Unclonable Encryption, introduced by Gottesman in 2003 [Quantum Inform. Comput. 3 (2003) 581], is a quantum protocol that guarantees the secrecy of a successfully transferred classical message even when all keys leak at a later time. We propose an Unclonable Encryption protocol with the additional property that the sender’s key material is allowed to leak even in the case of an unsuccessful run. This extra feature makes it possible to achieve secure quantum encryption even when one of the parties is unable to protect its keys against after-protocol theft. Such an asymmetry occurs e.g. in case of server–client scenarios, where the client device is resource constrained and/or located in a hostile environment. Our protocol makes use of a bidirectional quantum channel in a manner similar to the two-way protocol LM05 [Phys. Rev. Lett. 94 (2005) 140501]. Bob sends random qubit states to Alice. Alice flips the states in a way that depends on the message and a shared key, and sends the resulting states back to Bob. Bob recovers Alice’s message by measuring the flips. We prove that our protocol satisfies the definition of unclonable encryption and additionally that the message remains secure even if all of Alice’s keys leak after the protocol. Furthermore, we show that some of the key materials can be safely re-used. Our security proof is formulated in terms of diamond norms, which makes it composable, and allows for noisy quantum channels. We work out the details only for the asymptotics in the limit of long messages. As a side result, we construct a two-way QKD scheme with a high key rate. We show that its key rate is higher than the rate of the two-way QKD scheme LM05 proven by Beaudry et al. [Phys. Rev. A 88 (2013) 062302] for the case of independent channel noise.
Unclonable Encryption, introduced by Gottesman in 2003 [Quantum Inform. Comput. 3 (2003) 581], is a quantum protocol that guarantees the secrecy of a successfully transferred classical message even when all keys leak at a later time. We propose an Unclonable Encryption protocol with the additional property that the sender’s key material is allowed to leak even in the case of an unsuccessful run. This extra feature makes it possible to achieve secure quantum encryption even when one of the parties is unable to protect its keys against after-protocol theft. Such an asymmetry occurs e.g. in case of server–client scenarios, where the client device is resource constrained and/or located in a hostile environment. Our protocol makes use of a bidirectional quantum channel in a manner similar to the two-way protocol LM05 [Phys. Rev. Lett. 94 (2005) 140501]. Bob sends random qubit states to Alice. Alice flips the states in a way that depends on the message and a shared key, and sends the resulting states back to Bob. Bob recovers Alice’s message by measuring the flips. We prove that our protocol satisfies the definition of unclonable encryption and additionally that the message remains secure even if all of Alice’s keys leak after the protocol. Furthermore, we show that some of the key materials can be safely re-used. Our security proof is formulated in terms of diamond norms, which makes it composable, and allows for noisy quantum channels. We work out the details only for the asymptotics in the limit of long messages. As a side result, we construct a two-way QKD scheme with a high key rate. We show that its key rate is higher than the rate of the two-way QKD scheme LM05 proven by Beaudry et al. [Phys. Rev. A 88 (2013) 062302] for the case of independent channel noise.
We re-visit Unclonable Encryption as introduced by Gottesman in~2003 \cite{uncl}. We look at the combination of Unclonable Encryption and Key Recycling, while aiming for low communication complexity and high rate. We introduce a qubit-based prepare-and-measure Unclonable Encryption scheme with re-usable keys. Our scheme consists of a single transmission by Alice and a single classical feedback from Bob. The transmission from Alice to Bob consists entirely of qubits. The rate, defined as the message length divided by the number of qubits, is higher than what can be achieved using Gottesman's scheme~\cite{uncl}. We provide a security proof based on the diamond norm distance, taking noise into account.
Quantum states cannot be cloned. I show how to extend this property to classical messages encoded using quantum states, a task I call ``uncloneable encryption.'' An uncloneable encryption scheme has the property that an eavesdropper Eve not only cannot read the encrypted message, but she cannot copy it down for later decoding. She could steal it, but then the receiver Bob would not receive the message, and would thus be alerted that something was amiss. I prove that any authentication scheme for quantum states acts as a secure uncloneable encryption scheme. Uncloneable encryption is also closely related to quantum key distribution (QKD), demonstrating a close connection between cryptographic tasks for quantum states and for classical messages. Thus, studying uncloneable encryption and quantum authentication allows for some modest improvements in QKD protocols. While the main results apply to a one-time key with unconditional security, I also show uncloneable encryption remains secure with a pseudorandom key. In this case, to defeat the scheme, Eve must break the computational assumption behind the pseudorandom sequence before Bob receives the message, or her opportunity is lost. This means uncloneable encryption can be used in a non-interactive setting, where QKD is not available, allowing Alice and Bob to convert a temporary computational assumption into a permanently secure message.
In a two-way deterministic quantum key distribution (DQKD) protocol, Bob randomly prepares qubits in one of four states and sends them to Alice. To encode a bit, Alice performs an operation on each received qubit and returns it to Bob. Bob then measures the backward qubits to learn about Alice's operations and hence the key bits. Recently, we proved the unconditional security of the final key of this protocol in the ideal device setting. In this paper, we prove that two-way DQKD protocols are immune to all detector-side-channel attacks at Bob's side, while we assume ideal detectors at Alice's side for error testing. Our result represents a step forward in making DQKD protocols secure against general detector-side-channel attacks.
In a deterministic quantum key distribution (DQKD) protocol with a two-way quantum channel, Bob sends a qubit to Alice who then encodes a key bit onto the qubit and sends it back to Bob. After measuring the returned qubit, Bob can obtain Alice's key bit immediately, without basis reconciliation. Since an eavesdropper may attack the qubits traveling on either the Bob-Alice channel or the Alice-Bob channel, the security analysis of DQKD protocol with a two-way quantum channel is complicated and its unconditional security has been controversial. This paper presents a security proof of a single-photon four-state DQKD protocol against general attacks.
We propose a novel concept of securing cryptographic keys which we call “Unclonable Secret Keys,” where any cryptographic object is modified so that its secret key is an unclonable quantum bit-string whereas all other parameters such as messages, public keys, ciphertexts, signatures, etc., remain classical. We study this model in the authentication and encryption setting giving a plethora of definitions and positive results as well as several applications that are impossible in a purely classical setting. In the authentication setting, we define the notion of one-shot signatures, a fundamental element in building unclonable keys, where the signing key not only is unclonable, but also is restricted to signing only one message even in the paradoxical scenario where it is generated dishonestly. We propose a construction relative to a classical oracle and prove its unconditional security. Moreover, we provide numerous applications including a signature scheme where an adversary can sign as many messages as it wants and yet it cannot generate two signing keys for the same public key. We show that one-shot signatures are sufficient to build a proof-of-work-based decentralized cryptocurrency with several ideal properties: it does not make use of a blockchain, it allows sending money over insecure classical channels and it admits several smart contracts. Moreover, we demonstrate that a weaker version of one-shot signatures, namely privately verifiable tokens for signatures, are sufficient to reduce any classically queried stateful oracle to a stateless one. This effectively eliminates, in a provable manner, resetting attacks to hardware devices (modeled as oracles). In the encryption setting, we study different forms of unclonable decryption keys. We give constructions that vary on their security guarantees and their flexibility. We start with the simplest setting of secret key encryption with honestly generated keys and show that it exists in the quantum random oracle model. We provide a range of extensions, such as public key encryption with dishonestly generated keys, predicate encryption, broadcast encryption and more.