Fact-checked by Grok 2 weeks ago

Continuous-variable quantum information

Continuous-variable quantum information is a framework in quantum information science that encodes, processes, and detects quantum information using continuous degrees of freedom, such as the quadrature amplitudes of quantum harmonic oscillators or electromagnetic fields, rather than discrete two-level systems like qubits.^[1]^[2] This approach leverages the infinite-dimensional Hilbert space of continuous-variable (CV) systems, enabling protocols that exploit properties like non-orthogonality and entanglement for tasks in quantum communication and computation.^[2] Primarily implemented in quantum optics, CV quantum information has advanced through experimental techniques such as homodyne detection and squeezed light generation.^[1] At its core, CV quantum information relies on Gaussian states and Gaussian operations, which form the backbone of most practical implementations due to their mathematical tractability and compatibility with linear optical elements.^[3] Gaussian states, characterized by their displacement and covariance matrix in phase space, quantify correlations like entanglement and quantum discord using symplectic geometry and tools from quantum optics.^[3] Key protocols include continuous-variable quantum teleportation, dense coding, and entanglement distillation, often outperforming discrete-variable counterparts in terms of efficiency for certain bosonic systems.^[2] These operations typically involve three stages: preparation of CV-encoded states (e.g., via coherent or squeezed states), manipulation using beam splitters and phase shifters, and measurement via homodyne or heterodyne detection.^[1] In quantum computing, CV approaches emphasize measurement-based quantum computation (MBQC) using highly entangled cluster states generated from multimode squeezed light, offering scalability advantages over gate-based models in infinite-dimensional spaces.^[4] Notable developments include the creation of deterministic three-dimensional cluster states and nonlinear feedforward techniques to mitigate measurement-induced noise.^[4] Error correction in CV systems employs bosonic codes, such as Gottesman-Kitaev-Preskill (GKP) codes, which protect logical information against photon loss and dephasing by encoding in oscillator quadratures.^[4] Progress as of 2025, including multiplexing large-scale cluster states and complete architectures for CV MBQC using spatiotemporal modes, highlights pathways toward fault-tolerant CV quantum processors.^[5] Beyond computation, CV quantum information underpins secure communication protocols and sensing applications, benefiting from the robustness of Gaussian resources in noisy channels, with recent enhancements in CV quantum key distribution (CV-QKD) via adaptive filtering.^[3]^[6] While Gaussian methods dominate due to experimental feasibility, extensions to non-Gaussian states—via photon subtraction or Kerr nonlinearity—promise enhanced capabilities, such as universal quantum computation, though they face challenges in scalability and fidelity.^[3] Overall, the field continues to evolve, driven by integrations with hybrid discrete-continuous systems for broader quantum technologies.^[1]

Introduction

Definition and motivation

Continuous-variable (CV) quantum information refers to the encoding and processing of quantum information using continuous observables, such as the position X and momentum P quadratures of bosonic modes like optical fields, within infinite-dimensional Hilbert spaces.^[7] Unlike discrete-variable (DV) quantum information, which relies on finite-dimensional systems like qubits with bounded spectra, CV approaches exploit unbounded continuous spectra to represent and manipulate information directly analogous to classical continuous variables.^[8] This framework is particularly suited to systems modeled by the quantum harmonic oscillator, where the basic unit consists of infinite energy levels accessed via quadrature measurements, offering a natural extension beyond the two-level qubit paradigm.^[9] The motivation for CV quantum information stems from its compatibility with experimental platforms in quantum optics, where generating and controlling multiple modes is straightforward, enabling scalable implementations without the need for single-photon sources required in many DV protocols.^[7] It provides a natural fit for analog quantum simulation of continuous systems, such as molecular vibrations or many-body Hamiltonians, due to the direct mapping of continuous observables.^[9] Additionally, CV systems support fault-tolerant quantum computing through bosonic error-correcting codes, like Gottesman-Kitaev-Preskill codes, which encode logical qubits in oscillator states to mitigate photon loss and noise.^[10] In quantum communication, CV protocols demonstrate advantages over DV methods in noisy channels, tolerating higher losses and thermal noise while leveraging homodyne detection for efficient key distribution.^[11] While CV approaches enable the direct representation of continuous functions and offer experimental simplicity through linear optics and Gaussian resources, they introduce challenges such as unbounded error propagation, contrasting with the error-bounded nature of DV qubits.^[7] Nonetheless, these features position CV quantum information as a complementary paradigm, excelling in applications where scalability and noise resilience are paramount.^[9]

Historical overview

The field of continuous-variable (CV) quantum information emerged from foundational theoretical proposals in the early 1990s, building on quantum optics and the infinite-dimensional Hilbert spaces of bosonic modes. The initial concept of quantum teleportation for continuous variables was proposed by Lev Vaidman in 1994, extending the discrete-variable protocol to dynamical variables with continuous spectra, such as position and momentum quadratures of light fields. This idea laid the groundwork for manipulating quantum states without direct interaction. In 1998, Samuel L. Braunstein and H. Jeff Kimble advanced the theory by developing a practical protocol using squeezed-state entanglement and homodyne detection, enabling high-fidelity teleportation of coherent states. Concurrently, Seth Lloyd and Samuel L. Braunstein introduced the framework for universal quantum computing over continuous variables in 1999, demonstrating that Gaussian operations combined with non-Gaussian elements like cubic phase gates could perform arbitrary unitary transformations on CV states. Experimental milestones quickly followed, validating these theories and spurring further development. The first demonstration of unconditional CV quantum teleportation was achieved by Akira Furusawa and colleagues in 1998, using squeezed vacuum states to achieve 58% fidelity for coherent state input, surpassing the classical limit. Theoretical proposals for CV cluster states, essential for measurement-based quantum computing, were formalized by Nicolas C. Menicucci and co-authors in 2006, showing how time-multiplexed squeezed light could generate scalable entangled resources via linear optics. This was proposed in 2011 by Nicolas C. Menicucci using four offline squeezers and beam splitters to generate scalable temporal-mode CV cluster states, enabling one-way quantum computation primitives.^[12] Experimental realizations followed, including a large-scale 2D temporal-mode cluster state with 5 × 1240 sites in 2019.^[13] Progress in large-scale entanglement accelerated with the 2014 experiment by Chen et al., generating multipartite entanglement across 60 modes of a quantum optical frequency comb via parametric down-conversion, marking a key step toward fault-tolerant CV architectures. Photonic demonstrations of CV quantum computing advanced significantly in 2022 with the Borealis processor from Xanadu, which performed programmable Gaussian boson sampling on 216 squeezed modes, exhibiting quantum computational advantage over classical simulations.^[14] Recent advances from 2023 to 2025 have focused on practical applications, particularly in quantum communication and error handling. In CV quantum key distribution (QKD), enhancements to the GG02 protocol using adaptive signal processing have enabled satellite links with a threefold increase in secure key rates compared to standard Gaussian modulation, as demonstrated in low-Earth orbit simulations and ground-to-satellite tests.^[6] Integration of CV encoding schemes with machine learning techniques for error mitigation has also gained traction, with 2025 reviews highlighting neural network-based methods to correct quadrature noise in distributed CV systems, improving fidelity in noisy intermediate-scale quantum devices.^[15]

