Homodyne Detection of Temporally Resolved Quantum States

1. Problem Statement

Balanced Homodyne Detection (BHD) is a standard tool in quantum optics for measuring quantum states. However, a fundamental challenge arises when dealing with temporally resolved quantum states (states with spectral widths narrower than the detector bandwidth).

• The Mismatch: The quantum state exists in a specific "principal temporal mode" (defined by its generation process), while the homodyne detector naturally measures in a "time-bin basis" (dictated by the detector's sampling rate and local oscillator).
• The Consequence: Continuous homodyne detection effectively projects the single-mode quantum state onto a multi-mode detector basis. This results in a statistical mixture of the target state and vacuum noise across different time bins.
• The Gap: Existing methods often treat mode mismatch as a simple loss parameter. There is a lack of a rigorous formalism and simulation tool to model how this projection affects the reconstruction of the quantum state (marginal distributions and Wigner functions) under realistic experimental errors like timing jitter, phase noise, and modal mismatch.

2. Methodology

The authors propose a new formalism and a simulation algorithm to bridge the gap between the quantum state's natural basis and the detector's basis.

A. Theoretical Framework

• Modal Hilbert Space: The paper treats the time-dependent photocurrent as a basis vector in a modal Hilbert space.
• Basis Transformation:
  • Principal Basis: Defined by the quantum state's generation mode ($f_0$).
  • Detector Basis: Defined by time-bins ($h_k$).
  • Projection: The detector performs a projection of the principal mode onto the time-bin basis. If the modes are not perfectly matched, the measurement is modeled as a mixture of the quantum state and vacuum in orthogonal modes.
• Mathematical Formulation: The photocurrent $i(t)$ is expressed as a sum of weighted modes. The measurement of a specific quadrature is obtained by taking the inner product of the photocurrent with the target mode function: $q_\psi = \langle f(t) | i(t) \rangle$.

B. Simulation Algorithm (Algorithm 1)

The authors developed a numerical algorithm to simulate continuous homodyne traces:

1. Basis Construction: Given a quantum state in a temporal mode $f_0(t)$, an orthonormal basis $\{f_0, f_1, \dots, f_{N-1}\}$ is generated using QR decomposition or Gram-Schmidt.
2. Sampling:
   • The coefficient for the principal mode ($f_0$) is sampled from the quantum state's marginal distribution $P_r(q_\theta)$.
   • Coefficients for all orthogonal modes ($f_{1 \dots N-1}$) are sampled from the background state (typically vacuum).
3. Synthesis: The photocurrent trace is reconstructed by summing the weighted basis functions: $i(t_k) = g \sum_j x_j f_j(t_k)$.
4. Reconstruction: By integrating the simulated trace against the principal mode, the original quadrature sample is recovered.

C. Error Quantification

To evaluate the impact of experimental imperfections, the authors use the Bhattacharyya Coefficient ($B$) to measure the similarity between the theoretical marginal distribution and the reconstructed distribution:
$$B(P_{r1}, P_{r2}) = \int \sqrt{P_{r1}(t) P_{r2}(t)} dt$$
This metric ranges from 0 (distinguishable) to 1 (identical) and is preferred over overlap integrals because it applies directly to probability density functions (PDFs).

3. Key Contributions

1. Formalism of Mode Projection: The paper reframes mode mismatch not just as loss, but as a geometric projection of a quantum state from its principal basis into the detector's time-bin basis.
2. Simulation Algorithm: A robust, open-source algorithm (available on GitHub) that generates realistic continuous homodyne photocurrents for arbitrary quantum states and temporal modes.
3. Error Analysis Framework: A systematic method to quantify how specific experimental errors degrade state reconstruction without requiring full tomography.

4. Results and Findings

The authors applied their algorithm to simulate three common experimental errors:

• Modal Mismatch:

  • Simulated by reconstructing the state using a mode $\tilde{f}(t)$ that differs from the true mode $f(t)$.
  • Result: The Bhattacharyya coefficient decreases as the overlap $\eta = \langle f | \tilde{f} \rangle$ decreases. The effect is state-dependent; for example, low-amplitude coherent states overlap significantly with vacuum, making them less sensitive to mismatch than high-amplitude states or Fock states.

• Timing Jitter:

  • Simulated by introducing random temporal shifts $\tau$ (Gaussian distributed) to the signal mode before integration.
  • Result: As the standard deviation of the jitter ($\sigma$) increases, the reconstructed Wigner function degrades, eventually approaching the vacuum state. The coefficient $B$ drops significantly for non-Gaussian states (e.g., Cat states) even with small jitter.

• Phase Jitter:

  • Simulated by random rotations in phase space ($\theta \to \theta + \delta\theta$).
  • Result: Phase jitter acts as an angular average of the Wigner function projections.
  • Key Insight: Symmetric states (like Fock states $|n\rangle$) are insensitive to phase jitter because their quadrature distributions are rotationally invariant. Asymmetric states (like Cat states or squeezed states) suffer significant degradation in reconstruction fidelity due to phase noise.

5. Significance

• Quantum Computing Relevance: The work is particularly relevant for continuous-variable (CV) quantum computing using Gottesman-Kitaev-Preskill (GKP) states, where precise temporal mode matching is critical for high-fidelity logical operations.
• Experimental Guidance: The provided framework allows experimentalists to:
  • Predict the sensitivity of their setup to timing and phase errors.
  • Determine the necessary stability requirements (e.g., maximum allowable timing jitter) to achieve a desired reconstruction fidelity.
  • Optimize local oscillator shaping to maximize mode overlap.
• Open Science: The authors provide a complete, open-source implementation, enabling the community to simulate and correct for these errors in their own homodyne tomography experiments.

In summary, this paper provides a critical theoretical and computational bridge for understanding how continuous homodyne detection interacts with temporally structured quantum states, offering a practical toolkit for error analysis in high-precision quantum optics.