Table 1 Comparison of notable parameters from prepare-and-measure QKD experiments conducted in the last decade with similar physical channels, as indicated by the column showing loss and length

From: Practical continuous-variable quantum key distribution with composable security

 

Protocol implementation keywords

Loss/ Length [dB/km]

μ [PNU]

ξu [PNU]

ξQ [%]

B [MHz]

N( × 109) [symbols]

SKF [bits/symbol]

Huang et al. (ref. 33)

CV, Gaussian, Homodyne via RLO

5.0/25.0

2.00

0.025*

 

100

0.02

0.0010*

Xu et al. (ref. 40)

DV, phase coding, Avalanche diode

4.5/20.0

0.37

 

2.73

5

7.84

0.0001*

Islam et al. (ref. 41)

DV, time-bin coding, Superconducting

4.0/20.0

0.45

 

5.49

2500

62.5

0.0105

Wang et al. (ref. 20)

CV, Gaussian, Heterodyne via RLO

5.0/25.0

1.62

0.011

 

100

0.01

0.0185

Zhang et al. (ref. 21)

CV, Gaussian, Homodyne via TLO

4.4/27.3

7.19

0.002

 

5

100

0.0560

Current work

CV, Gaussian, Heterodyne via RLO

4.6/20.3

1.45

0.005

 

100

1

0.0471

  1. Values with a superscript * may be somewhat inaccurate as they were inferred from a graph. μ, mean photon number of the quantum state alphabet; ξu, untrusted noise (referred to channel output); ξQ, quantum bit error rate; B, repetition rate in pulsed or quantum data bandwidth in CW implementations; N, number of transmitted quantum data symbols or pulses in the experiment; secret key fraction (SKF), secret key length in bits divided by N. It is possible to parametrize ξu and ξQ by the same quantity, namely the mean number of noise photons from the channel, in CV and DV systems, respectively36. Also, assuming symmetry between the quadratures, 1 photon number unit (PNU) corresponds to a variance of 2 shot noise units (SNU).
  2. R/TLO: real/transmitted LO.