Mathematical Framework

Hilbert space and basic operators

In continuous-variable quantum information, the foundational mathematical structure is provided by the Hilbert space describing bosonic modes, such as those realized in quantum optical systems. For a single mode, the Hilbert space \mathcal{H} is infinite-dimensional and can be represented as L^2(\mathbb{R}), the space of square-integrable wavefunctions \psi(x) over the position variable x, where the inner product is \langle \psi | \phi \rangle = \int_{-\infty}^{\infty} \psi^*(x) \phi(x) \, dx. Equivalently, it is the Fock space spanned by the orthonormal number states |n\rangle (with n = 0, 1, 2, \dots), which are eigenstates of the photon number operator \hat{n} = \hat{a}^\dagger \hat{a}.^[2]^[16] The basic operators acting on this space are the annihilation operator \hat{a} and the creation operator \hat{a}^\dagger, satisfying the bosonic commutation relation [\hat{a}, \hat{a}^\dagger] = 1. These operators generate the number states via \hat{a} |n\rangle = \sqrt{n} |n-1\rangle and \hat{a}^\dagger |n\rangle = \sqrt{n+1} |n+1\rangle. The quadrature operators, which are central to continuous-variable protocols, are defined in dimensionless units (with \hbar = 1) as the position-like operator \hat{X} = \frac{\hat{a} + \hat{a}^\dagger}{\sqrt{2}} and the momentum-like operator \hat{P} = \frac{\hat{a} - \hat{a}^\dagger}{i\sqrt{2}}. They obey the canonical commutation relation [\hat{X}, \hat{P}] = i, mirroring the Heisenberg uncertainty principle.^[2]^[16] For multi-mode systems with N modes, the total Hilbert space is the tensor product \mathcal{H} = \bigotimes_{k=1}^N L^2(\mathbb{R}) (or equivalently, the multi-mode Fock space), allowing for the description of entangled states across modes. The quadrature operators generalize to a vector \hat{\mathbf{r}} = (\hat{X}_1, \hat{P}_1, \dots, \hat{X}_N, \hat{P}_N)^T, with commutation relations [\hat{r}_j, \hat{r}_k] = i \Omega_{jk}, where \Omega is the symplectic form \Omega = \bigoplus_{k=1}^N \begin{pmatrix} 0 & 1 \\ -1 & 0 \end{pmatrix}. This structure introduces a phase-space formalism, where the quadratures \hat{X}_k and \hat{P}_k are analogous to classical position and momentum coordinates, enabling the use of symplectic geometry to analyze multi-mode transformations and uncertainties.^[2]^[16] The vacuum state |0\rangle, which is the ground state of the quantum harmonic oscillator annihilated by \hat{a} |0\rangle = 0, serves as the reference for all continuous-variable states. It exhibits equal uncertainties in the quadratures, with variances \langle (\Delta \hat{X})^2 \rangle = \langle (\Delta \hat{P})^2 \rangle = \frac{1}{2}, saturating the uncertainty relation \langle (\Delta \hat{X})^2 \rangle \langle (\Delta \hat{P})^2 \rangle \geq \frac{1}{4}. This isotropic Gaussian distribution in phase space defines the shot-noise limit for measurements in the field.^[2]^[16]

Quantum states and phase-space representations

In continuous-variable quantum systems, quantum states are described using the density operator \rho, a positive semi-definite Hermitian operator with trace unity that generalizes both pure and mixed states. For pure states, \rho = |\psi\rangle\langle\psi|, where the state vector |\psi\rangle can be represented in the position basis as a wavefunction \psi(x) satisfying \int |\psi(x)|^2 dx = 1, or expanded in the Fock basis as |\psi\rangle = \sum_n c_n |n\rangle with \sum_n |c_n|^2 = 1, where |n\rangle are number states corresponding to n photons. Mixed states arise from statistical ensembles, expressed as \rho = \sum_i p_i |\psi_i\rangle\langle\psi_i| with probabilities p_i \geq 0 and \sum_i p_i = 1, capturing classical uncertainty or decoherence effects inherent in continuous-variable encodings. Coherent states, introduced as the quantum analogs of classical coherent radiation fields, are defined as displaced vacuum states |\alpha\rangle = \hat{D}(\alpha) |0\rangle, where \hat{D}(\alpha) = \exp(\alpha \hat{a}^\dagger - \alpha^* \hat{a}) is the displacement operator with complex parameter \alpha, and |0\rangle is the vacuum state annihilated by the bosonic annihilation operator \hat{a}. These states minimize the uncertainty in the quadrature operators \hat{X} and \hat{P}, satisfying the Heisenberg relation \Delta X \Delta P = 1/2 with equal variances \langle (\Delta X)^2 \rangle = \langle (\Delta P)^2 \rangle = 1/2, and exhibit expectation values \langle \hat{X} \rangle = \sqrt{2} \operatorname{Re}(\alpha) and \langle \hat{P} \rangle = \sqrt{2} \operatorname{Im}(\alpha), making them ideal for encoding classical information in quantum protocols. Squeezed states represent a class of nonclassical pure states where the variance in one quadrature is reduced below the vacuum level at the expense of the conjugate quadrature, such that \langle (\Delta X)^2 \rangle < 1/2 while preserving the uncertainty principle. These states are generated by applying the squeeze operator \hat{S}(\zeta) = \exp\left[ (\zeta^* \hat{a}^2 - \zeta (\hat{a}^\dagger)^2)/2 \right] with complex \zeta = r e^{i\theta} to the vacuum or a coherent state, enabling enhanced precision in measurements. Phase-space representations provide quasiprobability distributions for visualizing continuous-variable states, with the Wigner function serving as a key tool:

W(x,p) = \frac{1}{\pi} \int_{-\infty}^{\infty} \langle x + y | \rho | x - y \rangle e^{-2 i p y} \, dy.

This symmetric function yields marginal probability densities for quadratures via integration, \int W(x,p) \, dx = \langle p | \rho | p \rangle and \int W(x,p) \, dp = |\psi(x)|^2 for pure states, and its negativity signals nonclassical features like those in Fock or squeezed states. Thermal states describe mixed equilibrium states of a harmonic oscillator in contact with a heat bath, characterized by the density operator \rho = \sum_{n=0}^{\infty} p_n |n\rangle\langle n| with Bose-Einstein probabilities p_n = \frac{\bar{n}^n}{(\bar{n} + 1)^{n+1}}, where \bar{n} = \langle \hat{a}^\dagger \hat{a} \rangle = \frac{1}{e^{\beta \hbar \omega} - 1} is the average photon number depending on temperature T = 1/(k_B \beta) and oscillator frequency \omega. These states exhibit symmetric quadrature variances \langle (\Delta X)^2 \rangle = \langle (\Delta P)^2 \rangle = \frac{1}{2} (2\bar{n} + 1) > 1/2, representing excess noise that limits quantum resources in practical implementations.

