This thesis deals with position-based quantum cryptography (PBQC), a type of quantum cryptography. More concretely, we focus on the design and study of quantum position verification (QPV) protocols, which act as a building block for PBQC. We prioritise the development of QPV protocols that are loss-tolerant, error-robust, and feasible with current or near-future technology, concentrating on photon-based protocols and linear-optical equipment. We also theoretically study different attack settings on QPV in the context of state discrimination, dealing with the interplay between non-locality and interaction. We introduce a new QPV protocol based on the SWAP test and Hong-Ou-Mandel interference and study it in detail regarding attacks and practical feasibility. We further provide a solution to a major practical issue of QPV -- signal loss. A minor modification of the standard structure of QPV, namely introducing a small time delay and a commitment to play from the honest prover, allows us to provably make the transmission loss between the verifiers and the honest party irrelevant for security for a class of QPV protocols. Moreover, we generalise and study QPV protocols in the continuous-variable setting. Exploring theoretical aspects of QPV attacks, we focus on the difference between quantum and classical communication. We show new bounds on existing protocols, construct protocols that can be attacked with quantum communication but not locally, show a separation between quantum and classical communication attacks on QPV, and derive a necessary condition on the attack error based on the structure of the QPV input ensemble.
This thesis deals with position-based quantum cryptography (PBQC), a type of quantum cryptography. More concretely, we focus on the design and study of quantum position verification (QPV) protocols, which act as a building block for PBQC. We prioritise the development of QPV protocols that are loss-tolerant, error-robust, and feasible with current or near-future technology, concentrating on photon-based protocols and linear-optical equipment. We also theoretically study different attack settings on QPV in the context of state discrimination, dealing with the interplay between non-locality and interaction. We introduce a new QPV protocol based on the SWAP test and Hong-Ou-Mandel interference and study it in detail regarding attacks and practical feasibility. We further provide a solution to a major practical issue of QPV -- signal loss. A minor modification of the standard structure of QPV, namely introducing a small time delay and a commitment to play from the honest prover, allows us to provably make the transmission loss between the verifiers and the honest party irrelevant for security for a class of QPV protocols. Moreover, we generalise and study QPV protocols in the continuous-variable setting. Exploring theoretical aspects of QPV attacks, we focus on the difference between quantum and classical communication. We show new bounds on existing protocols, construct protocols that can be attacked with quantum communication but not locally, show a separation between quantum and classical communication attacks on QPV, and derive a necessary condition on the attack error based on the structure of the QPV input ensemble.
In this thesis we explore position-based quantum cryptography zooming in on the task of position verification. In position verification, the idea is to use an individual's geographical location as a cryptographic credential. In practice, such protocols can authenticate that a message originated from a specific location or ensure that messages can only be read at a certain location. In a position-verification protocol, the limitations imposed by the speed of light, as described by special relativity, are used to verify that a party is at their claimed location. It has been shown that any QPV protocol can be broken by attackers that use an amount of entanglement exponential in the input size. However, from a practical point of view, not all is lost. We can still hope for QPV protocols that are secure in practice. This raises the problem of designing protocols that are secure in a practical setting. An important problem that arises is that of signal loss. Signal loss can be detrimental as it allows attackers to only answer on a subset of rounds. We investigate several questions related to the practical implementation of such protocols. We propose a new experimentally feasible protocol, which is fully loss-tolerant against attackers restricted to classical communication. Furthermore, we consider the role of allowing quantum communication with no pre-shared entanglement for attackers and show that this setting is strictly stronger than the classical one, indicating that allowing for quantum communication may break security. Conversely, we then show that any protocol secure against classical communication can be transformed into a protocol secure against quantum communication. Finally, we invert the picture, and consider the task of non-local quantum computation (NLQC), which corresponds to the operations of the attackers in a QPV protocol. We connect NLQC to the wider context of information-theoretic cryptography by relating it to a number of other cryptographic primitives.
We study the role of quantum communication in attacks on quantum position verification. In this work, we construct the first known example of a QPV protocol that is provably secure against unentangled attackers restricted to classical communication, but can be perfectly attacked by local operations and a single round of simultaneous quantum communication indicating that allowing for quantum communication may break security. We also show that any protocol secure against classical communication can be transformed into a protocol secure against quantum communication. We further show, using arguments based on the monogamy of entanglement, that the task of Bell state discrimination cannot be done locally with a single round of quantum communication, not even probabilistically (when we allow attackers to say loss sometimes), making this the first fully loss-tolerant QPV task secure against quantum communication attacks. Finally, we observe that any multi-round QPV protocol can be attacked with a linear amount of entanglement if the loss is high enough.
