Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE

Enhancing QKD security with weak measurements

Jacob Farinholt, James E. Troupe

Published Oct 24, 2016• 1 citation

Abstract

Publisher’s Note: This paper, originally published on 10/24/2016, was replaced with a corrected/revised version on 11/8/2016. If you downloaded the original PDF but are unable to access the revision, please contact SPIE Digital Library Customer Service for assistance. In the late 1980s, Aharonov and colleagues developed the notion of a weak measurement of a quantum observable that does not appreciably disturb the system.1, 2 The measurement results are conditioned on both the pre-selected and post-selected state of the quantum system. While any one measurement reveals very little information, by making the same measurement on a large ensemble of identically prepared pre- and post-selected (PPS) states and averaging the results, one may obtain what is known as the weak value of the observable with respect to that PPS ensemble. Recently, weak measurements have been proposed as a method of assessing the security of QKD in the well-known BB84 protocol.3 This weak value augmented QKD protocol (WV-QKD) works by additionally requiring the receiver, Bob, to make a weak measurement of a particular observable prior to his strong measurement. For the subset of measurement results in which Alice and Bob's measurement bases do not agree, the weak measurement results can be used to detect any attempt by an eavesdropper, Eve, to correlate her measurement results with Bob's. Furthermore, the well-known detector blinding attacks, which are known to perfectly correlate Eve's results with Bob's without being caught by conventional BB84 implementations, actually make the eavesdropper more visible in the new WV-QKD protocol. In this paper, we will introduce the WV-QKD protocol and discuss its generalization to the 6-state single qubit protocol. We will discuss the types of weak measurements that are optimal for this protocol, and compare the predicted performance of the 6- and 4-state WV-QKD protocols.

DOI: 10.1117/12.2241059

Related papers

Quantum Cryptography with Weak Measurements

James E. Troupe, Jacob Farinholt

In this article we present a new prepare and measure quantum key distribution protocol that decouples the necessary quantum channel error estimation from its dependency on sifting, or otherwise post-selecting, the detection outcomes. Rather than estimating Eve's coupling to the quantum channel from the statistics of the sifted key, we infer this information from weak measurements made equally on all of the received photons immediately prior to post-selection by the photon detectors. We prove that the accuracy of the weak measurement parameter estimation is robust to reasonable device imperfections, even in an adversarial environment, and hence the asymptotic security of this protocol can be inferred from the security analysis of BB84. In addition to eliminating detector basis-dependent attacks, such as detector blinding, we demonstrate that this new prepared and measure QKD protocol is immune to a very powerful class of measurement-side device attacks that also allow an adversary control of the weak measurement outcomes given two modest requirements placed on the measurement-side devices. Finally, we compare the asymptotically achievable secure key rate of a decoy state version of the weak measurement protocol and show it is essentially equal to that of BB84 with decoy states and significantly higher than MDI-QKD for realistic system parameters.

arXiv:1702.04836Feb 16, 20170 citations

Weak values in continuous weak measurement of qubits

Lupei Qin, Liang, Pengfei, Xin-Qi Li

For continuous weak measurement of qubits, we obtain exact expressions for weak values (WVs) from the post-selection restricted average of measurement outputs, by using both the quantumtrajectory- equation (QTE) and quantum Bayesian approach. The former is applicable to short-time weak measurement, while the latter can relax the measurement strength to finite. We find that even in the "very" weak limit the result can be essentially different from the one originally proposed by Aharonov, Albert and Vaidman (AAV), in a sense that our result incorporates non-perturbative correction which could be important when the AAV's WV is large. Within the Bayesian framework, we obtain also elegant expressions for finite measurement strength and find that the amplifier's noise in quantum measurement has no effect on the WVs. In particular, we obtain very useful result for homodyne measurement in circuit-QED system, which allows for measuring the real and imaginary parts of the AAV's WV by simply tuning the phase of the local oscillator. This advantage can be exploited as efficient state-tomography technique.