Core Resources

Gaussian states

Gaussian states represent the most extensively studied class of quantum states in continuous-variable quantum information, owing to their central role in theoretical descriptions and practical implementations of quantum protocols. These states are defined as those whose Wigner function is Gaussian in phase space, which allows for a complete characterization using only the first and second moments of the quadrature operators.^[16] For an N-mode system, the quadratures are collected in the vector \mathbf{R} = (X_1, P_1, \dots, X_N, P_N)^T, where X_j and P_j satisfy [X_j, P_k] = i \delta_{jk} (in units where \hbar = 1). A Gaussian state is fully specified by its displacement vector \mathbf{d} = \langle \mathbf{R} \rangle, representing the means of the quadratures, and its covariance matrix V, a $2N \times 2N real symmetric positive semidefinite matrix with elements

V_{ij} = \frac{\langle R_i R_j + R_j R_i \rangle}{2} - \langle R_i \rangle \langle R_j \rangle.

^[17]^[16] The displacement \mathbf{d} can be shifted locally without altering the quantum correlations encoded in V, making the covariance matrix the primary descriptor for many applications. For physical realizability, V must satisfy the uncertainty principle V + \frac{i}{2} \Omega \geq 0, where \Omega is the symplectic form \Omega = \bigoplus_{j=1}^N \begin{pmatrix} 0 & 1 \\ -1 & 0 \end{pmatrix}.^[16] In single-mode systems (N=1), Gaussian states provide intuitive examples that illustrate their diversity. The coherent state, a displaced vacuum, has covariance matrix V = \frac{1}{2} I, where I is the $2 \times 2 identity matrix, corresponding to equal variances of \frac{1}{2} in both quadratures and serving as the benchmark for classical-like behavior.^[17] The squeezed vacuum state, generated by the squeezing operator, features V = \frac{1}{2} \operatorname{diag}(e^{-2r}, e^{2r}) for squeezing parameter r > 0, reducing the variance below \frac{1}{2} in one quadrature at the expense of the orthogonal one, enabling enhanced precision in measurements.^[16] Thermal states, which are mixed and classical, possess V = \frac{2\bar{n} + 1}{2} I, where \bar{n} \geq 0 is the mean thermal photon number, reflecting increased noise from thermal occupation.^[17] The purity of a Gaussian state, a measure of its mixedness ranging from 0 (completely mixed) to 1 (pure), is determined by the covariance matrix via \mu = [\det(2V)]^{-1/2}.^[16] For physicality across N modes, \det V \geq (1/2)^{2N}, with equality holding for pure states; this bound arises from the requirement that the symplectic eigenvalues \nu_j of V (solutions to the characteristic equation of |i \Omega V|) satisfy \nu_j \geq 1/2 for all j = 1, \dots, N.^[17] Pure Gaussian states occur when all \nu_j = 1/2, corresponding to \det V = (1/2)^{2N}. In multimode settings, the symplectic eigenvalues further facilitate decomposition into a tensor product of single-mode thermal states in the normal form, aiding analysis of global properties.^[16] Non-classicality in Gaussian states manifests primarily through squeezing, where at least one quadrature variance falls below the vacuum level of $1/2, violating classical probability distributions and enabling quantum advantages such as reduced noise in interferometry.^[17] All Gaussian states, including pure ones, possess non-negative Wigner functions, distinguishing them from non-Gaussian states that can exhibit negativities signaling stronger non-classical features; however, squeezed or entangled Gaussians remain non-classical as they cannot be expressed as statistical mixtures of coherent states alone.^[16]

Gaussian operations and transformations

Gaussian unitaries in continuous-variable systems are linear transformations generated by quadratic Hamiltonians in the annihilation and creation operators, preserving the Gaussian form of states under evolution.^[16] These unitaries form a fundamental class of operations in quantum optics and quantum information processing, enabling manipulations of quadrature amplitudes without introducing non-Gaussian features.^[2] Key single-mode Gaussian unitaries include the displacement operator, which shifts the state in phase space; the squeezing operator, which reduces uncertainty in one quadrature at the expense of the other; and the phase shifter, which rotates the phase-space distribution. The displacement operator is defined as

D(\alpha) = \exp(\alpha a^\dagger - \alpha^* a),

where \alpha \in \mathbb{C} is the displacement amplitude and a^\dagger, a are the creation and annihilation operators.^[16] The squeezing operator is given by

S(\xi) = \exp\left(\frac{\xi^* a^2 - \xi (a^\dagger)^2}{2}\right),

with \xi = r e^{i\theta} specifying the squeezing strength r and phase \theta.^[16] The phase shifter acts as

R(\phi) = \exp(i \phi a^\dagger a),

rotating the field by angle \phi.^[16] For two modes, the beamsplitter mixes them via

BS(\theta) = \exp\left[\theta (a^\dagger b - a b^\dagger)\right],

where \theta controls the mixing angle, facilitating interference in multimode setups.^[16] In the multimode setting, Gaussian unitaries correspond to elements of the symplectic group Sp(2N, \mathbb{R}), which act on the vector of quadrature operators \mathbf{R} = (q_1, p_1, \dots, q_N, p_N)^T via \mathbf{R}' = S \mathbf{R} + \mathbf{d}, where S \in Sp(2N, \mathbb{R}) is a symplectic matrix satisfying S \Omega S^T = \Omega and \mathbf{d} is a displacement vector.^[2] The symplectic form \Omega is the block-diagonal matrix with $2 \times 2 blocks \begin{pmatrix} 0 & 1 \\ -1 & 0 \end{pmatrix}, ensuring the canonical commutation relations [\mathbf{R}_i, \mathbf{R}_j] = i \Omega_{ij} are preserved.^[2] This group structure underpins the algebraic description of Gaussian evolutions in systems with N modes. Gaussian channels extend these operations to non-unitary evolutions, such as those encountered in realistic quantum communication. These completely positive trace-preserving maps send Gaussian states to Gaussian states and include models like lossy transmission with transmissivity \eta \in [0,1], which attenuates the signal while adding vacuum noise, and additive classical noise channels that inject Gaussian noise.^[16] In terms of the covariance matrix V of a Gaussian state, a general Gaussian channel acts as V \mapsto T V T^T + N, where T is a real matrix and N a positive semidefinite noise matrix satisfying N + i \Omega \geq 0.^[16] Kraus representations exist for specific channels, such as the beamsplitter model for loss, but the covariance transformation provides a compact description.^[16] Any Gaussian unitary can be decomposed into a sequence of the primitive operations—displacements, squeezers, phase shifters, and beamsplitters—along with local and global rotations, establishing their universality within the Gaussian class.^[2] Achieving full universality for quantum computing in continuous-variable systems, however, requires non-Gaussian gates, which can be implemented via photon counting or subtraction to introduce cubic or higher-order nonlinearities.^[16]

