Limitations on Practical Quantum Cryptography

Abstract

The use of quantum bits (qubits) in cryptography holds the promise of secure cryptographic quantum key distribution schemes. Unfortunately, the implemented schemes can be totally insecure. We provide a thorough investigation of security issues for practical quantum key distribution, taking into account channel losses, a realistic detection process, and modifications of the "qubits" sent from the sender to the receiver. We first show that even quantum key distribution with perfect qubits cannot be achieved over long distances when fixed channel losses and fixed dark count errors are taken into account. Then we show that existing experimental schemes (based on "weak-pulse") are usually totally insecure. Finally we show that parametric downconversion offers enhanced performance compared to its weak coherent pulse counterpart.Pacs: 03.67. Dd, 42.50.Dv, 03.65.Bz, 89.80.+h Quantum information theory suggests the possibility of accomplishing tasks which are beyond the capability of classical computer science, such as information-secure cryptographic key distribution [1,2]. The lack of security proofs for standard (secret-and public-) key distribution schemes, and the insecurity of the strongest classical schemes against "quantum attacks" [3], emphasizes the need for information-secure key distribution. Whereas the security of idealized quantum key distribution (qkd) schemes has been investigated against very sophisticated collective and joint attacks (e.g., [4,5]), the experimental qkd schemes have been proven secure against the simple individual attack only recently [6] (via the application of ideas presented here).In the four-state scheme [1], usually referred to as Bennett-Brassard-84 (BB84), the sender (Alice) and the receiver (Bob) use two conjugate bases (say, the rectilinear basis, +, and the diagonal basis, ×) for the polarization of single photons. In basis + they use the two orthogonal basis states |0 + and |1 + to represent "0" and "1" respectively. In basis × they use the two orthogonal basis states2)[|0 + − |1 + ] to represent "0" and "1". The basis is revealed later on via an unjammable and insecure classical channel. The signals where Bob used the same basis as Alice form the sifted key on which Bob can decode the bit value. The remaining signals are being discarded. Finally, they test a few bits to estimate the error-rate, and if the test passes (the tested error-rate is less than some pre-agreed threshold), they use errorcorrection and privacy amplification to obtain a potentially secure final key [7,8].The security of that scheme, which assumes a source of perfect qubits as well as losses and errors which are bounded by some small threshold, has been investigated in various works. Very simple attacks already render realistic qkd impossible, as we show here.The experiments are usually based on weak coherent pulses (wcp) as signal states with a low probability of containing more than one photon [7,9]. Initial security analysis of such weak-pulse schemes were done [7,10], and evidence of some potentia...

On‐chip single photon sources using planar photonic crystals and single quantum dots

“…The easiest and most straightforward way is to attenuate pulsed lasers [5], though in this case the production of single photons is probabilistic. Therefore, there may be no photons, several photons, or many photons, since the photon number generated is subject to Poissonian statistics; this can be problematic for a number of applications such as quantum cryptography because of possible photon number-splitting attacks ("eavesdropping") [6,7]. Another more practical approach for obtaining single photons is to exploit spontaneous parametric down conversion in nonlinear crystals [8], where higher-frequency pump photons incident on a nonlinear crystal are occasionally split into a pair of lower-frequency photons; one of these photons (the "heralding" photon) is used to herald the arrival of the second photon ("heralded" photon), and thus the second photon can be well isolated and manipulated.…”

Using quantum key distribution for cryptographic purposes: A survey

“…Instead, weak coherent laser pulses (WCP) with poissonian photon statistics are used. This has security implications since pulses of light containing more than one photon can be exploited by the eavesdropper to learn potentially all the information about the encoded bit, using the so-called photon-number-splitting (PNS) attack [43]. Considering this attack, it remains possible to prove the security of BB84 performed with WCP, but one must decrease the mean number of photons in the pulse (and hence the rate) as the distance increases [43].…”

“…This has security implications since pulses of light containing more than one photon can be exploited by the eavesdropper to learn potentially all the information about the encoded bit, using the so-called photon-number-splitting (PNS) attack [43]. Considering this attack, it remains possible to prove the security of BB84 performed with WCP, but one must decrease the mean number of photons in the pulse (and hence the rate) as the distance increases [43]. In 2003, a more radical response to the PNS attack has been proposed, that consists in modifying the QKD protocol with the adjunction 5.2 Device-independent security: fundamental quantum mechanics as a tool against side-channels of decoy-states to actively test the influence of the eavesdropper on the photon statistics.…”

Practical photon number detection with electric field-modulated silicon avalanche photodiodes

“…Similarly, in linear optics quantum computing, the successful operation of a quantum logic gate is often heralded by the detection of a photon number state 6 . Low-noise photon number detection would also be useful for characterizing non-classical light sources 7,8 and security analysis in quantum cryptography 9 . In combination with parametric down-conversion sources, photon number detection can also be used for the generation and conditioning of photonic Fock states 10 , as well as more complex quantum light states 11 .…”