photon number increases the chance of getting a useful single photon, but also increases multi-photon emissions. Those extra photons constrain the secure key rate. Quantum dots can suppress multi-photon emissions, allowing the model’s QD source to operate at a higher mean photon number without the same penalty.
The comparison uses decoy-state analysis for both types of source and sets the QD antibunching parameter, g(2)(0), to 0.02. The researchers test QD mean photon numbers of 0.05, 0.3 and 0.8 against a weak-pulse configuration whose mean is numerically optimized at each transmission loss. With these assumptions, the QD source’s threshold for beating the optimized laser case is about 0.26 in the low-loss regime, falling to about 0.2 as loss rises.
Indistinguishability narrows the modeled reach
Source statistics alone do not settle the comparison. MDI-QKD relies on Hong–Ou–Mandel interference between photons from Alice and Bob; if they are distinguishable, the measurement degrades and the photons may reveal information about the encoded states. In a separate analysis, the paper incorporates imperfect indistinguishability into its security analysis, using the interference visibility as an experimentally measurable indicator.
For a modeled single-photon indistinguishability of 0.9999, the calculated Hong–Ou–Mandel visibility is 0.983297 for the QD source and 0.499015 for the weak coherent pulses. At that setting, the study finds positive key rates only up to roughly 40 kilometres for QD sources and 25 kilometres for weak pulses. Those distance figures come from a separate comparison in which both source types use a fixed mean photon number of 0.3; the laser intensity is not optimized by distance in that analysis.
The model stops short of a network trial
The paper assumes an asymptotic regime, effectively an infinite number of transmitted signals, so it does not include the statistical costs of a finite-key implementation. Its two links have equal losses, and Alice and Bob are assumed to use sources with identical characteristics. The authors also hold g(2)(0) at 0.02; the modeled advantage depends on maintaining high brightness and indistinguishability together.
For a deployed system, finite data, unequal links and real source performance could change the key-rate comparison. The study establishes a modeled advantage under specified conditions, not a measured MDI-QKD range. Finite-key effects and asymmetric links remain outside its analysis.