Entanglement in continuous-variable systems

Entanglement in continuous-variable systems primarily arises from Gaussian states, with the two-mode squeezed vacuum serving as a canonical example of a bipartite entangled resource. This state, denoted as |\text{TMS}\rangle = \sum_{n=0}^{\infty} \frac{ (\tanh r)^n }{ \cosh r } |n, n\rangle, where r is the squeezing parameter, is generated via parametric down-conversion in a nonlinear optical medium, such as a χ² crystal pumped by a laser, which correlates the photon numbers in two output modes.^[18] In the limit of infinite squeezing (r \to \infty), the two-mode squeezed vacuum exhibits perfect EPR-like correlations, where the quadratures satisfy \hat{X}_1 - \hat{X}_2 \to 0 and \hat{P}_1 + \hat{P}_2 \to 0 with vanishing variances, realizing the ideal Einstein-Podolsky-Rosen paradox in continuous variables.^[18] Detection of entanglement in these systems relies on separability criteria tailored to continuous variables. For bipartite systems, the Duan et al. criterion provides a sufficient condition for inseparability: for any separable state, the inequality \langle (\Delta (\hat{X}_1 - \hat{X}_2))^2 + (\Delta (\hat{P}_1 + \hat{P}_2))^2 \rangle \geq 2 holds in units where \hbar = 1, with violation indicating entanglement; this is particularly effective for Gaussian states with EPR-type correlations.^[19] For Gaussian states specifically, Simon's positive partial transpose (PPT) criterion offers a necessary and sufficient test: a state is separable if and only if its partial transpose, obtained by time-reversal on one subsystem's covariance matrix V, yields a physical covariance matrix satisfying \tilde{V} + \frac{i}{2} \Omega \geq 0, where \Omega is the symplectic form; symplectic eigenvalues less than 1/2 after partial transposition signal entanglement. Quantification of entanglement in continuous-variable systems often employs distillable measures computable from the covariance matrix. The logarithmic negativity, defined as E_N(\rho) = \log_2 \|\rho^{\text{PT}}\|_1, where \|\cdot\|_1 is the trace norm and \rho^{\text{PT}} is the partial transpose, serves as an upper bound on the distillable entanglement and is efficiently calculable for Gaussian states via their symplectic eigenvalues.^[18] For symmetric two-mode Gaussian states, the entanglement of formation—the minimum average entropy of pure-state decompositions—has a closed-form expression involving the squeezing parameter, showing that two-mode squeezed states maximize EPR correlations for a given entanglement amount.^[20] Multipartite entanglement extends these concepts to graph-based Gaussian states, such as cluster states, which are generated by applying controlled-phase gates on a lattice of single-mode squeezed states and enable measurement-based quantum computation (MBQC). These states are characterized by nullifiers \hat{\delta}_j = \hat{X}_j - \sum_{k \in N(j)} \hat{P}_k, where N(j) denotes neighboring nodes in the graph; small variances of these operators (approaching zero in the infinite-squeezing limit) confirm the multipartite entanglement required for universal CV MBQC.

Protocols and Applications

Measurement techniques

In continuous-variable quantum information (CVQI), measurement techniques are crucial for extracting information from quantum states encoded in continuous degrees of freedom, such as the quadratures of light fields. These methods enable the verification of quantum resources like squeezing and entanglement, as well as the implementation of protocols requiring projective measurements. The primary approaches include homodyne and heterodyne detection, which are Gaussian measurements well-suited to optical implementations, and photon counting, which probes non-Gaussian features but faces practical limitations. Efficiency and noise considerations are integral to all techniques, affecting the fidelity of state reconstruction and the detection of quantum correlations.^[18]^[21] Homodyne detection projects the signal field onto a specific quadrature phase, defined as X_\theta = X \cos \theta + P \sin \theta, where X and P are the amplitude and phase quadratures, respectively, and \theta determines the measurement basis. This is achieved by mixing the signal with a strong local oscillator (LO) at a beamsplitter, followed by balanced photodetection of the output intensities, which yields the quadrature value proportional to the difference in photocurrents. The balanced setup minimizes common-mode noise and allows measurement of the variance \langle \Delta X_\theta^2 \rangle, enabling detection of squeezing below the vacuum level of 1/2 when expressed in natural units. Homodyne detection is highly efficient for single-quadrature measurements and has been pivotal in verifying continuous-variable entanglement in optical experiments.^[18]^[21] Heterodyne detection simultaneously measures both quadratures X and P by splitting the signal field at a 50:50 beamsplitter, with each output mixed with a LO of appropriate phase. This dual-rail approach effectively doubles the measurement rate but introduces an additional half-unit of vacuum noise per quadrature due to the beamsplitter's vacuum input, resulting in a total added noise of 1/2 compared to the shot-noise limit. It is particularly useful for quantum state tomography, where repeated measurements reconstruct the full phase-space distribution, such as the Wigner function, providing a complete characterization of Gaussian states. Despite the noise penalty, heterodyne's ability to access both quadratures without phase adjustments makes it preferable in scenarios requiring rapid full-state information.^[18]^[21] Photon counting measures the photon number in Fock states |n\rangle, offering access to non-Gaussian quantum features essential for resource generation beyond Gaussian operations, such as cat states or Gottesman-Kitaev-Preskill encoding. In optical CVQI, this is implemented using avalanche photodiodes or superconducting nanowire detectors, which register discrete clicks corresponding to photon arrivals. However, challenges arise from high losses in free-space or fiber optics, which degrade the number resolution, and the low detection efficiency for single photons, often below 90%, limiting its application to low-photon-number regimes. Photon counting thus complements Gaussian methods but is less routine in CVQI due to these practical hurdles.^[18]^[21]^[22] Quantum efficiency \eta < 1 in these detectors introduces added noise, modeled as an effective loss channel that mixes the signal with vacuum, reducing the measured squeezing or entanglement by a factor related to \eta. For homodyne and heterodyne setups, electronic noise from photodiodes and amplifiers further contributes variance, while continuous measurements induce backaction that disturbs the quantum state via the measurement process itself. Achieving near-unity efficiency, as demonstrated in integrated photonic detectors with \eta > 95\%, is vital for scaling CVQI applications, though residual noise sources remain a key limitation in experimental realizations.^[21]^[18]