arXiv:1502.02396Feb 9, 20150 citations

Security and performance analyses of quantum-key-distribution systems

Chi‐Hang Fred Fung

Quantum cryptography, or, more precisely, quantum key distribution (QKD) allows two distant parties, Alice and Bob, to expand a previously shared secret key by sending quantum states through a quantum channel. The goal here is for Alice and Bob to communicate securely in the presence of an eavesdropper, Eve. The most well-known QKD protocol is the Bennett-Brassard 1984 (BB84) protocol, which has been proved unconditionally secure against any attacks allowed by quantum mechanics. The main advantage of QKD over conventional cryptography is its absolute security. Thus, it is imperative to evaluate the security and performance of QKD protocols and realistic QKD systems. In this thesis, we study the security and performance aspects of QKD in four parts. In the first part, we study the Scarani-Acin-Ribordy-Gisin 2004 (SARG04) protocol. This protocol differs from the BB84 protocol only in the classical post-processing step and uses the same four quantum states as the BB84 protocol. We compared these two protocols and showed that in typical situations, the classical post-processing software for BB84 should be used in order to have a higher performance. In the second part, we prove the security of a three-state protocol. This protocol is similar to the BB84 protocol in that they share the same three qubit states. In the third part, we consider a practical Trojan horse attack on practical QKD systems implementing the BB84 protocol. The attack works by taking advantage of the finite response time of practical phase modulators to remap the states sent out by the legitimate user, Alice. We showed that if Alice and Bob are unaware of our attack, the final key shared between them can be compromised in some situations. In the fourth part, we consider another type of Trojan horse attack, which is based on mismatch between detection efficiencies of detectors. Practical detectors in QKD systems usually do not have identical efficiencies and an eavesdropper Eve may be able to exploit this efficiency mismatch to steal information about the final secret key. We analyze the security of the attack and show that it is possible for Eve to acquire some information on the final secret key if Alice and Bob are unaware of this attack.

Jan 1, 20080 citations

Quantum key distribution with prepare-and-measure Bell test

Yong-Gang Tan

The prepare-and-measure quantum key distribution (QKD) has the merits of fast speed, high key generation rate, and easy implementation. However, the detector side channel attacks greatly undermine the security of the key bits. The eavesdropper, Eve, exploits the flaws of the detectors to obtain illegal information without violating quantum principles. It means that she can intervene in the communication without being detected. A prepare-and-measure Bell test protocol will be proposed. By randomly carrying out Bell test at the side of the information receiver, Bob, Eve's illegal information gain within the detector side channel attack can be well bounded. This protocol does not require any improvement on the detectors used in available prepare-and-measure QKD. Though we only illustrate its application in the BB84 protocol, it is applicable for any prepare-and-measure QKD.

Scientific ReportsOct 13, 20166 citations

A Contextuality Based Quantum Key Distribution Protocol

James E. Troupe, Jacob Farinholt

In this article we present a new prepare and measure quantum key distribution protocol that uses an experimentally accessible measure of single qubit contextuality to warranty the security of the quantum channel. The definition of contextuality used is that due to Spekkens in which any underlying hidden variable model of the physical system is noncontextual if the probability distribution of the model's hidden variables is independent of the preparation and measurement context. Under this more general definition of noncontextuality the measurement outcomes of a single qubit can be shown to require a contextual model in order to reproduce the results of some non-projective, positive operator valued measurements (POVMs). The proposed protocol utilizes a particular set of POVMs to exhibit the degree of contextuality of the qubits exiting the QKD system's quantum channel. The contextuality measure is defined in terms of the weak values of a set of projective measurements that are implemented as POVMs using weak measurements. The resulting QKD protocol should have a key rate that is equal to or greater than BB84 since none of the sifted key is used to test for the presence of an eavesdropper. Additionally, the new protocol is shown to be immune to detector based attacks.

arXiv:1512.02256Dec 7, 20154 citations

Another Look at Symmetric Incoherent Optimal Eavesdropping against BB84

Arpita Maitra, Goutam Paul

Abstract. The BB84 protocol is used by Alice (the sender) and Bob (the receiver) to settle on a secret classical bit-string by communicating qubits over an insecure quantum channel where Eve (the Eavesdropper) can have access. In this paper, we revisit a well known eavesdropping technique against BB84. We claim that there exist certain gaps in understanding the existing eavesdropping strategyin terms of cryptanalytic view and we try to bridge those gaps in this paper. First we refer to the result where it is shown that in the six-state variant of the BB84 protocol (Bruß, Phys. Rev. Lett., 1998), the mutual information between Alice (the sender) and Eve (the eavesdropper) is higher when two-bit probe is used compared to the one-bit probe and hencethetwo-bitprobeprovidesastrongereavesdroppingstrategy.However, from cryptanalytic point of view, we show that Eve has the same success probability in guessing the bit transmitted by Alice in both the cases of the two-bit and the one-bit probe. Thus, we point out that having higher mutual information may not directly lead to obtaining higher probability in guessing the key bit. It is also explained in the work of Bruß that the six-state variant of the BB84 protocol is more secure than the traditional four-state BB84. We look into this point in more detail and identify that this advantage is only achieved at the expense of communicating more qubits in the sixstate protocol. In fact, we present different scenarios, where given the same number of qubits communicated, the security comparison of the four and six-state protocols is evaluated carefully.

Jan 1, 20120 citations