The goal of position-based cryptography is for an honest party to use her spatio-temporal position as her only credential in a cryptographic protocol. In particular, Position-Verification aims at verifying that a certain party, called the prover, holds a given position in spacetime. Such a protocol typically goes as follows: a set of verifiers will coordinate and send some challenge to the prover, and it is expected that only someone sitting in the supposed position of the prover can successfully pass the challenge. Two papers established that information theoretic security is not possible for Position-Based Quantum Cryptography (PBQC) protocols: neither in the classical [1], nor in the quantum [2] regime. More precisely, it is always possible for a coalition of adversaries to convince the verifiers, even if none of the adversary sits in the spatiotemporal region where the prover is supposed to be. At the same time, the best known attack strategies require the adversaries to share very entangled states: the attacks described in [2] are based on Vaidman’s protocol for nonlocal computation [3] and consist in the cheaters teleporting some quantum state back and forth, with a total cost in terms of required EPR pairs that scales double exponentially in n, the number of qubits involved in the PBQC protocol. Later, new generic attacks based on port-based teleportation[4] were proposed and require only exponential entanglement to succeed [5]. Finally, it is also know that a linear amount of entanglement is not su cient to break the PBQC protocol [5], [6]. A major open question in the field is to bridge the gap between lower and upper bounds: does there exist a protocol that requires exponentially many EPR pairs to break, or are all PBQC protocols insecure against a coalition of adversaries that shares polynomially many EPR pairs? While we do not settle this question in the present paper, we make some progress towards it by exhibiting new attack strategies with polynomial complexity against a large family of practical protocols. These attacks rely on teleportation gates and their precise complexity can be quantified with the notion of Cli↵ord hierarchy [7] of a quantum gate. We then introduce a new family of protocols, that generalizes a construction by Lau and Lo [8], which appears to be immune against these polynomialcomplexity attacks. Another main challenge is to determine whether there exist some PBQC protocol resistant to channel imperfections, either losses or noise, between the verifiers and
Position-based quantum cryptography (PBQC) allows a party to use its geographical location as its only credential to implement various cryptographic protocols. Such a protocol may lead to important applications in practice. Although it has been shown that any PBQC protocol is breakable if the adversaries pre-share an arbitrarily large entangled state, the security of PBQC in the bounded-quantum-storage model is still an open question. In this paper, we study the performance of various PBQC protocols over a lossy channel under the assumption that no entanglement is pre-shared between adversaries. By introducing the decoy state idea, we show that an extended Bennett-Brassard-1984--type PBQC protocol implemented with a weak coherent source and realistic single-photon detectors can tolerate an overall loss (including both the channel loss and the detection efficiency) of 13 dB if the intrinsic quantum bit error rate is 1%. We also study a few continuous variable PBQC protocols and show that they suffer from a 3-dB loss limitation.
Properties of quantum mechanics such as entanglement and the no-cloning theorem enable new forms of cryptography that are unattainable in classical models of computation. Quantum information, however, is harder to process, store, and transmit than classical information. As such, it is desirable to construct semi-quantum cryptographic protocols where only a subset of the parties have quantum capabilities while the rest operate classically. In this dissertation, we analyze the security of semi-quantum protocols based on two approaches: (i) nonlocality-based techniques and (ii) the hardness of the Learning with Errors (LWE) problem. In the first part of the dissertation, we leverage a form of Bell test to construct a quantum position verification (QPV) protocol. Initial QPV protocols required verifiers to prepare and send quantum states to an honest prover to achieve security against adversaries with bounded memory or entanglement. Later work constructed QPV protocols with classical verifiers but required additional assumptions such as the hardness of LWE or of random circuit sampling. We construct a variant of QPV where classical verifiers verify the spatial locations of two provers. We use the nonlocal correlations produced by the two provers to show security against adversaries with bounded quantum memory without making any computational intractability assumptions. In the second part of the dissertation, we use trapdoor claw-free functions (TCFs) to construct semi-quantum cryptographic protocols. TCFs, which are constructible from the hardness of LWE, have been shown to be immensely valuable in the construction of classically verifiable quantum cryptographic protocols. As opposed to previous work where the server is asked to prepare a claw state and perform Pauli-X and Z measurements, we use a technique that leverages the entire range of qubit measurements on the XY-plane. We use this technique to construct three protocols: (i) a one-round blind remote state preparation protocol that delegates the preparation of an arbitrary state on the XY-plane up to a Pauli-X correction; (ii) a two-round proof of quantumness whose security is directly in terms of the hardness of LWE and only requires the prover to prepare and measure a single claw state; and (iii) a 1-of-2 Puzzle with a single qubit auxiliary state.
This doctoral thesis explores quantum position verification (QPV)---a cryptographic task where one attempts to confirm someone’s geographical location. The core question that QPV aims to answer is: Are you truly at the location you claim to be? To achieve this, QPV combines two pillars of the fundamental laws of nature: (i) special relativity, which limits the speed at which information can travel to the speed of light in vacuum, and (ii) quantum mechanics, which governs the behavior of quantum particles in ways that defy intuition. QPV protocols rely on (i) timing communication between the entities involved in the protocol, and (ii) transmitting information encoded in quantum particles. However, quantum hackers may attempt to pretend to be at the claimed location while actually being elsewhere. This raises the critical question: Can one be certain that the claimed location is genuine and not forged by hackers? This thesis presents new advances toward a fundamental understanding of QPV, its security against powerful quantum hackers, and the feasibility of secure QPV protocols despite experimental challenges that, if exploited by hackers, can severely compromise their security. Furthermore, related to attacks on QPV protocols, this thesis analyzes quantum correlations that emerge in broader cryptographic primitives.