Quantum communication protocols

Continuous-variable (CV) quantum communication protocols exploit Gaussian states and operations to enable secure and efficient information transfer, leveraging the infinite-dimensional Hilbert space of bosonic modes for enhanced capacities compared to discrete-variable counterparts. These protocols typically rely on shared entanglement resources, such as two-mode squeezed vacuum states, to surpass classical limits in tasks like state transfer, capacity enhancement, and key generation. Central to their implementation are Gaussian measurements, such as homodyne or heterodyne detection, which extract quadrature information while preserving quantum correlations. Security in these protocols often stems from the no-cloning theorem and uncertainty principles, with proofs against collective attacks ensuring robustness in practical noisy channels. One foundational CV protocol is quantum teleportation, which transfers an unknown quantum state from a sender (Alice) to a receiver (Bob) using a shared entangled resource and classical communication. Proposed by Braunstein and Kimble in 1998, the protocol employs an Einstein-Podolsky-Rosen (EPR)-like pair generated from two-mode squeezing, where Alice performs joint homodyne measurements on the input state and her mode of the EPR pair, sending the results to Bob, who applies displacement operations to reconstruct the state.^[23] The average fidelity of teleportation for coherent input states is given by F = \frac{1}{1 + e^{-2r}}, where r is the squeezing parameter of the EPR resource; in the limit of infinite squeezing (r \to \infty), F \to 1, while classical measure-and-prepare strategies achieve at most F = 0.5.^[23] This protocol was experimentally realized the same year using optical fields, achieving a fidelity of $0.58 \pm 0.02 > 0.5, confirming quantum advantage. Continuous-variable dense coding enables a sender to encode more classical information into a quantum channel than classically possible, effectively doubling the capacity when entanglement is available. Introduced by Braunstein and Kimble in 1999, the protocol uses a shared two-mode squeezed state, where Alice applies a displacement operation to her mode based on two bits of classical information before sending it to Bob, who jointly measures both modes via homodyne detection. In the strong-squeezing limit, this achieves a capacity of 2 bits per mode, twice the 1 bit of a classical channel, by exploiting the anticorrelations in the squeezed quadratures.^[24] A variant employs a one-way classical squeezing channel to prepare the entanglement, maintaining the doubling effect while simplifying resource generation.^[25] CV quantum key distribution (QKD) protocols generate shared secret keys secure against eavesdropping, using Gaussian-modulated coherent states for practical implementation. The seminal GG02 protocol, developed by Grosshans and Grangier in 2002, has Alice send coherent states with quadratures modulated according to a Gaussian distribution, while Bob performs heterodyne detection to measure both quadratures simultaneously.^[26] Post-processing includes parameter estimation, error correction, and privacy amplification; reverse reconciliation—where Bob's measurements serve as the key basis—enhances security by mitigating excess noise on Alice's side.^[26] Security against collective attacks is proven using entropic uncertainty relations, yielding positive key rates over distances up to 100 km in fiber with modest modulation variance.^[27] Recent enhancements in 2025 have optimized the GG02 framework for satellite-to-ground links, dynamically adjusting parameters to handle atmospheric turbulence and achieve up to threefold improvements in key rates compared to static configurations.^[6] Entanglement distribution in CV systems propagates quantum correlations over lossy channels using two-mode squeezed states as the resource. These states, generated via parametric down-conversion, maintain logarithmic negativity as a measure of entanglement even under symmetric loss, enabling distribution over tens of kilometers in optical fiber before distillation is needed.^[28] Entanglement swapping extends this by linking independent distribution links: two parties each share a two-mode squeezed state with a central node, which performs joint Gaussian measurements (e.g., beam splitters and homodyne detection) to project the outer modes into an entangled state, effectively bridging distant nodes without direct transmission.^[29] This Gaussian swapping operation preserves the overall entanglement, with fidelity scaling as e^{-2r} for squeezing r, and is crucial for scaling quantum networks over lossy infrastructures.^[29]

Quantum computing approaches

Continuous-variable (CV) quantum computing encompasses several models that leverage the infinite-dimensional Hilbert space of bosonic modes, such as optical or microwave fields, to perform computational tasks. The circuit model implements quantum gates through a process known as gate teleportation, where an input state is entangled with ancillary resources using Gaussian operations like beam splitters and phase shifters, followed by homodyne measurements to project the desired gate onto the target mode. This approach allows for the realization of Gaussian unitaries, which form the backbone of CV computations, but requires an additional non-Gaussian element for universality. Kerr nonlinearities, generated via interactions like χ(3) processes in optical fibers or superconducting circuits, or photon counting measurements that introduce projective nonlinearity, enable the implementation of operations such as the cubic phase gate, completing a universal gate set.^[30] An alternative paradigm is measurement-based quantum computing (MBQC), which uses large-scale CV cluster states as universal resources instead of sequential gates. These cluster states, generated from squeezed vacuum states via Gaussian entangling operations, encode computational wires and nodes in a graph-like structure, where adaptive homodyne measurements on individual modes drive the computation by applying local corrections based on measurement outcomes. This model inherits the entanglement properties from CV systems, allowing for scalability; recent experiments have demonstrated generation of cluster states with thousands of temporal or spatial modes using integrated photonic chips and time-multiplexing techniques. Fault tolerance in CV quantum computing is achieved through bosonic error correction codes that protect logical information against noise in the continuous quadratures. The Gottesman-Kitaev-Preskill (GKP) code encodes a discrete qubit into a superposition of position-momentum grid states within a single oscillator, enabling correction of small displacement errors via syndrome measurements on ancillary modes.^[31] Complementary to GKP, cat codes encode information in superpositions of coherent states (|α⟩ ± |-α⟩), which naturally suppress phase-flip errors and support bias-preserving gates like controlled-phase operations that maintain the code's error bias toward bit flips. Recent advances in 2025 have improved GKP stabilization using reservoir engineering, where engineered dissipation in superconducting circuits confines the oscillator to the logical subspace, achieving lifetimes exceeding milliseconds and enabling repeated error correction cycles. CV quantum computing demonstrates computational advantages in specific tasks, such as analogs of boson sampling, where Gaussian boson sampling (GBS) with squeezed states and linear interferometers produces output distributions that are exponentially hard to simulate classically, as verified in photonic experiments with up to 100 modes. Hybrid approaches combining CV and discrete-variable systems further extend capabilities, allowing for efficient approximation of continuous functions through variational circuits that leverage bosonic modes for high-dimensional embeddings and qubits for nonlinear activations.^[32]

Implementations and Experiments

Optical realizations

Squeezed light, a fundamental resource for continuous-variable quantum information processing, is primarily generated through spontaneous parametric down-conversion (SPDC) in nonlinear optical crystals such as potassium titanyl phosphate (KTP) or lithium niobate, where a pump photon splits into correlated signal and idler photons exhibiting reduced quantum noise in one quadrature.^[33] This process, first demonstrated in optical cavities in the 1980s, enables the creation of single-mode squeezed vacuum states essential for tasks like quantum metrology and error reduction in protocols.^[33] Experimental setups typically involve a continuous-wave laser pumping the crystal within an optical resonator to enhance efficiency, producing squeezing levels that surpass the standard quantum limit. By 2023, laboratory experiments achieved up to 15 dB of squeezing, sufficient for fault-tolerant thresholds in certain computational schemes when combined with low-loss optics. Integrated photonics has emerged as a compact platform for realizing Gaussian operations in continuous-variable systems, leveraging silicon or thin-film lithium niobate waveguides to implement beamsplitters, phase shifters, and modulators with sub-decibel losses.^[34] These chips enable on-chip generation and manipulation of squeezed states via integrated SPDC sources, reducing alignment challenges associated with bulk optics and facilitating scalable circuits for quantum communication and simulation.^[35] For instance, silicon photonic integrated circuits support homodyne detection for quadrature measurements, while lithium niobate platforms offer electro-optic tunability for dynamic control of quantum states. Xanadu's Strawberry Fields software framework further supports the design and simulation of such programmable continuous-variable circuits on photonic hardware, integrating with experimental devices to optimize gate decompositions and state preparation.^[36] Large-scale entangled resources, such as cluster states, are generated in optical systems using time- or frequency-multiplexing of squeezed light through networks of beamsplitters and delay lines, enabling measurement-based quantum computing. A landmark experiment in 2013 produced a fully characterized 10,000-mode continuous-variable cluster state via time-domain multiplexing, demonstrating scalability through temporal encoding of modes.^[37] By 2022, experimental progress included the creation of multimode squeezed states with hundreds of independent modes using frequency combs and integrated arrays, approaching thresholds for practical quantum advantage in boson sampling tasks.^[38] Readout of these states relies on balanced homodyne detection arrays, where local oscillator phases are scanned to project onto arbitrary quadratures, with integrated versions achieving gigahertz bandwidths for real-time processing.^[35] Key challenges in optical realizations include mitigating photon loss, which degrades squeezing and entanglement, addressed through quantum error correction codes tailored for continuous variables, such as Gottesman-Kitaev-Preskill schemes that encode logical qubits in bosonic modes.^[39] These codes enable fault-tolerant operations by detecting and correcting loss-induced errors via syndrome measurements on ancillary modes, with recent photonic demonstrations showing improved fidelity in lossy channels.^[40] Additionally, achieving stable room-temperature operation remains critical, as thermal noise in nonlinear crystals and detectors can limit squeezing, though advances in cryogenic-free setups and active stabilization have enabled robust, ambient-condition experiments without liquid helium cooling.

