arXiv:2610.16198
Time-Frequency Continuous-Variable Quantum Teleportation with Single Photons
Mohamed Meguebel, Nadia Belabas, Nicolas Fabre
Abstract
Quantum teleportation, a key building block for quantum technologies, enables distant parties, Alice and Bob, to transfer an unknown quantum state using entanglement, local operations, and classical communication. In this work, we propose a quantum teleportation protocol for single-photon time-frequency continuous variables, where time is the time of arrival at a detector. We detail the operational requirements of the protocol, notably the nonlinear interactions needed for the photon-photon time-frequency entangling gate. By explicitly treating the detector position as the reference frame for time of arrival, we show that, even with ideal classical and quantum channels, the input-output fidelity is not unity unless the receiving party can correct the detector-position mismatch. This results in a trade-off between teleportation fidelity and the temporal bandwidth of the input photon wavepacket. The protocol itself can be used for remote detector synchronization. In particular, we show that time-frequency cat and Gottesman-Kitaev-Preskill states provide an internal temporal ruler through fidelity revivals. We further derive the classical and best-copy fidelity thresholds for single-photon Gaussian wavepackets, finding that the relevant vacuum-equivalent shot noise is input-dependent, in contrast with quadrature-based continuous-variable quantum teleportation. Our proposal derives from a quantum theory of space-time events, whereafter we establish space as a reference for the time degree of freedom, thereby grounding our work in ongoing research on reference frames and superselection rules in quantum optics and quantum information. Beyond protocols harnessing time-frequency continuous variables of single photons, this work points toward remote synchronization in quantum communication architectures.
AI Summary
The paper proposes teleportation of a single photon’s continuous time–frequency state, asking how arrival time’s dependence on a detector reference affects state transfer. It adapts Braunstein–Kimble teleportation using a time–frequency-entangled photon pair, a nonlinear two-photon mode mixer, joint measurement of commuting time-difference and frequency-sum observables, and conditional displacements. A geometric event-based formulation defines arrival time conditionally on detector position; chronocyclic Wigner functions provide output-state and fidelity calculations. For an ideal entangled resource, an uncorrected detector-reference mismatch $\Delta x$ produces a temporal displacement, with fidelity equal to the input wavepacket’s squared autocorrelation. For a Gaussian pulse of temporal standard deviation $\sigma_t$, $\mathcal F=\exp[-\Delta x^2/(4c^2\sigma_t^2)]$. Thus narrower pulses require tighter reference calibration. Teleporting known probe states and optimizing the correction is proposed for remote calibration. Time–frequency cat and GKP states exhibit autocorrelation revivals that provide structured temporal markers. For fixed-width Gaussian input alphabets, the reported measure-and-prepare fidelity threshold is $1/2$, established through correspondence with quadrature coherent states. A symmetric Gaussian-cloning analysis gives $2/3$; this is not an unrestricted non-Gaussian cloning bound. The effective noise reference depends on input pulse width rather than a universal vacuum scale. The analysis assumes ideal projective measurements, lossless channels, and a perfect nonlinear mode mixer. Efficient implementation of that mixer remains experimentally undemonstrated. Timing jitter, finite spectral resolution, and delay fluctuations are identified as additional noise sources whose quantitative treatment is left for future work.
Artificial intelligent agent \textit{Claude Sonnet 5.5} was used exclusively for checking spelling, typos and grammar.
Central claim may not hold: For the explicitly specified finite Gaussian input prior, the optimal classical fidelity is (1+lambda)/(2+lambda), not 1/2; the reported value requires the unstated broad-prior limit. The claimed guarantee that fidelity above 2/3 excludes a better retained copy fails without a Gaussian-cloner restriction, as demonstrated by the non-Gaussian cloning fidelity approximately 0.6826 acknowledged in the paper. The fidelity expression advertised for a general Gaussian channel omits the squared covariance between time-difference and frequency-sum noise and is valid only when that covariance vanishes.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.