Emerging platforms and recent advances

In trapped ion systems, continuous-variable quantum information can be encoded in the motional states of ions, where vibrational modes act as bosonic degrees of freedom manipulated via laser interactions to perform Gaussian operations such as squeezing and displacement. This approach leverages the ions' internal electronic states for readout and control, enabling precise preparation of coherent superpositions in the motional Hilbert space.^[41] Mechanical oscillators coupled to trapped ions further extend this platform by integrating optomechanical interactions, allowing for the generation of entangled continuous-variable states between motional modes and mechanical resonators for enhanced quantum sensing and simulation tasks. Proposals for hybrid continuous-variable quantum computing in trapped ions, emerging in the mid-2010s, combine discrete internal qubit states with continuous motional variables to realize universal gates, including non-Gaussian operations essential for fault-tolerant computation.^[42] These hybrid schemes exploit the Jaynes-Cummings interaction between ionic states and vibrations to implement conditional beam-splitter gates, bridging discrete- and continuous-variable paradigms for scalable processing.^[43] Superconducting circuits provide another key platform for continuous-variable quantum information, employing microwave cavities as harmonic oscillators to host bosonic modes suitable for encoding logical information.^[44] These cavities enable the implementation of Gottesman-Kitaev-Preskill (GKP) codes, which discretize the continuous phase space into a grid for error correction against analog noise.^[45] In 2024, significant progress was made in stabilizing cat states within these circuits using Kerr nonlinearity and parametric driving, achieving robust superpositions that mitigate bit-flip errors through engineered dissipation.^[46] Such demonstrations have pushed coherence times in superconducting cavities to tens of milliseconds, facilitating practical bosonic encoding.^[47] From 2023 to 2025, continuous-variable quantum key distribution (CV-QKD) advanced with measurement-device-independent (MDI) protocols demonstrated over 100 km of optical fiber, achieving secure key rates while countering detector-based vulnerabilities through untrusted relay architectures. Distributed continuous-variable processing networks have emerged as enablers for quantum internet applications, interconnecting remote nodes via fiber links for shared entanglement and multipartite computations. Machine learning-aided techniques for continuous-variable encoding have gained traction, optimizing fault-tolerant strategies by learning adaptive discretizations that enhance error thresholds in bosonic codes. In 2025, notable advances include the deterministic generation of an eight-mode entangled continuous-variable state on an integrated optical chip and high-rate CV-MDI-QKD over 100 km achieving 2.6 Mbit/s secure key rates.^[48]^[49] Scalability in these emerging non-optical platforms hinges on coherence times, with trapped ion motional states sustaining coherence up to seconds through sympathetic cooling and low-heating traps, outperforming superconducting microwave cavities' typical millisecond lifetimes limited by dielectric losses.^[50] In contrast, optical systems benefit from near-unlimited propagation coherence in low-loss fibers, though non-optical alternatives offer superior integration for on-chip quantum networks despite ongoing challenges in extending coherence for large-scale operations.^[47]

Advanced Topics

Classical simulation methods

Classical simulation methods for continuous-variable (CV) quantum information focus on efficiently emulating quantum processes on classical computers, particularly for systems lacking quantum advantage. Gaussian states and operations, which preserve the Gaussian form under linear transformations, can be simulated deterministically by tracking the evolution of the covariance matrix under Gaussian channels.^[51] This approach exploits the fact that Gaussian states are fully characterized by their first and second moments, allowing polynomial-time computation for linear optical networks involving beam splitters, phase shifters, and homodyne or heterodyne measurements.^[51] For pure Gaussian processes, the simulation complexity scales as O(n^3) for n modes, making it feasible for moderate system sizes.^[52] A key extension arises from the positivity of the Wigner function, which renders certain CV quantum computations classically simulable. If the Wigner function of a quantum state remains nonnegative throughout the computation, the system behaves classically, as probabilities can be directly sampled from this quasiprobability distribution without negative interference terms.^[53] This result generalizes the Gottesman-Knill theorem from discrete-variable stabilizer states to CV systems, enabling efficient simulation of Gaussian operations on states with positive Wigner functions, such as coherent and squeezed vacuum states under Clifford-Gaussian evolutions.^[54] Hudson's theorem underpins this by guaranteeing that nonnegative Wigner functions correspond to classical probability distributions.^[53] For non-Gaussian extensions, where the Wigner function develops negativities, exact simulation becomes exponentially costly, but approximate methods using tensor networks provide practical solutions for small non-Gaussianities. Matrix product states and their functional variants represent CV wavefunctions or density operators in a compressed tensor format, allowing simulation of circuits with limited entanglement or depth. These techniques efficiently handle non-Gaussian inputs or operations, such as single-photon additions, by contracting the network to compute expectation values or sample outcomes, though full Wigner negativity incurs exponential resource demands.^[55] Simulation thresholds highlight the boundary of classical efficiency: Gaussian CV systems are simulable up to hundreds of modes, such as 256 modes as of 2025, on current hardware, beyond which memory and time requirements grow prohibitive without approximations for certain tasks like Gaussian boson sampling.^[56]^[57] Quantum speedup in CV protocols typically requires non-Gaussian resources, such as photon loss introducing effective non-Gaussianity or Kerr nonlinearities generating cat states, which evade efficient classical simulation and enable tasks like universal quantum computation.

Hybrid discrete-continuous systems

Hybrid discrete-continuous systems integrate discrete-variable (DV) quantum information processing, typically using qubits, with continuous-variable (CV) systems that employ infinite-dimensional Hilbert spaces such as bosonic modes. This hybridization leverages the strengths of both paradigms: the fault-tolerant logical encoding possible in DV systems and the scalability and ease of generating entanglement in CV optical platforms. Interfaces between these systems enable protocols that combine CV for efficient quantum communication over long distances with DV for robust computation, addressing limitations in purely CV or DV architectures. A key example is the Gottesman-Kitaev-Preskill (GKP) encoding, which embeds logical DV qubits into the CV Hilbert space of a harmonic oscillator to protect against small displacement errors in position and momentum. The logical zero state in the square-lattice GKP code is approximated as a superposition of position eigenstates:

|0_L\rangle \approx \sum_{m=-\infty}^{\infty} |q + \sqrt{\pi} m \rangle_{\text{GKP}},

with a similar form for the logical one state shifted by \sqrt{\pi}/2. This encoding stabilizes the logical qubit against small errors by projecting measurements onto a discrete lattice, allowing standard DV gates to be implemented approximately on the CV mode while enabling error correction via syndrome measurements. The GKP code thus serves as a bridge, facilitating hybrid operations where CV modes act as error-protected carriers for DV information.^[31]^[58] Discrete qubits or qudits can also approximate computations involving continuous functions, such as path integrals or eigenvalue problems, by discretizing the underlying continuous spaces. For instance, quantum algorithms for evaluating path integrals in quantum mechanics achieve a query complexity scaling as O(1/\epsilon) to achieve precision \epsilon, in contrast to classical methods that scale exponentially as \exp(1/\epsilon) for certain smooth integrands. Similarly, for approximating the lowest eigenvalue of a continuous Hamiltonian via quantum phase estimation on a discretized grid, DV systems provide polynomial scaling in $1/\epsilon and system dimension, enabling efficient simulation of continuous dynamics that would be intractable classically. These approaches use DV hardware to perform finite approximations of infinite-dimensional problems, often outperforming direct CV methods in error-prone environments. Hybrid protocols exploit CV systems for communication and DV for processing, with transduction mechanisms converting between the domains. In quantum networks, CV optical modes transmit entangled states over fiber or free space, while DV superconducting or ion-trap qubits perform local gates; transduction is achieved via electro-optic or optomechanical devices that map CV quadratures to DV states. For example, microwave-to-optical transducers have enabled the readout of superconducting qubit states via optical fibers, preserving quantum coherence with fidelity of approximately 85% in recent demonstrations.^[59] Such interfaces support hybrid quantum key distribution (QKD) networks, where CV links distribute keys at high rates over metropolitan scales, interfaced with DV repeaters for long-haul security.^[59] The advantages of hybrid systems include combining the room-temperature scalability of optical CV entanglement generation with the mature error-correction codes of DV platforms, potentially enabling fault-tolerant quantum networks. Challenges persist in achieving high-fidelity transduction, with current efficiencies limited by noise in the conversion process, though 2025 advances in hybrid QKD encoders—integrating DV modulation with CV homodyne detection—have demonstrated secure key rates up to 1 Mbit/s over 50 km fibers, paving the way for interoperable quantum infrastructures.^[60]^[61]

References

[1]
[1008.3468] Continuous Variable Quantum Information Processing
Aug 20, 2010 · In this paper we review the basic principles of CV quantum information processing with main focus on recent developments in the field.
[2]
[quant-ph/0410100] Quantum information with continuous variables
Oct 13, 2004 · We review the progress in quantum information based on continuous quantum variables, with emphasis on quantum optical implementations.
[3]
Continuous variable quantum information: Gaussian states and ...
Jan 19, 2014 · In this article, we provide a brief, and hopefully didactic, exposition of Gaussian state quantum information and its contemporary uses.
[4]
Quantum Information Processing and Continuous Variable ... - Nature
Continuous variable: An approach in quantum information that encodes information in continuous degrees of freedom, such as the quadrature amplitudes of the ...
[5]
https://www.nature.com/articles/s41598-025-02899-8
[6]
https://www.nature.com/articles/s42005-025-02317-5
[7]
https://doi.org/10.1103/RevModPhys.77.513
[8]
https://arxiv.org/pdf/quant-ph/0410100.pdf
[9]
Comparison of Discrete Variable and Continuous ... - Quantum Journal
Jun 20, 2024 · We find that in the low phase noise regime but high thermal noise regime, CV-QKD can tolerate more loss compared to DV-QKD.
[10]
Quantum computational advantage with a programmable photonic ...
Jun 1, 2022 · Here we report quantum computational advantage using Borealis, a photonic processor offering dynamic programmability on all gates implemented.
[11]
Enhanced continuous-variable quantum key distribution protocol via ...
Oct 15, 2025 · Quantum key distribution (QKD) provides secure communication using quantum mechanics, with continuous-variable QKD (CV-QKD) being an attractive ...
[12]
(PDF) Physics-Inspired Machine Learning for Quantum Error Mitigation
Jun 11, 2025 · Experimental results demonstrate that NNAS outperforms current methods across a spectrum of metrics, including error mitigation capability, ...
[13]
[1110.3234] Gaussian Quantum Information - arXiv
Oct 14, 2011 · This review ... Access Paper: View a PDF of the paper titled Gaussian Quantum Information, by Christian Weedbrook and 6 other authors.
[14]
Gaussian states in continuous variable quantum information - arXiv
Mar 31, 2005 · A special emphasis has been given to the phase space analysis of quantum dynamics and to the role of Gaussian states in continuous variable quantum information.
[15]
Quantum information with continuous variables | Rev. Mod. Phys.
Jun 29, 2005 · This article reviews the progress in quantum information based on continuous quantum variables, with emphasis on quantum optical implementations.Abstract · Article Text · Continuous Variables in... · Quantum Communication with...
[16]
Inseparability criterion for continuous variable systems - arXiv
Aug 17, 1999 · An inseparability criterion based on the total variance of a pair of Einstein-Podolsky-Rosen type operators is proposed for continuous variable systems.
[17]
Entanglement of Formation for Symmetric Gaussian States
Sep 3, 2003 · We show that for a fixed amount of entanglement, two-mode squeezed states are those that maximize Einstein-Podolsky-Rosen–like correlations.
[18]
Gaussian quantum information | Rev. Mod. Phys.
May 1, 2012 · Christian Weedbrook ... The large subject matter means that this review will take a mostly theoretical approach to Gaussian quantum information.
[19]
Continuous‐variable quantum information processing - Andersen
Apr 23, 2010 · In this paper we review the basic principles of CV quantum information processing with main focus on recent developments in the field. We will ...
[20]
Teleportation of Continuous Quantum Variables | Phys. Rev. Lett.
A protocol is presented for teleporting the wave function of a single mode of the electromagnetic field with high fidelity using squeezed-state entanglement.
[21]
[quant-ph/9910010] Dense Coding for Continuous Variables - arXiv
Oct 4, 1999 · This paper presents a scheme for dense quantum coding of electromagnetic field quadrature amplitudes using shared entanglement, achieving high ...Missing: original | Show results with:original
[22]
Dense Coding for Continuous Variables - Caltech Authors
Oct 15, 2019 · Abstract. A scheme to achieve dense quantum coding for the quadrature amplitudes of the electromagnetic field is presented.Missing: doubles | Show results with:doubles
[23]
Continuous Variable Quantum Cryptography Using Coherent States
Jan 16, 2002 · Continuous Variable Quantum Cryptography Using Coherent States. Frédéric Grosshans and Philippe Grangier. Laboratoire Charles Fabry de l ...
[24]
Continuous variable quantum cryptography using coherent states
Sep 18, 2001 · Title:Continuous variable quantum cryptography using coherent states. Authors:Frédéric Grosshans, Philippe Grangier. View a PDF of the paper ...
[25]
Continuous-variable entanglement distillation over a general lossy ...
Sep 9, 2009 · This paper begins by describing the important characteristics of the scheme in [5] and finds that the purification power of this scheme is ...
[26]
Continuous-Variable Entanglement Swapping - MDPI
We present a very brief overview of entanglement swapping as it relates to continuous-variable quantum information. The technical background required is ...
[27]
[2002.01402] Universal Gate Set for Continuous-Variable Quantum ...
Feb 4, 2020 · We provide an explicit construction of a universal gate set for continuous-variable quantum computation with microwave circuits.
[28]
[quant-ph/0008040] Encoding a qubit in an oscillator - arXiv
May 13, 2001 · Title:Encoding a qubit in an oscillator. Authors:Daniel Gottesman, Alexei Kitaev, John Preskill. View a PDF of the paper titled Encoding a ...
[29]
[2507.01146] Quantum Computing in Discrete- and Continuous ...
Jul 1, 2025 · This thesis develops a theoretical framework for hybrid continuous-variable (CV) and discrete-variable (DV) quantum systems, with emphasis on quantum control.
[30]
Generation of Squeezed States by Parametric Down Conversion
Nov 17, 1986 · Squeezed states of the electromagnetic field are generated by degenerate parametric down conversion in an optical cavity.
[31]
Integrated photonics on thin-film lithium niobate
May 3, 2021 · Quantum operations can be performed using low-loss optical waveguides, beam splitters, and phase shifters. With the growing complexity of ...
[32]
Integrated photonic platform for quantum information with ... - Science
Dec 7, 2018 · Integrated quantum photonics provides a scalable platform for the generation, manipulation, and detection of optical quantum states by confining light inside ...
[33]
Strawberry Fields: A Software Platform for Photonic Quantum ...
Mar 11, 2019 · We introduce Strawberry Fields, an open-source quantum programming architecture for light-based quantum computers, and detail its key features.Missing: CV | Show results with:CV
[34]
Ultra-large-scale continuous-variable cluster states multiplexed in ...
Nov 17, 2013 · We deterministically generate and fully characterize a continuous-variable cluster state containing more than 10,000 entangled modes. This is, ...Missing: Menicucci 10000 mode
[35]
2022 Roadmap on integrated quantum photonics - IOPscience
This technology, together with the recent generation of large CV cluster states and theoretical progress for fault tolerant computation hold great potential ...
[36]
Logical states for fault-tolerant quantum computation with ... - Science
Jan 18, 2024 · In terms of wave functions, GKP states have a periodic sharp-peaked grid structure in both of their quadratures, which allows for the detection ...
[37]
Quantum error cancellation in photonic systems
We show that our quantum error cancellation protocol can undo photon losses in expectation value estimation tasks.
[38]
Continuous variables quantum computation over the vibrational ...
Aug 15, 2017 · In this paper, we investigate the idea of exploiting the vibrational modes of trapped ions as the physical platform of quantum computing, i.e. ...
[39]
Universal hybrid quantum computing in trapped ions | Phys. Rev. A
The Jaynes-Cummings-type interaction between the (discrete) internal states of ions coupled to their (continuous) motional states can generate Gaussian [28–30]Article Text · INTRODUCTION · THEORY · NUMERICAL SIMULATIONS
[40]
Hybrid Quantum Computing with Conditional Beam Splitter Gate in ...
Apr 27, 2020 · Trapped ions, with discrete internal states and motional modes that can be described by continuous variables in an infinite-dimensional ...
[41]
Advancements in Superconducting Microwave Cavities and Qubits ...
Apr 18, 2023 · Superconducting microwave cavities with ultra-high Q-factors are revolutionizing the field of quantum computing, offering long coherence times ...Missing: CVQI | Show results with:CVQI
[42]
Quantum Error Correction with the Gottesman-Kitaev-Preskill Code
Jun 23, 2021 · The Gottesman-Kitaev-Preskill (GKP) code was proposed in 2001 by Daniel Gottesman, Alexei Kitaev, and John Preskill as a way to encode a qubit in an oscillator.
[43]
Stabilizing the Kerr arbitrary cat states and holonomic universal control
Jun 19, 2024 · In this work, we use a parametrically driving KNR to propose a novel engineering Hamiltonian that can stabilize arbitrary cat states.
[44]
Superconducting Cavity Qubit with Tens of Milliseconds Single ...
Sep 14, 2023 · We report a single-photon qubit encoded in a novel superconducting cavity with a coherence time of 34 ms, representing an order of magnitude improvement.Missing: CVQI | Show results with:CVQI
[45]
Quantum machine learning via continuous-variable cluster states ...
Jun 2, 2025 · We propose a new approach for a photonic platform suitable for distributed quantum machine learning and exhibiting memory.
[46]
[2007.04149] Coherence properties of highly-excited motional states ...
Jul 8, 2020 · We present a study of the coherence properties of a variety of motional states of a single ion confined in a Penning ion trap.
[47]
Efficient Classical Simulation of Continuous Variable Quantum ...
Sep 11, 2001 · We obtain sufficient conditions for the efficient simulation of a continuous variable quantum algorithm or process on a classical computer.Missing: Wigner positivity
[48]
Classical simulation of Gaussian quantum circuits with non ...
Jul 6, 2021 · We consider Gaussian quantum circuits supplemented with non-Gaussian input states and derive sufficient conditions for efficient classical strong simulation of ...
[49]
Positive Wigner functions render classical simulation of quantum ...
Aug 17, 2012 · Noting the fact that Clifford and Gaussian operations preserve the positivity of the Wigner function, our result generalizes the Gottesman-Knill ...
[50]
Positive Wigner Functions Render Classical Simulation of Quantum ...
Dec 4, 2012 · Here we present a generalization of the Gottesman-Knill theorem stating that one can simulate, by efficient classical sampling, quantum circuits ...
[51]
https://arxiv.org/abs/quant-ph/0109047
[52]
Simulating complex networks in phase space: Gaussian boson ...
Validation based on low-order correlations [29] or direct classical simulation may be susceptible to mock-ups. ... 50 modes [32] . Since a direct ...
[53]
Encoding a qubit in an oscillator | Phys. Rev. A
Jun 11, 2001 · Quantum error-correcting codes are constructed that embed a finite-dimensional code space in the infinite-dimensional Hilbert space of a system described by ...
[54]
All-optical superconducting qubit readout | Nature Physics
Feb 11, 2025 · Here we realize radio-over-fibre qubit readout at millikelvin temperatures. We use one device to simultaneously perform upconversion and downconversion.
[55]
Hybrid Encoder for Discrete and Continuous Variable QKD
May 20, 2025 · This work demonstrates a hybrid QKD encoder that bridges the advantages of discrete and continuous variables. Utilizing a self-compensating ...
[56]
https://link.aps.org/doi/10.1103/PhysRevA.105.012427